The Internet Marketing Driver: Glenn Gabe's goal is to help marketers build powerful and measurable web marketing strategies.

sunday, April 25th, 2010

A Baker’s Dozen: A Quick Update on Kati’s Kupcakes, The Winner of The Search a Small Business Holiday Giveaway [PODCAST]

If you’re a frequent reader of my blog, then you probably remember the Search a Small Business Holiday Giveaway I launched this past December. The purpose of the contest was to give an ultra-small business in New Jersey a free online marketing audit, which would produce plan for enhancing the company’s digital strategies.

Continue reading this post>

monday, April 19th, 2010

Conversion Goals and Events in Google Analytics: What’s The Difference and When To Use Them

As online marketing evolves, more and more companies are realizing the power of effectively tracking their marketing efforts via web analytics. I’m finding myself doing a lot more analytics strategy work for clients and I absolutely love it (on multiple levels). I’ve worked with a wide range of web analytics packages since 1995 and it’s amazing to see how the industry and technologies have progressed.

Continue reading this post>

Thursday, March 04th, 2010

Advanced Segmentation in Google Analytics: How to Set Up and Use Advanced Segments to Analyze Social Media Traffic

How to set up and use advanced segments in Google Analytics to analyze social media traffic.When I’m helping clients analyze website traffic and performance, I try and help them avoid the time-consuming process of “report browsing”. Instead, I emphasize entering the process with a very specific goal in mind. You need a purpose when diving into reporting or else you run the risk of spending hours scanning metrics with nothing to show at the end but a headache and a bottle of Visine. I also emphasize focusing on actionable data, or information you can analyze and then make decisions based on. This is why segmentation is so important. I’ll explain more about segments below, but for now think of a segment as a slice of your site traffic (based on traffic source, type of visitor, etc.) For the example I provide in this post, visitors from Social Media websites could be a segment of your site traffic.

Back to web analytics and actionable data. Aggregate data from a broad view of your site traffic doesn’t tell you very much. However, data related to specific traffic sources, locations, keywords, and campaigns can reveal incredible information (and you can act on that data). For example, an aggregate website bounce rate of 70% tells you almost nothing. You cannot take action from that metric alone, since you might have dozens of traffic sources all with varying bounce rates. Some may be low (15-20%), while others may be extremely high (90%+). If you just focus on the average bounce rate at the site level, you won’t be able to make an impact easily. On the flip side, if you had a 70% bounce rate for a specific ad group in paid search (which focuses on a specific theme based on your product line), then you know there’s a problem. That’s actionable data. You can then start to analyze the keywords you are bidding on, the ad text you are using, the landing pages you are driving visitors to, etc. And actionable data impacts conversion, revenue, registrations, and overall campaign performance.

The Social Media Segment
With all the buzz about social media marketing, many companies are trying to figure out how to effectively analyze traffic from social media sites. Sure, it’s easy to see visits from social media sites, but in order to understand the impact of that traffic, you need to dig deeper and have a clearer view. It’s sometimes hard to analyze the specific data you want when several sources of traffic are mixed in your reporting. It can get extremely frustrating to say the least. For example, what content on your site do social media visitors consume the most, how engaged is that traffic segment, how much revenue do they generate, do they return to your site, so on and so forth. So, wouldn’t it be great to isolate that traffic and then run Google Analytics reporting just for that custom segment? The good news is that you can set this up using one of the most powerful features of Google Analytics – Advanced Segmentation.

What is Advanced Segmentation?
Setting up advanced segments in Google Analytics enables you to analyze very specific slices of traffic. Instead of analyzing reporting based on major types of traffic, you can slice and dice the traffic to glean actionable insights. For example, you can set up segments for social media traffic, visitors from specific countries or cities, visitors that searched for specific keywords, campaign traffic, etc. You get the picture. It’s extremely flexible and the segments you choose to set up are based on your specific online marketing initiatives. Once you set up an advanced segment, you will only view data for that segment while you traverse your reporting in Google Analytics. Advanced Segmentation is incredibly handy, and again, you gain actionable intelligence from the reporting for the segment you are analyzing. You can view the Google Analytics help area for more information about advanced segmentation.

Setting Up Your Social Media Segment
Let’s say you’ve been focusing heavily on social media marketing and want to gain a clearer picture of how that traffic is performing. For argument’s sake, let’s say you have a Facebook page and accounts at Twitter, Stumbleupon, Delicious, and Digg. You hired a social media marketer who is managing each account and that person has started gaining traction. Based on your social media efforts, you want to find out as much as possible about how that segment is performing. Sure, you could go into referring sources and view some top-level data for each traffic source, but you want more. You want to drill into several more reports to see what content they are viewing, how much revenue they are generating, which events they are triggering, where they are located geographically, etc. Let’s get started.

How to Set Up Your Social Media Segment in Google Analytics:

1. Log into Google Analytics and find the Advanced Segments dropdown in the upper right-hand corner of your reporting. It will be located above the date range and the default segment will say “All Visits”.

Finding the advanced segments dropdown:
Finding advanced segments in Google Analytics.

2. Click the “All Visits” dropdown and find the link on the left-hand side that reads “Create a new advanced segment”.

Creating a new advanced segment:
Creating a new advanced segment.

3. Now you will see a slick drag and drop interface for creating your custom segment. I love that Google Analytics made this so intuitive. On the left hand side, you will find a list of dimensions and metrics that you can use to create your custom segment. On the right-hand side, you will find an area where you can drag those dimensions or metrics and then define them. Clicking the arrows next to each category on the left hand side will reveal all of the dimensions and metrics you can utilize.

Dragging metrics and dimensions to define a new segment:
Dragging metrics and dimensions to create a new segment.

4. For our purposes, we want to define several referring sources as our segment (various social media websites). Click the arrow icon next to “Traffic Sources” and then drag the “Source” tab to the area that says “Dimension or Metric”. The “Source” tab is located near the bottom of the list under “Traffic Sources”. When you drag the source tab over the box labeled “dimension or metric”, you will see the bounding box change from a light grey dotted line to a dark grey dotted line (indicating that you can drop it there). Once you drop the source tab in the box, Google Analytics will let you type the first few letters of the site in a text field to select the specific traffic source. Google Analytics will also auto-populate the field with your current traffic sources (as you type). So, if you start typing Twitter, you will be able to select You will also notice a “condition” dropdown, which gives you the flexibility for setting matching options. For our purpose, we want to use “Matches Exactly” as we want the exact social media website.

Entering specific traffic sources to define the custom segment:
Entering specific traffic sources to define a segment.

5. Once you set up, you can add more traffic sources by clicking the “Add or statement” link and then dragging another “Source” tab to the dimension or metric box. Start typing Digg and then select Note, Google Analytics will only auto-populate sites where visits exist for your website. So if you don’t have any visitors from, then it won’t show up. You will need to manually enter in the field if that’s the case.

Adding more traffic sources to your custom segment:
Using the add or statement to include more metrics or dimensions.

6. Add traffic sources for Stumbleupon, Facebook, and Delicious as explained above.

7. Name your custom segment by typing in the text field below the drag and drop section you have been using up to this point. You can enter something like “Social Media Traffic” for this example.

Naming your advanced segment:
Naming your advanced segment.

8. At this point, you can click “Test Segment” to see the data that Google Analytics will pull for the segment. The “Test Segment” link is located on the right side of the screen next to your various social media traffic sources. It’s not required that you test the segment, but it’s always a good idea to ensure you set up your custom segment properly.

9. Finally, when you are ready, click “Create Segment”, which is located next to the Name Segment field mentioned earlier.

10. After creating your segment, you will be sent back to your Google Analytics reporting. Note, your new segment will not be active at this point. You will still be viewing “All Visits” until you manually select your segment. To do this, find the “Advanced Segments” area again in the upper right-hand corner of the reporting and click the “All Visits” dropdown. You should see your new segment titled “Social Media Traffic” in the list. You can click the checkbox next to “All Visits” to remove that segment from your reporting and instead check the box next to “Social Media Traffic” to include the segment in your reporting. When you click “Apply” at the bottom of the advanced segments form, you will be able to view only social media traffic in your Google Analytics reports (as defined by your segment). Voila, you’re done.

