Monday, November 29, 2010

Google Analytics Data Skewed Because Of Instant Previews !!


There are confirmed reports that Google Instant Previews may be skewing your web analytics data.

It appears that in some cases Google will conduct an on-demand fetch of your page to dynamically create an Instant Preview. The on-demand fetch happens when a searcher places his mouse over the search result on Google and the image preview comes up. Some analytics tools, including Google Analytics, may consider that a visit, because Google Instant Preview is actually visiting the page in real-time to get that on-demand Instant Preview.

There are several complaints about this issue in the Google Help forums. Google’s John Mueller replied saying they are working on fixing this from showing up in Google Analytics. “We’re working on a solution for this, to prevent Google Instant Preview on-demand fetches from executing Analytics JavaScript,” John said.

Ideally, There is no estimated time for when this may be resolved. So if you have seen a skew in your analytics data in the past two weeks, this may be the reason.

Read More: Google Analytics Data Skewed Because Of Instant Previews

Friday, November 19, 2010

Troubleshooting Google Analytics Goals and Funnels !!


Objective:-
In this module you'll learn some of the most common reasons for why goals and funnels aren't functioning properly, and how to fix them.
Troubleshooting Goals That Aren't Being Tracked

One way to check if you have written your Goal URL correctly is to see if the page is being tracked. Search the Top Content report for the goal page to confirm the page is properly tracked and counted as a goal. Please see the examples below for further information.

I. Exact Match/Head Match:

Search the Top Content report (found underneath the 'Content' section) for the request URI of the goal URL.

For example, if your goal URL is www.example.com/cats/prettycatcheckout.html, then the Request URI is everything after the domain name '/cats/prettycatcheckout.html.'

If the request URI appears in the Top Content report, then the goal URL is written correctly. However, if it doesn't appear, please see troubleshooting tips here.

II. Regular Expression Match

Search your Top Content report using your regular expression. The report filter allows regular expressions, so your goal page should appear if your regular expression is written correctly. If the goal page doesn't appear, please see the troubleshooting tips here.

Tip: If the goal page doesn't appear in your Top Content report the first time you search, try modifying your search until it does appear. Then use the search result from your modified search query as your Goal URL.
Troubleshooting Funnel Drop-Offs

If you have funnel steps with URLs to different domains or subdomains, and the tracking code isn't customized as described in the following articles, then all visitors that go from the website in the first step to any other domain or subdomain will appear to drop off in your reports. Learn how to track multiple domains or subdomains in Section 1: Installing Google Analytics on Complex Websites.

For example, suppose a funnel is setup as follows:

Goal URL: www.secondsite.com/jkl.html

Funnel Steps:
Step 1 (You're checked off 'Required step' for Step 1 from your Goal settings page): www.firstsite.com/abc.html
Step 2: www.firstsite.com/def.html
Step 3: www.secondsite.com/ghi.html

In this example, if the tracking code is not customized as needed for multiple domains, then users will appear to drop off at Step 2 because a new session will be created when the user goes from www.firstsite.com to www.secondsite.com. In the new session, Google Analytics will not know that the user actually visited Step 1 because it was done in the previous session. Since the first step is required in this funnel, it will also appear as if the user did not convert towards the goal, and Step 3 and the Goal URL will not be recorded in the funnel since it didn't occur in the same session.

Read more: Overview of Troubleshooting Google Analytics Goals and Funnels


Monday, November 15, 2010

In web analytics, everything is relative !!


What's a good bounce rate for my web site?

I get that kind of question a lot. What's a 'good' bounce rate? A 'good' time on site?

The answer, I'm afraid, is: Better than your current bounce rate. Better than your current time on site.

In web analytics, it's best to focus on your own data and on improving. Use yourself as the benchmark. This is your best strategy for two reasons:
Lack of accurate benchmarks

Accurate, internet- or industry-wide data on keyword searches, or competitors, or just about anything else, is scarce. Non-existent, really.

1. 'Panel'-based statistics like Compete.com (which I love) and Alexa (which I'm starting to like again) sweep in an incredibly wide range of web sites. The bounce rate on your online bike shop won't compare to, say, the bounce rate on the New York Times web site.
2. Statistics within your own industry will include outliers at both end of the spectrum: At one end are the companies that have invested 100x your budget to become the shining pinnacle of conversion rate optimization. At the other, you'll be comparing yourself to the sites designed according to 1992 best practices. Even if you can narrow down the data in #1, it'll be inaccurate..
3. Keyword data from Google is about as trustworthy as a credit default swap.
4. Keyword data from other sources may be more trustworthy, but shows you a tiny sliver of total search traffic.

Numbers lie

Even if you could get accurate benchmarks, they still lie. Your business isn't like your competitors', no matter how similar they seem. Competitor A just fired his head of sales, so conversion rates tanked for a month. Competitor B happened to get on Channel 5 News. Her traffic tripled, lowering her conversion rate, too - but her sales skyrocketed.