Now comes the fun part. Go ahead and browse your reporting to view data just for your social media segment. This includes the content the segment is consuming, locations they are visiting from, conversions, revenue, event tracking, return visitors, etc. Pretty cool, right? Note, you can also activate multiple segments at one time to compare them in your reporting. But, that’s for another blog post. :)

Go Ahead, Segment Away…
Although this was a quick tutorial, I hope you have a better understanding of what advanced segments are and how to use them to analyze specific slices of traffic. Based on how flexible and powerful advanced segmentation is, I plan to write more about the topic in future blog posts. So go ahead and log into Google Analytics and create some custom segments. Don’t worry about corrupting your reporting or messing up any of your data. Advanced Segments won’t hurt any of your current profiles or reporting. It simply filters data for you based on the dimensions and metrics you choose. It’s like a segmentation sandbox (less the shovel and pail of course). Have fun.


Labels: , , , , ,

Monday, December 07, 2009

Announcing The “Search a Small Business” Holiday Giveaway from G-Squared Interactive

The Search a Small Business Holiday Sweepstakes from G-Squared Interactive.The holidays are always a great time to reflect upon the past year in online marketing. Looking back at 2009, it’s interesting to analyze how various companies utilized new technologies and marketing channels to increase sales and engage prospective customers. I feel fortunate to be in a position where I get to speak with many marketers from a wide range of companies (both large and small) to learn which tactics they are using to grow their businesses. I think it's been an amazing year, with Search, Social Media, and Mobile attracting a lot of attention from a wide range of companies and organizations.

However, looking back on my conversations and projects over the past year, it’s hard to ignore the lack of resources available to ultra small businesses. These small businesses unfortunately don’t have the time or budgets to tackle online marketing the way larger companies can. They also happen to be a critical component of our economy, so it just doesn’t seem right.

Ultra-Small Businesses & Online Marketing
To me, ultra small businesses are companies run by one or two individuals, employ less than ten people, generate under $500K per year in revenue, and move at light speed to keep their businesses moving. Ultra small businesses are critically important for our economy, but tend to be overshadowed by news from larger brands and companies. As article after article is written about multi-billion dollar powerhouses, the small business owner remains somewhat anonymous. Yet, those very business owners in aggregate employ millions of people and are an important part of the framework of commerce in the United States.

This got me thinking. What could I do this holiday season to help a small business start 2010 the right way? Let’s face it, many small businesses don't have the time to learn about the latest in online marketing and how to leverage those tactics to increase sales. But, just because they don’t have huge budgets and big brand names doesn’t mean they have to be left out in the cold!

The “Search a Small Business” Holiday Giveaway
So in the spirit of the holidays, I decided to launch The “Search a Small Business” Holiday Giveaway here at G-Squared Interactive. Over the next week, small businesses that meet the requirements listed below can simply send us an email to enter the contest. The winning business will receive a free online marketing audit, which includes an analysis of how their current website is performing. And more importantly, the analysis will provide recommendations for improving the website and various online marketing efforts. Insights from the analysis could include recommendations for improving Search Engine Optimization (SEO), Paid Search (SEM), Social Media Marketing, Website Optimization, and Web Analytics. The goal is to help the winning company quickly understand changes that can impact its business. The analysis will be performed by myself and Matt Leonard, an incredibly smart online marketer and good friend of mine that has agreed to help. For those of you on Twitter, Matt is @mjleonard and you should follow him now if you aren’t already. Together, we plan to arm a small business with key information for improving its online marketing efforts.

Please review the following requirements before entering the contest to make sure your business is eligible.

In order to be eligible, you must:
* Have less than 10 employees.
* Already have a website. Since this is an online marketing analysis, we need something to analyze. :)
* Be located less than 60 miles from Princeton, NJ and be a NJ business. This is because Matt and I will present the results to you in person at your office. I recommend you check Google Maps to see if you are eligible.
* Be willing to let us write follow-up blog posts about the giveaway and project. We would like to provide updates about how the winner is using the information provided in our analysis.
* Be willing to make changes! The analysis can only go so far. You will need to act on the recommendations in order to see an impact.

To read the official rules and regulations, please click here.

So let us help your business start off 2010 with the right online marketing strategies in place! Enter now by emailing us at Be sure to include all of your contact information so we know how to get in touch with you, including your full name, business name, business address, number of employees, phone number, and website URL. We will be accepting emails for the contest from Monday, December 7th, 2009 through Monday, December 14th, 2009. We will announce the winner on December 15th on this blog (and directly contact the winner via the information they provide when entering the contest).

Happy Holidays!

Glenn Gabe and Matt Leonard

Labels: , , , , , ,

Tuesday, November 17, 2009

Invalid Clicks and Click Fraud in Local Search Marketing (SEM) - Giving a Whole New Meaning to the Term HyperLocal

The Impact of Click Fraud on Local AdvertisersI’m currently helping several businesses focused on local advertising with both SEO and SEM (Paid Search, PPC). Depending on the industry and market, Local PPC can be both extremely competitive and pricey. Of course, the upside is capturing those highly targeted clicks and turning them into paying customers, which could yield hundreds or thousands of dollars per conversion. When the difference between page one and page two could be significant amounts of revenue, the companies vying to gain those clicks can become hyper-competitive (and that’s an understatement). I’ve heard stories about some companies incorporating clicking through competitor ads as part of their morning routine… That’s not cool, but very real for the industry and market they are located in.

Based on what I just explained above about, the dark side of paid search ends up rearing its ugly head for some local businesses. In highly competitive industries, and in highly competitive markets, click fraud can run rampant. The thought process is simple (and unethical). Eat up your competitor’s budget so you have more of an opportunity to catch highly targeted clicks. As mentioned above, those highly targeted clicks could yield thousands of dollars per day from new customers (depending on the industry).

I think a lot of people have heard about click fraud, but few have actually explored the problem and how it’s affecting their campaigns. For many local businesses attempting to land the ultra-targeted, “ready to buy” customer, click fraud can be a real click, I mean thorn, in their side. That's not good for anyone involved (including Google and the other search engines).

How Big of a Problem is Click Fraud?
So how much of a problem is click fraud for local businesses? It depends on the industry and market, but I’ve seen click fraud rates as high as 35%. Click Forensics publishes the Click Fraud Index and found that the industry average for Q3 2009 was 14.1%. That’s definitely high, but the abnormally high click fraud rates for local search give a whole new meaning to the term hyperlocal. :) Click fraud rates that high can make a serious dent in your budget, put a strain on ROI for your paid search campaigns, and can end up intensifying the overall click fraud problem (increasing the amount of click fraud as some business owners retaliate). So yes, click fraud is a big problem (and can be especially fierce in local advertising).

Defining Click Fraud:
For those of you not that familiar with click fraud, I’ll provide a quick rundown. There are actually several types of click fraud (and reasons for committing it), but I’ll focus on the act of attempting to deplete a competitor’s paid search budget by clicking on their paid search advertisements. And of course there is no intention of taking action on the competitor’s website once clicking through. In a nutshell, it’s Business A clicking on Business B’s ads in order to deplete Business B’s daily budget. Also note that it doesn’t have to be in the form of repetitive clicks from one location. Business A might hire other people or companies to help click on competitor ads (which can be accomplished via click farms, bots, etc.) You can read more about click fraud in Google’s Ad Traffic Quality Resource Center.

Google and Invalid Clicks
Many local businesses running paid search have no idea that Google actually provides statistics on the “invalid clicks” they catch. And by the way, “invalid” is a nice way of saying “click fraud.” :) Google provides some great reporting functionality as part of AdWords and I think too many companies (especially small businesses that are moving at light speed) never tap into the reporting to track campaign performance.

To access the reporting interface in AdWords, you can click the Reporting Tab, and then Reports. Then you can Create a New Report and choose to run a Campaign Performance Report. As part of setting up this report, you can click Add or Remove Columns to customize the report. Then you can click the checkboxes for Invalid Clicks and Invalid Click Rate to view the statistics at an account or campaign level. Depending on your line of business and where you are located, you might be surprised at how many invalid clicks were recorded for your campaigns…

Accessing AdWords Reporting Functionality:
Creating a new report in Google AdWords

Running An Invalid Clicks Report:
Running an invalid clicks report in Google AdWords

What is an Invalid Click?
Google’s system is continually analyzing clicks and looking for patterns that may be fraudulent. For example, clicks from the same IP address, duplicate clicks, clicks from “known sources of invalid activity”, etc. You can read more about how Google calculates invalid clicks in AdWords help. The system is essentially looking for any type of suspicious activity.

Local Showing a Higher Rate of Click Fraud:
Based on running invalid click reports for a wide range of clients, I typically see a much higher percentage of invalid clicks for companies focused on local search. How much higher? Well, comparing invalid click rates across industries, I’ve seen local-centric clients receive 4X to 5X the percentage of invalid clicks. That’s a lot of clicks, and more importantly, a lot of potential money at risk. Now you might be asking, “Glenn, if Google catches the invalid clicks, then the companies shouldn’t get charged, right?” True, but that’s only for what Google catches… Their system isn’t flawless (especially because well-crafted click fraud is nearly impossible to identify). That’s just an unfortunate reality. So, if you see a 20% invalid click rate, it just might be 30-35%.

The Impact on Budget
Let’s add a monetary value to the click fraud problem I mentioned above. In some industries, local businesses are paying $20-$30 per click (yes, you read that correctly). For argument’s sake, let’s say you receive 100 clicks per day at $20 per click. If Google picks up a 20% invalid click rate, and we estimate that it’s really 30% (just for this example), then there is 10% still getting past Google’s filters. So, the 100 clicks coming through are “actual clicks” according to Google (since it won’t charge you for the invalid clicks, or the 20% it caught). Out of the 100 actual clicks that you are being charged for, the 10% of invalid clicks that slip through equate to 10 clicks at $20 per click (or $200 per day). Over a month, that’s over $6000 per month potentially wasted. For many small businesses, that may be too much to overcome. And that’s exactly what the people committing click fraud want to happen. They want to push competitors to the point of quitting AdWords (and paid search in general), which leaves the fraudsters in control of the paid listings. Needless to say, this isn’t good for the paid search industry, the local businesses getting hit by fraud, and of course Google (since Google makes a majority of its money from paid search).

What Can Local Businesses Do About It?
Although click fraud is a big problem, and one that’s hard to overcome, there are some things you can do to stay on top of the problem. I’ve provided a list of recommendations below to help you stay informed, track your clicks, and potentially fight click fraud. The more you understand what’s going on, the more you can develop a strategy for documenting and combating the problem.

Here’s what you can do:

1. Run invalid click reports on a regular basis. This will help you understand how many invalid clicks are occurring, if they spike during certain times, and which campaigns they are impacting. You can also speak with your Google rep (or any rep at AdWords) about the problem, based on the data you collect.

2. Break up your campaigns logically. You can run invalid click reports on an account or campaign level (but not ad group). If you lump all of your ad groups into one campaign, you won’t get as clear of a picture of the click fraud problem impacting your business.

3. Analyze your log files to determine problematic IP’s. Unfortunately, Google isn’t going to provide details about the invalid clicks they find. They will just show you a total number and not reveal who is committing the click fraud. I think that’s unfortunate, but it’s just the way it is right now. But, you can get in touch with your hosting provider (or your IT department) to analyze your server logs. If you competitor is clicking from a specific IP (like their office down the block from you), you might be able to pick it up. Then work with Google and your lawyer on next steps.

4. There are third party solutions that can help you track and identify click fraud. If you believe that your business is the victim of severe click fraud, you might want to go down this path. For example, Click Forensics (which also publishes The Click Fraud Index mentioned earlier) provides services for ad networks, publishers, agencies, and advertisers. There are also several other solutions for tracking fraudulent clicks that are relatively easy to set up. Do some research and demo the various solutions. They could end up saving you a lot of money.

5. Get familiar with Google’s Ad Traffic Quality Resource Center. There’s some good information about click fraud in the center, including an overview of the problem, key definitions, a help center, ways to contact Google’s quality team, etc.

6. Don’t participate in click fraud. Although it should be obvious, contributing to the overall click fraud problem isn’t going to help anything. You should focus your time and attention on running ethical and ROI-driven paid search campaigns and then deal with click fraud legally. Work with Google, your IT staff, your hosting provider, third party solutions for tracking click fraud, and your lawyer in determining the best path to take.

Not All Clicks Are Created Equal
Is click fraud a problem for local businesses? You bet. But you don’t have to sit there in the dark as your competitors click your ads. You should educate yourself about click fraud, stay vigilant, remain white hat (ethical), and analyze the situation to the best of your ability.

As mentioned earlier, Click Forensics says the industry click fraud rate was 14.1% in Q3 of 2009. As a business owner focused on local advertising, you need to decide if you’re ok with that number... Is click fraud just part of doing business in Local PPC or should you fight to save your budget (and the potential customers that would come from that budget?) Like I said earlier, local click fraud gives a whole new meaning to the term Hyperlocal.


Labels: , , , , ,

Monday, November 09, 2009

FaceYahoogle – The Impact of Facebook, Yahoo, and Google on Website Traffic

The Power of Google, Yahoo, and Facebook on Site Traffic
It’s hard to get through a conversation about online marketing right now without bringing up Google, Facebook, and Yahoo (among other popular companies). However, if you’re not heavily involved in online marketing, and you’re not close to the actual referring traffic numbers from Google, Yahoo, and Facebook, then their influence can easily become nebulous. It’s easy to say, “Google is a powerhouse” or “Facebook has 325 million members”, and “You need to be there”, but how powerful are they really?

From a traffic perspective, the three companies are so powerful that I’ve given them their own combined name, or FaceYahoogle. The power of FaceYahoogle ends up becoming very real for my clients after I complete a competitive analysis (which includes identifying major sources of traffic for their company, as well as their competitors). The numbers I present across websites typically show extremely high referral data from FaceYahoogle, and by viewing the actual traffic numbers, you start to get a feel for how much impact the three entities have traffic-wise and potentially revenue-wise.

Digging Deeper into FaceYahoogle
If you’ve read previous posts of mine, then you already know that I’m a big believer in using data versus opinion to make decisions. The power of analytics in online marketing enables you to see granular performance data across a number of key metrics. And the more websites I analyze, the more I see a significant trend across industry categories. I see FaceYahoogle sending large amounts of traffic to a wide range of sites. The abnormally high percentage of traffic coming from Google, Yahoo, and Facebook is not only amazing to see, it’s actually scary. With thousands and thousands of potential referring sites on the web, to see FaceYahoogle send that high of a percentage of traffic is alarming. I think you begin understand how Google built up a $22 billion war chest! :)

I think many people would suspect Google being high on the referring sites list, based on having ~70% market share in search and also having Gmail, Google Maps, Google Docs, etc. However, I’m not sure many know how much actual traffic is coming from Googleland. Also, we hear that Facebook has over 300 million members, which is powerful, but are those members visiting your site via the social network? I’ll answer that question below via screenshots. And then you have Yahoo, with turmoil somewhat cloaking the power of its sites. How much traffic actually comes from Yahoo Search, Yahoo Mail, Yahoo News, Finance, Answers, etc?

So that’s my quick introduction to FaceYahoogle. Now let’s take a look at some numbers! I have provided Compete data (September 09) for a number of popular websites across a range of categories so you can view their referring sources. Note, I know Compete isn’t perfect, but it does provide a wealth of information to analyze for each website (especially for sites that receive large amounts of traffic).

Referring Sites for
31% from FaceYahoogle (and 17% from Google alone…)

Referring Sources for The New York Times

Referring Sites for LinkedIn
36% from FaceYahoogle, and over 8% from Facebook.

Referring Sources for LinkedIn

Referring Sites for
24% from FaceYahoogle

Referring Sources for

Referring Sites for JCrew
31% from FaceYahoogle

Referring Sources for JCrew

Referring Sites for The Huffington Post
33% from FaceYahoogle (and almost 8% from Facebook)

Referring Sources for The Huffington Post

Referring Sites for Yelp
A whopping 55% from FaceYahoogle (and 43% of that from Google!)

Referring Sources for Yelp

Referring Sites for ESPN
25% from FaceYahoogle (and nearly 10% from Facebook)

Referring Sources for ESPN

Referring Sites for
25% from FaceYahoogle (cha-ching…)

Referring Sources for

Referring Sites for
28% from FaceYahoogle

Referring Sources for

Let’s throw in a military site to see how the 3 headed monster works here:
Referring Sites for
Over 40% of referring traffic from FaceYahoogle

Referring Sources for The US Airforce

The screenshots above make it a little more tangible, right? FaceYahoogle is accounting for 40%+ of referring traffic for some websites. If you analyze website traffic often, then you know how insane those numbers are… But that’s not the whole story. The downstream data is important too. It ends up that a large percentage of traffic from these websites is going back to FaceYahoogle. Let’s take a look at just a few from above.