Unless you've got the whole story, the numbers will lie. And you can't get the whole story.
Focus on improvement

So, if you're trying to figure out how many visitors you should be getting for 'slobber knocker', the answer is? More than you get right now.

If you're trying to figure out where your conversion rate should be? Yep. Better than what you're getting right now.

That's what web analytics are for: Helping you improve. Which, as it happens, is also how you beat your competitors.


Read more: In web analytics, everything is relative

Saturday, November 6, 2010

Web Analytics Can also Track Offline Campaigns !!

Web analytics is not just a tool for measuring Web site traffic. Web analytics applications can also help companies measure the results of traditional print advertising campaigns.

That's what New York-based retailer BuiltNY discovered when it began running a four-issue print campaign in Dwell magazine in August. The company's "design-focused" neoprene tote bags for things like wine, lunch, or laptops are sold directly on its Web site and through resellers in 30 countries.

The campaign began in Dwell's September issue with a quirky expose of what various items look like in a BuiltNY bag through an airport x-ray machine. The ad itself featured a candid letter from BuiltNY, superimposed on one of the images, describing the events leading up to the creation of the ad.

To track the print campaign, BuiltNY put a unique, easy-to-remember unique URL in the ad, which was only in use for that campaign. That landing page shows colorful x-ray images of objects like wine bottles, lunches, seashells and beach gear, all inside the appropriate BuiltNY bag.

Through Google Analytics, BuiltNY was able to attribute an 800 percent boost in traffic when the ad hit newsstands, and a 40 percent increase in online sales from visitors that came through that URL, Steve Bowden, art director for BuiltNY, told ClickZ.

"We can read it like the Wall Street Journal for our own Web traffic," Bowden said. "Every morning we get an update on how our Web, print and e-mail campaigns are doing, correlated to sales."

"Instead of gathering around the table scratching our heads, we actually have data to show how the campaign is performing," added Aaron Lown, a principal at BuiltNY and its co-creative director.

The second ad in the series, in Dwell's October issue, follows the bare-bones letter approach, with an invitation to browse BuiltNY's new line of cases for cell phones, MP3 players, laptops and other electronics. That ad links to an online game BuiltNY developed with Justin Bakse of VolcanoKit.com, where users battle alien neoprene electronic accessories with a "ballistic champagne bottle."

BuiltNY ran a more traditional test ad in Dwell earlier this year, before it began using Google Analytics. "But I have no idea if it worked," Lown said, since he had no way to track its success. Prior to implementing Google Analytics about the same time the X-ray campaign began, BuiltNY hadn't used any Web analytics products for its first three years in business. "We just had too many other things to do, like design new products, run our business...," Lown said.

These print ads account for a majority of BuiltNY's marketing spend, which is balanced out primarily with public relations outreach, quarterly e-mail campaigns, and a small AdWords campaign, Lown said. "This is the first advertising we've done, and while it's small in the grand scheme of things, it's our largest ad effort," he said.

BuiltNY is now using Google Analytics to track all of its online and offline efforts. It's also using the data to optimize future campaigns, such as determining the best day and time to send out e-mail communications to existing customers. Because it can now accurately determine the value of its online and offline campaigns, BuiltNY can more confidently spend its limited marketing budget knowing what the return will be, Lown said.

Web analytics applications are often underutilized by small and mid-sized businesses like BuiltNY, which currently has about 30 employees. For many small businesses, the only Web analytics available are simple traffic counting applications available through their hosting provider, which provide little actionable value, Greg Dowling, senior analyst at JupiterResearch, told ClickZ. Those companies eschew more sophisticated analytics applications for reasons of cost, information overload, or confusion over how to utilize the data in their business, he said.

"I think the primary barrier is the lack of internal resources required to effectively establish, monitor, and maintain a Web analytics installation, as well as the cost-prohibitive nature of enterprise class Web analytics platforms," Dowling said. "Additionally, tool complexity prevents users who actually install these applications from getting any real value out of them if they don't have dedicated Web analysts supporting these installations."

Google tackled the first issue head-on by offering its analytics product for free, and is taking on the others with a robust help center and knowledge base of practical applications it calls Conversion University. The money saved by Google Analytics being free can also be put into hiring and training Web analysts or paying for consulting from existing partners, Dowling said.

While the tracking method BuiltNY used for its print ads is straightforward, there are hidden pitfalls that have prevented more implementations, Dowling said.

"While it is technically easy to track offline campaigns through the use of redirect or vanity URLs, the practice is often wrought with complexities and is prone to error, making the data collected highly suspect," he said. "Tighter integration with direct marketing systems that would allow for the tracking of campaign respondents across campaigns both on and offline is becoming available as Web analytics vendors enhance data integration capabilities, but widespread adoption and usage is limited."