Downstream Data for
26% of visitors leave and go back to FaceYahoogle

Downstream Traffic from

Downstream Data for
31% of visitors leave and go back to FaceYahoogle

Downstream Traffic from

I saw the same trend across the other sites.

So, FaceYahoogle is driving enormous amount of traffic, but it’s also the top recipient of traffic from many sites. In particular, Facebook provides some unique opportunities with regard to downstream traffic. Give your visitors something to take back and you can possibly end up with even more traffic (WOM-based or possibly viral-based). And with some Google and Yahoo traffic going to back to Gmail, Yahoo Mail, Yahoo Answers, etc., you also have opportunities for spreading the word about your products, company, brand, etc. Let’s quickly take a closer look at each part of FaceYahoogle below.

As you can see, Google is an absolute powerhouse, even showing 43% of Yelp's overall referring traffic. That’s outrageous! And it’s not just any traffic, right? Many of the visitors from Google just searched for specific products or services that each site provides (AKA, high quality visitors). Imagine the revenue impact of Google traffic for those sites. In case you are wondering, Google traffic numbers include Search, Maps, Mail, Docs, Video, etc.

Seeing the high percentages from Google across sites, you can start to understand why SEO and SEM have been incredibly hot in online marketing… Some companies survive based on Google traffic alone (via paid and organic search traffic). A slip in rankings can be catastrophic for some websites, with the potential of impacting millions of dollars of revenue. Think about it. If you have 40% of your traffic coming from Google and slip to page two, three, or beyond, you will lose many targeted visitors, and the money they would have spent on your site. So is Google powerful? You bet it is. The numbers combined with my experience tell me so. :)

Facebook has grown by leaps and bounds over the past few years and is estimated to have 325 million members now. Clearly people are signing up in droves, using the platform at a staggering pace (104 billion pageviews based on Compete September 09), and oh yeah, they are visiting websites from Facebook. As you can see in the screenshots above, Facebook ranks in the top five referring sites for many of the properties I checked. Actually, it was typically in the top three. And in case you’re wondering, Twitter is moving up the charts too. Depending on the focus on the site in question, I see Twitter sending large amounts of traffic (and that doesn't count desktop clients which many Twitter members use). On that note, to read an example of how Twitter can impact exposure, traffic, and subsequent SEO power, check out my post about the Twitter effect on SEO. It’s a great example of how Search works with Social.

So, if your company is ignoring social media, then go back through the screenshots above and take note of the percentage of referring traffic from Facebook again. In meetings, I find myself saying more and more that if you ignore social media (and especially Facebook and Twitter), do so at your own risk. Again, the numbers are pretty convincing.

Although Yahoo has taken a back seat recently, the numbers are still strong from a referring source perspective. Between Yahoo Search, Yahoo Mail, Yahoo Answers, Yahoo News, Finance, etc. there are still millions of people visiting each property per month. And yes, those sites end up as top referring sources (impacting traffic, sales, sign-ups, twitter followers, Facebook fans, etc.) Yahoo consistently showed up in the top five referring sites, and often number one or two. Don’t count out Yahoo just yet. If you do, you’d be dismissing a huge traffic source (when you take all of their properties into account).

The Future of FaceYahoogle
I’m sure you are wondering which sites will be the major sources of traffic in 2010 and beyond? Will Twitter beat out Facebook, will Bing surpass Google, will Yahoo be non-existent? The beauty of the web (and technology) is that we never know. But the data does tell us something… don’t ignore Search and Social, and how they can work together.

People are searching and people are talking. And the people that are talking can impact how people that are searching find your website. And people searching can lead to sharing on social networks, based on what they find. Look at the numbers again. Don’t discount Facebook because you think people are tagging photos or playing games all day. You also shouldn’t disregard Google’s dominance. It is too powerful to ignore. And Yahoo isn’t dead yet. There are millions of people visiting Yahoo Sites on a regular basis.

Last, to emphasize that we never really know what will take off, I have provided Twitter’s trending below (Compete data over the past 24 months). I bet many people don’t even know that it was around in 2006 and 2007… and that it crept along until 2008 when it gained serious traction. So, is the next Twitter out there right now slowly growing and about to gain traction? Good question. :)

Click the image to view a larger version:
Twitter Trending 24 Months


Labels: , , , , , , ,

Tuesday, February 24, 2009

The Connection Between High Search Engine Rankings, The Latest Google Heatmap Study, and The Long Tail of SEO: My Guest Post on Search Engine Journal

Glenn Gabe's Guest Post on Search Engine Journal, The Connection Between Page 1 Rankings and The Long Tail.
I’ve been a big fan of Search Engine Journal for a number of years now. They provide excellent search-related posts and information, as well as outstanding coverage of the Search industry. So, as you can imagine, I was extremely excited after speaking with them about becoming a contributor. My first guest post went live yesterday and it details an experiment that I recently ran across a number of websites I manage.

A Custom Google Analytics Filter for SEO
Based on a great blog post by Andre Scholten, I set up a custom filter in Google Analytics to track where each keyword ranked in Google when people clicked through to the websites I tracked for the experiment. This enabled me to view all of the keywords (head, torso, and long tail keywords) leading to the websites I tracked, but also let me quickly see where those keywords ranked in Google when people clicked through. Yes, I could use a number of search tools to run a position analysis on target keywords, but that’s not realistic when you include all of your long tail keywords, since you might be analyzing thousands of keywords at a time.

The Results Were Pretty Darn Compelling…
So I set up a custom filter on a number of websites I manage and waited for the data to stream in. It only took a few hours before keywords started showing up in my reports (along with their rankings). And, I picked up an interesting trend pretty darn quickly... I started to see a strong connection between page one rankings and the long tail of SEO. I’ve written about the powerful long tail of SEO before on my blog, and I believe it’s often overlooked by many people outside of the Search industry. You can start to connect the dots if you add the latest Google heat map study, which found that people are quickly scanning the first few organic results in Google, and if they don’t find what they need, they are refining their searches. My experiment definitely started to connect the dots...

Eager to see the results? Well, you’ll have to check out my guest post on Search Engine Journal to find out more about my experiment and to view the results! It’s a thorough post, and if you have enjoyed the rest of my posts on the Internet Marketing Driver, I think you’ll really dig this one. :)

My guest post: The Long Tail of Page 1 Rankings

Also, definitely feel free to post your comments after reading my guest post. I’d love to hear your own experiences with the long tail of SEO.


Labels: , , ,

Sunday, February 08, 2009

Flash Tutorial, How to Track Flash in Google Analytics Using the Google Analytics for Adobe Flash Component (GAforFlash) Part 2 of 2

Flash Tutorial for Using the GAforFlash Component (AS3)In part 1 of this series I introduced the concept of tracking flash applications, why flash has been hard to track for many marketers, and I introduced the Google Analytics for Adobe Flash Component. To quickly review, the GA for flash component enables you to easily make calls directly from your ActionScript code in order to track pageviews, events, conversions, etc. in flash. It’s an important step for understanding how your flash elements contribute to the success of your website, and not just because you think it’s a killer flash app. ;-) In part 2 of my series, I will walk you through how to actually set up flash tracking using the gaforflash component (step by step in the flash authoring environment.) By the end of this tutorial, you should have a solid understanding of using the flash component and making calls to Google Analytics from within your ActionScript code. I will reference the flash movie that I created during my own testing and show the resulting Google Analytics reporting to tie it all together. So without further ado, let’s start coding. :)

Quick Disclaimer, Code Ahead:
As I mentioned in my first post, if you have some knowledge of developing flash movies and writing ActionScript code, then you should be able to follow along. If you aren’t familiar with developing in flash, grab your flash developer and possibly your web analyst and set up a working lunch. By the end of the tutorial, I’m confident you will see the power of using this technique to track your flash elements.

Visit the Google Code Project and Download the Components
First, visit the Google Code Project for Google Analytics for Flash (gaforflash). Click the downloads tab at the top of the page and download the zipfile listed. The current version as of this blog post was v1.0.1. Once you download the zipfile, unzip the contents to a directory on your hard drive. Open the readme textfile and follow the instructions for copying the components to the proper directory in your Adobe flash folder. This will differ depending on if you are running a pc or a mac and you will be creating a new folder within the components directory where you will copy the files (the instructions tell you to create a Google folder and drop the components there). Once you copy the components into the new directory, then go ahead and launch flash. Again, make sure you follow the readme before launching flash.

The Google Code Project for GaforFlash:
The Google Code Project for GAforFlash

Determine What You Are going To Track:
To me, this is the most important step (and I bet the web analysts reading this post agree!) Mapping out what you want to track is essential to having clean reporting and a structured hierarchy. You can really have some messy reporting without working through this step… We’ll keep this example very simple to keep the amount of coding down, so here’s the scenario.

You will be adding a new flash element to a category page on your website and want to track how visitors engage the flash movie and how that flash element contributes to the success of the website. In our sample flash movie, there will be a start button, which launches the rest of the flash movie. The start button is there for a reason and you’ll learn why in a second. Once someone clicks the start button, they will be presented with two product thumbnails. The goal of our sample flash movie is to get visitors to learn more about each product and then click an email button at the bottom of the flash movie to get in touch with sales. Again, this is completely made up and simple, but we need to map this out in order to know what to track. Also, we'll track when users hover over each thumbnail and then when they click each one to reveal more information. In addition, they obviously want to track the email link at the bottom of the flash movie, since clicking the button will be a conversion in GA.

To summarize the key pageviews, events and conversions we will track:
1. A start button will trigger a pageview so you know how many people engaged the flash movie. This is so you know that the flash movie was triggered (and not just sitting on the page).
2. The two product thumbnails will trigger events when someone hovers over the thumbnail and when someone clicks them.
3. The email link at the bottom of the flash movie will trigger a pageview for when someone clicks the button. That pageview will also trigger a conversion.
4. Note: you will need to set up a conversion goal in Google Analytics for the email pageview that we trigger when visitors click the email button. This is easy to do and then will start showing up within your Goals tab in GA. You can read more about tracking goals in the GA help center.

Open Up Flash and Create the Necessary Assets
In order to work though this tutorial, you will need to create some simple flash assets (buttons). Don’t worry about how they look. The core point of this tutorial is that you learn how to use the flash component to make calls to Google Analytics and not to win design awards. ;) You will need to create a start button, a product thumbnail button, and an email button (which can be simply text if you want). Once you have quickly created each button, proceed with the rest of the tutorial.

Import the Tracking Libraries
First, make sure your current flash movie is targeting ActionScript 3. Open up the publish settings dialog box (control shift F12), click the flash tab, and use the dropdown to select an ActionScript version to target. Choose ActionScript 3. Next, you need to drag an instance of the AnalyticsLibrary Component to the stage in order to import the code libraries. Create a new layer in flash and open the components panel (control F7). You should see a category named Google (which you created earlier in this tutorial). If you don’t see the Google category, then go back to the beginning of this tutorial and follow the directions again in the readme text file that was part of the download. Click the plus sign (+) next to Google and drag the AnalyticsLibrary component to the stage. Don’t worry where you place it on the stage. It should now be present in your project library (you should also see it listed in your library. Click Control L to see your project library.) Keep in mind I’m referring to your project library, which holds all of the assets you create in your flash movie (buttons, movie clips, images, components, etc.) I’m not referring to code libraries, which we will discuss shortly.

Finding the AnalyticsLibrary Component in Flash:
Drag the AnalyticsLibrary component to the stage.

Next, create an Actions layer and select frame 1 of the timeline in that layer. Click F9 to open the Actions window (which is where you write ActionScript code). Since we are using the Analytics Library Component to make calls directly from our ActionScript 3 code, you will need to import the libraries that you will use to instantiate a tracking object and make calls to Google Analytics. If you don’t import the libraries, you will not be able to make calls to GA.

Here is what you should add to frame 1 of the actions layer:
var tracker:AnalyticsTracker = new GATracker( this, "UA-111-222", "AS3", true );

A Quick Explanation of the Code:
Note, you would obviously want to add your own GA account number so the data is sent to the correct profile. i.e. Don’t keep 111-222 as the web property ID. The first two lines import the necessary code libraries and the third line of code initializes the tracking object. Keep “this” as the first parameter, which references the current display object. Enter your own GA account number for the second parameter, you can keep AS3 as the third parameter for this tutorial, and keep “true” as the fourth parameter. That sets the debug mode. When true, you will see a trace of all tracking events occurring. When false, this will not be visible. Also, you can read the gaforflash documentation to learn more about each parameter.

Making Calls From ActionScript to GA:
Now, if you create a start button that visitors have to click in order to see anything in your flash movie, then we can logically target that click as a pageview. Then we can view in GA how many people started the flash movie. That would answer one of the first questions from your CMO, right? :) So create a new layer in your flash movie called Start Button and add your button to frame 1 of this layer. To stop the flash movie from playing before someone clicks the start button, simply add a stop action on frame 5 in your Actions layer. Create new keyframe on frame 5 and enter this.stop(); as the code. Again, the intent of this tutorial is not to teach you how to use flash, but I’ll add some tips as we move along. Select the start button on the stage by clicking it once and give it an instance name of start_btn in the properties panel. You can click Control F3 to bring up the properties panel if it’s not on screen. We need to give the button an instance name in order to target it in ActionScript. If you don’t give it an instance name, your code will not work.

Creating an Instance Name in the Properties Panel:
Give your button in flash an instance name.

Now go back to frame 1 of your actions layer and open the actions window again (click F9). Below the code we first added (importing the libraries and initializing a tracking object), add the following lines of code. I will explain them in a second… You can give yourself some space by clicking enter a few times. Your code does not need to be lumped together!

//functionality for start button
start_btn.addEventListener( MouseEvent.CLICK, startExample );

function startExample( event:Event):void {
//we know that the user engaged the flash movie (they clicked start)

The Code Explained:
The first line is a comment, which is good programming practice. You can simply describe the code that follows. This will not be visible to anyone but you as a programmer. The second line is an event handler for the start button. It targets the instance name start_btn, which we set up earlier. This is why we needed to create an instance name. Now our code will look for a mouse event (CLICK) for the start button and then trigger the function called startExample, which I will cover in a second. Event handlers enable you to react to events in your flash movies. They are critical to creating advanced functionality in your flash applications.

The function startExample() will be called when someone clicks the start button. This function uses our tracker object to trigger the trackPageview method and then plays the main timeline in your flash movie, using the play(); action. Remember, we want the visitor to see the rest of our flash movie after clicking the start button. The trackPageview method tells Google Analytics to track a specific pageview when something happens. The page will show up in GA reporting as /GAFlash/Start in your content tab, as if someone actually visited a page on the website. Pretty cool, right? So, you’ll be able to go into GA and click the content tab and see how many times this “page” was triggered. Nice. That wasn’t so bad, was it? That was all done in less than 15 lines of code.

Tracking the Product Thumbnails:
You will use the same methodology for tracking the two product thumbnails in your flash movie. Remember, we want to know when someone clicks each thumbnail, but we’ll be adding one more event…when someone hovers over the thumbnails. You might find that people were interested enough to hover over a product, but not trigger it. If you see enough of this behavior, you might want to dig deeper to find out what’s causing it. That’s just a simple example and you should work with your web analyst to determine what to track for your specific website and flash content.

A Screenshot of the 2 Product Thumbnails for this Example:
Adding the product thumbnails in flash.

We used the trackPageview method for the start button, but we will use trackEvent for the thumbnails. Examples of events might be clicking a button, hovering over that button, visiting a specific area of a flash movie, etc. GA now provides an Event Tracking tab within your Content tab. Within the Event Tracking tab, you can view categories of events, the specific actions users took, the labels associated with those events, trending, etc. So, we’ll track two events with each product thumbnail in our flash movie. I will only cover the first thumbnail here and you can copy this process to apply event tracking to the second thumbnail.

After someone clicks the start button, you used the play(); action to play the main timeline. At frame 20 on your main timeline, create a new stop action in your actions layer. Add a new keyframe at frame 20 and open the Actions windows. Enter this.stop(); as the code. Then create a new layer for your product thumbnails and add the buttons that you created earlier in this tutorial. Note, for this example, add them to frame 1 so your code can target the button instances. You can place them off the stage (off-screen) in frame 1 so they aren't visible and then move them on-stage later in the flash movie when visitors need to see them. If they aren't on the stage in frame 1, the event handlers you create targeting these buttons won't work. Give each instance on the stage a unique instance name, like product1_btn and product2_btn. You do this by selecting each button and then entering an instance name in the properties panel. Again, we need to give them instance names in order to target them in our ActionScript code. Once you do this, go back to frame 1 of your Actions layer, click F9 to open the Actions window again and add the following code:

//functionality for product btn 1
product1_btn.addEventListener( MouseEvent.CLICK, clickProduct1 );
product1_btn.addEventListener( MouseEvent.MOUSE_OVER, hoverProduct1 );

function hoverProduct1( event:Event):void {
tracker.trackEvent("Products", "HoverProduct1");

function clickProduct1( event:Event):void {
tracker.trackEvent("Products", "ClickProduct1");

So we have two event handlers for the product1_btn. The first event handler will handle the CLICK mouse event and the second will handle the MOUSE_OVER mouse event. MOUSE_OVER is when someone hovers over the button (as you probably guessed). Checking the clickProduct1 function, you see that we are using trackEvent to send an event tracking call to Google Analytics. The two parameters are Category and Action respectively. I lumped both product buttons under the same category called “Products” and then gave specific events to each click and hover (HoverProduct1 and ClickProduct1). You will be able to drill into event categories in your GA reporting and then see specific actions, along with trending.

Important: When you set up the code for the second product button, make sure your event handlers target product2_btn versus product1_btn and that you trigger specific functions, such as clickProduct2 and hoverProduct2 versus clickProduct1 and hoverProduct1. You definitely don’t want to call the wrong functions, as this will ultimately skew your reporting or throw errors in your flash movie. i.e. Someone might click the profile 2 button, but you triggered the profile 1 tracking.

The Email Link, Our Conversion in Flash
Last, but not least, we want to track when people click the email button. We have decided that clicking this button will be a conversion for our flash movie. Create a new layer in your flash movie and add your email button to frame 1 of this layer. Select the email button on the stage by clicking it and give it an instance name of emailMe_btn. Then go back to frame 1 of your Actions layer and open the actions window (hit F9). Below the code we added earlier, add the following lines of code:

//functionality for email button, this is also our conversion
emailMe_btn.addEventListener( MouseEvent.CLICK, onButtonClick );
function onButtonClick ( event:Event ):void

OK, so when visitors click the email button, we will trigger the trackPageview method to track that click as a pageview in Google Analytics. The page will show up as /GAFlash/EmailMe in your Content tab. In addition, since this is a conversion, you can set up a conversion goal in Google Analytics targeting that page. Then conversions will show up in your Goals tab within Google Analytics. Now you can tell your CMO how many visitors are converting within your flash movie. That’s not as generic as “Sorry, we don’t know”, right? ;-)

Quickly Test and Track Your Flash Movie
In the first piece of code we added in this tutorial, we included this line:
var tracker:AnalyticsTracker = new GATracker( this, "UA-111-222", "AS3", true );

As mentioned earlier, the “true” parameter tells flash whether you want to run debug mode. If you set this to true, then you can see a trace of your calls to Google Analytics. I recommend turning this on during your testing and then when you are ready to go live, you can turn if off by setting the parameter to “false”. Go ahead and test your movie and click the start button. You should see a message when the call is made to GA, and if it was successful. Then as you interact with your product thumbnails, you will also see the calls being made. Last, as you click the email button, you should also see the call being made. If all looks good, then you can add your published flash movie to an html page and then upload the files to your website. Remember to set debug to false as mentioned earlier or else everyone viewing with your flash movie will see the trace of your calls. Then click away and thoroughly test out your flash movie, try different browsers, platforms, etc. Make sure you trigger each of the elements enough and wait for GA to show you the results in your reporting.

Enabling Debug Mode to View a Trace of Your Calls:
Visual debugging with the gaforflash component

My Working Example of Using GAforFlash
I uploaded my example of using the Google Analytics for Adobe Flash Component in case you wanted to see how it looked and worked. I know it's hard sometimes to follow along without seeing the final product.

Checking Your Google Analytics Reporting
I’m going to show you what the reporting looked like for my own example. The first thing I did was click the Content Tab in Google Analytics and then Event Tracking within that tab. I immediately could see the total number of events that took place, visits with an event, etc. {See screenshots below.} Clicking the Categories tab displayed the event categories we set up earlier, such as Products (Note, I used UserProfiles in my example versus Products so that's what you will see in the screenshots below). This category includes the actions of hovering and clicking the product thumbnails. This is why mapping out what you want to track is important. You want a clear hierarchy in your reporting. Drilling into each category, I could see the actions that took place like HoverProfile1, ClickProfile2, etc. Note, I used the word "profile" in my testing versus products. If you click the Actions tab (below categories), you will see all of your actions listed, regardless of category. We didn’t add the optional parameter for Labels so you won’t see anything there (in case you were wondering).

The Event Tracking Overview Report in Google Analytics:
Viewing the event categories in Google Analytics

The Events Category Page in Google Analytics:
Viewing the event categories in Google Analytics

The Event Actions Report in Google Analytics:
Viewing the event categories in Google Analytics

Pageviews and Conversions
Remember we wanted to know how many people started our flash movie (after clicking the start button)? Well if you go into the Content Tab and check top content, you will see the /GAFlash/Start page listed. The number of pageviews shows how many times visitors started the flash movie. In addition, you can view /GAFlash/EmailMe, which shows the number of times the email button was clicked. And since we set this up as a conversion, you can click the Goals tab and see your TrackFlash conversion listed. Drilling into that conversion goal will also reveal conversion rate. And since it’s a conversion goal, you can see conversion by traffic source, campaign, keyword, etc. That’s the benefit of setting up conversion goals… So, you might find that organic search traffic converted 10x as much as paid search, or that visitors arriving via your email marketing didn’t even trigger your flash movie, so on and so forth.

Viewing Flash Pageviews in the Top Content Report in Google Analytics:
Tracking flash pageviews in Google Analytics

Viewing a Flash Conversion in Google Analytics:
Tracking flash pageviews in Google Analytics
OK, we've definitely covered a lot in Part 2. Either your head is reeling or you're excited to use the gaforflash component in your own projects (or both!) Take your time and test our different scenarios. Work with your flash development team and your web analysts to map out how to best track your flash apps. You never know, it very well could lead to more flash application work… like if you told your CMO that 25% of the people interacting with your flash movie converted! :) Imagine if you could glean insights from your flash content versus it just being slick and pretty. And I’ll take data over pretty any day of the week. ;-)


Read Part 1 of this Series on Tracking Flash in Google Analytics

Labels: , , ,

How To Track Flash In Google Analytics (GA) Using The Google Analytics For Adobe Flash Component (GAforFlash), Part 1 of 2

Tracking Flash in Google Analytics Using the GAforFlash ComponentI wanted to start this post with some quick points about flash:
* flash content can be extremely engaging.
* flash apps are a great way to create something viral.
* flash is currently used extensively on the web for everything from video to games to product demos.
* Creative Directors love flash.
* Brand Managers also love flash.
* But unfortunately, flash has been hard and confusing to track...

I apologize if you were feeling pretty good about flash until the last bullet! ;-) That is the reality, though.

Why Has Flash Been Confusing And Hard To Track?
Let’s take a quick look at why flash tracking has been an issue. First and most basic, there are many people that don’t know you can even track flash applications. That’s a tough obstacle to overcome, right? Second, the Analytics team is sometimes not involved during the planning of flash-based projects or campaigns. I’m sure some web analysts reading this post probably believe that’s an understatement! Third, you need to coordinate mechanisms for tracking flash with both flash developers and analysts. Fourth, there has been confusion about how to actually track flash even if you already have your flash developers and analysts in the same room. Last (at least for this initial list), some people will focus on the end result (conversions only) and not track the specific elements on the website that lead to the conversion. i.e. As long as people buy something or sign up, I don’t care what they do on the site. As you can probably guess, I’m not a big fan of the latter… I think the more information you can gather about how visitors interact with your website, the more you can optimize the website to increase conversion (whatever conversion is for your specific business).

Yes, You Can Track Flash
I’m here to tell you that you can track flash and you can see how visitors are interacting with your flash applications. The approach I am going to show you is a more elegant method for tracking flash than what's been used in the past. My hope is that this two part series can save you from a potentially embarrassing moment. You know, when your CMO asks how the 6 month flash development project that cost $75K, that looks incredible, is highly engaging, and wins awards is contributing to the success of the website. Unfortunately, many people run for the hills at that moment, or act like they didn’t hear the question (Chevy Chase-style). I’d hate for you to say, “We’re not actually sure Mr. CMO…” ,and I’d rather hear you say, “Absolutely, here’s detailed reporting of how visitors are engaging our flash content, as well as the tangible effects on conversion.” Yes, I want you to be the flash analytics rock star. :)

Introduction to the Google Analytics for Adobe Flash Component (GAforFlash)
The GA for flash tracking component enables you to track specific events and functionality in your flash movies and seamlessly communicate with Google Analytics for tracking. It gracefully handles any DOM issues that could cause problems in other types of flash tracking using GA. It’s an open source initiative between Google and Adobe Systems and you can find more information on the Google Code Project website.

OK, But What Is It Exactly?
There are two components you can use in flash that enable you to track events, pageviews, conversions, etc. One is a simple flash component that you can customize in the component inspector in flash and the other component involves importing the tracking libraries into your project and then making calls to GA from within your ActionScript code (AS3). I’m going to cover the second approach, since it gives you the most flexibility. And don’t get scared with the way it was explained above. It’s actually straight forward if you are comfortable working in flash.

A Quick Tip For The Code-Averse:
I’m going to walk you through my example step by step in Part 2 of my series on tracking flash in google analytics. If you have worked in flash and written some ActionScript code, I’m confident you will be able to follow along. If not, grab your flash developer and your web analyst and then schedule a working lunch. You can all walk through my example together as a team. After going through my second post, I have a feeling you will collectively brainstorm several ways to use the GA for flash component to track your flash elements, websites, and applications.

So Are You Excited To Start?
Good, then you’re ready to read my second post, which covers how to track flash in Google Analytics using the GA for flash Tracking Component.


Labels: , , , ,

Monday, January 12, 2009

Check Your Search Engine Rankings, Why Your Competitors in Organic Search Might Not Always Be Who You Think They Are

Finding your actual competition in organic search.When beginning a new SEO project, there are some questions that always come up during initial meetings. How does organic search work? Which keywords should we target? Do we need to redesign our entire website? And…how do we compare to our competition in natural search? I’m going to focus on the last question in this post, because there’s an important point I’d like to make. Whenever I ask someone who their competition is in natural search, I typically hear the names of their core competitors (business-wise). Although that’s true in a pure business sense, that’s not necessarily the case in natural search. So, I often run a competitive position analysis to determine where a site ranks in the search engines as compared to its competition. It helps you (and your client) understand who their actual competition is and then sets the stage for deeper competitive analysis.

Don’t drop names with Google…
Outside of search, you might be able to throw a big brand name around and get somewhere. Unfortunately, the search engines don’t necessarily care. That’s one of the reasons you’ll see all types of websites ranking for highly competitive keywords. Actually, I’d argue that some smaller online businesses can easily outmaneuver larger websites and companies in SEO. When it comes down to it, the engines care about quality content, a good user experience, relevancy, and popularity. In other words, create outstanding content that can be easily crawled and indexed, optimize that content based on keyword research, make it easy for your visitors to find and use your content, and if those visitors find that content valuable, you might gain important inbound links (AKA votes). If that happens, subsequent rankings can follow… BTW, you’ll notice I didn’t mention that you need to be a big brand or a multi-billion dollar company to do this. That’s part of the reason blogs have become so powerful. They give the small guy a voice…and that small guy can often outrank large companies in the SERPs. Empowering, yes? Scary to large businesses and big brands, you bet.

Seriously? That’s My Competition in Natural Search??
Yes, I hear this often (with a few other words that I cannot put on my blog!) Once you run a position analysis using competitive keywords (based on keyword research), you and your client can clearly see who owns the SERPs for those keywords. Sure, the rankings can change over time, but you have a snapshot of which sites are ranking at that point in time. Then, you can take the next step and perform a competitive analysis to help you determine what type of content ranks, how the websites structure their content, and which sites link to them. Remember, quality and relevant inbound links are the lifeblood of SEO.

Let’s take a look at a few examples:
Note, since search engine rankings change often, you might not see the same exact results that I did during my test.

Flatscreen TVs
HD TV’s are obviously hot, so let’s check out a competitive keyword like flatscreen tvs:

Google search rankings for flatscreen tvs.

This is a great example. There are only 2 manufacturers in the top 10 (Samsung at #7 and Westinghouse Digital at #10, which isn't visible in the screenshot above). More on Samsung in a minute. The rest of the sites include reviews, news stories, a buying guide, an e-commerce website, etc. I’ll bet if you walked into LG, Panasonic or Samsung and asked who they compete with in natural search for a keyword like flatscreen tv's, you would hear the other big brands and not the sites in this list. Note, Samsung was pretty smart with creating the page that ranks on their website. Someone at Samsung (cough, cough, SEO guy), understood what people are searching for and provided that content on (Pagerank 8...) There are other good things about this page that I'd like to cover, but that’s for another post. :)

Cabernet Sauvignon
Any wine drinkers out there? Imagine you owned a winery and had an award winning cab? You would probably want to rank highly, right? Let’s see which sites rank for the keyword cabernet sauvignon:

Google search rankings for cabernet sauvignon.

Wow, I’m not exactly a wine connoisseur, but I don’t see any popular wine brands here (other than in the shopping results, which I'll tackle in a minute). You have Wikipedia (big surprise),, an article about Obama, and then a spattering of other wine-related educational pages. Needless to say, this list of websites is probably not what a leading winery would expect to find ranking for cabernet sauvignon.

Enter Universal Search: Also, in the middle of the page you will see shopping results listed. This is Universal Search in action, where Google is mixing additional types of results within the organic rankings. More on this below, but you should start to think about all the different ways you can rank in organic search beyond traditional webpage content. For example, video, images, news, shopping, local, etc.

HD Video Camera
HD Video is all the rage, let’s take a look at the keyword HD Video Camera:

Google search rankings for hd video camera.

Very interesting. There’s only one big brand in the list (Canon at #4). The rest of the list includes reviews at cnet, an announcement from, two YouTube videos (more on this in a minute), some news results, and then Again, if you walked into Sony, Panasonic, or Canon, do you think they would guess that they are competing against YouTube videos? Probably not. On that note, you can see Universal Search in action here again, with two video thumbnails in the organic results (at least at the time of my test). One is from Tiger Direct and the other is from Chris Pirillo! Great job Chris, you outrank major manufacturers of HD video cameras. :)

So, if you haven’t started thinking about Universal Search and the impact that it can have, just take a closer look at the screenshot above. I think you’ll change your mind. There's also a news result right under the video thumbnails. Both the video results and news results are powerful, especially since they have thumbnails associated with them. If you are interested in learning more about optimizing your video content, then check out my post about Video SEO.

Fuel Efficient Cars
Based on the spike in gasoline prices during 2008, let’s check out a search for fuel efficient cars:

Google search rankings for fuel efficient cars.

Holy smokes, there’s not 1 car manufacturer in the list. Not 1. I highly doubt that Ford, Honda, Toyota, Chrysler, and GM would take me seriously if I walked in and said that they compete with a government agency, a green publication, and a newspaper in natural search! That said, you need content on your site in order to rank... I was shocked to see several car manufacturer websites without a single occurrence of the keyword fuel efficient cars. I had to check a few times to believe it...

You will also see another example of Universal Search in the screenshot above. Google is providing news results mixed in the organic rankings (in the middle of the page). This is just another reason to start thinking about all of the ways to rank in organic search (and the different types of content you can optimize). All of your digital assets come into play with universal search.

So, are you ready to conduct a position analysis?
OK, I think you get the point. Performing a position analysis is an important step in understanding your actual competition in natural search. I would begin the process by identifying your competitive keywords via extensive keyword research and then determine where you rank against your competition for those keywords. Then, once you know the competition, you should complete a thorough competitive analysis to see how you can strengthen your organic search power and increase your search engine rankings.

Good luck and be ready for some interesting looks as you tell people who they really compete with in natural search. :)


Labels: , , , ,

Thursday, November 06, 2008

Web Analytics and Tracking Your Online Marketing Campaigns, Why Starting With a Basic Analytics Foundation is a Smart Way to Go

Web analytics, basic setup and strong foundation.You woke up earlier than usual this morning, eager to start the day. This is not your typical week... You’ve got a new product launching and you are having some final meetings to make sure everything is covered with your online marketing campaigns. In one of your last meetings before the launch, your CMO walks in and says, “Great work on developing the campaign and I’m excited to see the results. When can I expect to see some reporting detailing how each channel is performing?” Crickets chirp… {Since this is my blog post, I have the power to freeze time for 10 minutes so I can explain more about web analytics and help you craft your answer to your CMO. Please continue reading.}

There’s no reason that tracking online marketing campaigns should be an issue, although unfortunately, many times it is. There is a lot of talk about bleeding edge web analytics, and believe me, I’m excited about those advancements. But I would be careful with how you implement your web analytics package, or more importantly, how fast you move to an advanced tracking setup.

I think everyone would agree that it's never a good thing when campaigns go live without the proper tracking and measurement in place. It actually pains me to see that occur…especially knowing how some basic reporting can provide powerful and actionable insights. That's right, I said basic reporting and actionable insights in one sentence. For example, wouldn’t you like to track each aspect of your campaign to see which ones perform best? You can use this data to help you determine how to proceed in future campaigns (or even how to tweak current campaigns that are live). Is paid search generating the most revenue, which categories of keywords within paid search are driving that success, is email marketing generating high click throughs, but low conversion? Are your product pages ranking in natural search, how much traffic is coming from Google, and what’s the conversion rate for specific categories of keywords from organic search? If you think that these questions are hard to answer, you are wrong. I’m here to tell you that with some relatively basic tracking in place, you can find out answers to all of these questions, and more importantly, you can pass those insights to senior management at intervals during the campaign. As you can imagine, having campaign data is extremely valuable (even when it's negative). And, that information is easily digestible by all levels of the organization.

Start with a stable and accurate web analytics foundation and build upon it…
Here’s a quick analogy. You just spent thousands of dollars buying state of the art windows for your home. In addition, you decided to put in hardwood floors throughout the house. That’s great, but you’ve got a small problem. Your foundation is badly cracked. Considering that you need to spend a lot of money trying to fix your foundation, now how do you feel about adding all of the extras? You suddenly don’t seem to care, right? Web Analytics is the same way. What good is jumping to advanced levels of tracking when you can't even get basic performance data?? That's why I always recommend starting with a relatively basic implementation. Then, make sure your reporting is accurate and providing you with actionable information. Once you have a solid web analytics foundation in place, you can enhance it and test the new functionality in bite size pieces. For example, advanced segmentation, event tracking, tracking visitor engagement, implementing an advanced testing platform, etc. I don’t recommend jumping into the most advanced analytics setup right out of the gates. I can tell you with almost 100% certainty that you won’t be in a good place. There will be confusion, disappointment, frustration, and then you’ll probably revert to the basic setup like I recommended in the first place! By that time, you might have wasted countless hours, days, and months trying to get the advanced setup working. Even worse, there may be people in your company that have been using the reporting to make decisions... and decisions based on poor data is not good, to say the least.

What type of information can you get from relatively basic reporting?
Let’s go through a hypothetical campaign so you can see what I’m referring to. Maybe you have a new version of a product launching soon. You’ll be running paid search, display advertising, email marketing, and then optimizing the new section of the website for organic search. You’ve decided to use Google Analytics to track your campaigns and have installed the tracking code on each page of your website. For our example, there are two conversions, an e-commerce sale and an email list signup. You will be running paid search in Google and Yahoo, your display advertising is running on a number of industry-specific websites, you will be blasting out several email campaigns to your segmented in-house list, and you’ve optimized your new pages based on keyword research for natural search. With Google Analytics in place (a package I’ve written about often), you will be able to track each aspect of your campaign to determine the effectiveness of your efforts. With the proper tracking in place, you won't be surfing your web analytics reporting aimlessly for hours. Instead, you will be able to drill into GA and pull relevant information that can help you understand what worked and what didn't.

But Glenn, how do I track my campaigns?
That’s a good question and one I hear often. You’ve already added your GA code to your website, which is the first step. The next step (for our example) is to make sure GA tracks conversions and then revenue. You can learn how to set up conversion goals in Google Analytics here. After you learn how to set up conversion goals, you can read about how to set up e-Commerce tracking. It’s not hard to do and should take your developers a relatively short amount of time to set up. When that’s completed, you will be able to see conversions and revenue by channel (Paid Search, Natural Search, Email Marketing, Display Advertising, etc). Even better, you can drill into your campaigns to see which ad groups are driving the best performance, which keywords, which email creative, which creative elements are working best, etc. For example, you might find that one version of your email creative outperformed other email creative by 65%. That’s the type of powerful information you can glean from even a basic setup like this.

Now, GA natively tracks your AdWords campaigns so you are covered there without any additional tagging. For your other campaigns, you will need to tag your creative using GA’s tracking parameters. You can learn more about how to tag your links here. They are basically querystring parameters that enable GA to identify specific campaigns, and then will enable you to run reports on what those visitors do on your site. So for our example, you would want to tag your Yahoo paid search campaigns, your email marketing campaigns, and your display advertising so they can be uniquely identified by Google Analytics. BTW, I’ve written a post about how to tag your email marketing campaigns so you can track each element clicked in your emails. I’ve also written about tagging YSM campaigns using dynamic variables. Once you tag your campaigns, you can access your reporting within the Campaigns tab in Google Analytics (under Traffic Sources).

But can I really track valuable information with this setup?
You bet, but I’ll let you be the judge. Take a look at the bullets listed below and mark down how many you think would be valuable or important when tracking your campaigns:

*Visitors from each channel and then each campaign within that channel. i.e. Paid Search campaigns, email marketing campaigns, banners, etc.
*Conversions and conversion rate by channel, by campaign, search engine, keyword, email creative, banners, etc.
*e-Commerce revenue by channel, campaign, search engine, keyword, email creative, banner, etc.
*Bounce rate of all campaign landing pages.
*Exit rate of pages within your site and campaign section.
*Conversion Funnel analysis, or where people drop off when trying to complete a conversion.
*Trending over time per channel (and per campaign within each channel.)
*Top products and revenue during the campaign time period.
*Referring websites that are driving traffic to your campaign landing pages, including conversions and revenue from those traffic sources.
*Which geographic regions generate the most revenue or conversion.

I can keep going, but I’ll stop there.

How many bullets did you identify as valuable? I hope all of them (or at least most of them). Once you have this data, you can easily compare the reporting to previous campaigns, you can use it to refine the current campaign (on the fly), or use it to improve future campaigns. It’s actionable data. For example, you might find that display advertising cost you $50,000 and generated only $10,000 in revenue. Drilling into your display advertising, maybe two websites outperformed the others by a huge margin. Maybe you’ll find that paid search generated a 350% ROI. Drilling in further, your brand keywords accounted for most of the revenue and you already rank in natural search for those keywords, so do you need to run brand terms next time? Maybe your email marketing generated a lot of click throughs, but almost no conversion. You also notice a 90% bounce rate from email. Why?

OK, I think you get the picture. You will gain all of the information I listed above, and more, by using a fairly basic analytics setup with some minimal tagging. Can you see why it’s frustrating to some people in web marketing when campaigns go live without the proper analytics setup or tracking in place? Now, would I love to track even more than this by using an elaborate web analytics setup? You bet, but compared to having no tracking in place or unreliable tracking, I would be happy with this level of reporting! Wouldn’t you?

Back to your CMO for a second:
{Now I will unfreeze time so you can answer your CMO.} Earlier in the post, you were ready to answer a question from your CMO about campaign reporting by channel. Now instead of crickets chirping, I hope you’re chomping at the bit to answer his question. Maybe something like this will do, “Tracking? Absolutely, you’ll receive reporting 48 hours into the campaign and then twice per week for the life of the campaign. Then we’ll create a presentation detailing our findings once the campaign ends.” He smiles, and then walks out with a confident look in his eye. You turn around and open Outlook, create a task, and enter “Send Glenn a quick thank you.” ;-)

I’m going to leave you with one last question. If you were the CMO and had 2 senior marketing managers provide you their campaign reporting and one provides you the level of detail that I listed above (from a relatively basic analytics setup), and the other provides you with almost no reporting, or sketchy reporting at best, which one would you allocate more budget to next year?


Labels: , ,