Don't waste another minute on web analytics tag QA

Checking to be sure web analytics are implemented correctly on a website can be a burdensome task. Recognizing this, many tools have been developed to help speed things along. Most of these tools are focused on checking individual tags on individual pages and they're great for spot-checks. The question these tools leave unanswered, however, is whether the website's analytics as a whole is correctly implemented. The most you can do with many tools is look for and work on problems you already know about. But what about problems you don't know about? Or even worse, problems that you don't even know to look for - unknown unknowns?

Let's consider this concept of "unknown unknowns" for a moment. When we go to solve any problem, whether it's a lack of hot water, a war in a foreign country, or missing data in a reporting suite, there are essentially three classifications of data we use to solve problems.

Known knowns: things we know, i.e., I turned the hot water faucet; terrorism exists; my entry pages report doesn't look right.

Known unknowns: things we know we need to find out about, i.e., I don't know whether my water heater pilot light is lit; we don't know where the terrorists are hiding; I think my web analytics implementation is broken.

Unknown Unknowns: things we don't know, and don't know to look for, i.e., a construction crew ruptured the natural gas pipeline 3 blocks away; the terrorists are not in any one place; JS code added by a remote dev team last week is bad, and in turn is breaking the analytics code on 10 of your most popular landing pages.

So the question, when it comes to an analytics implementation is: How can you use a QA tool to find and fix every unknown unknown?

Well, with most tools, the best you can do is address problems you know you have, and do some QA to try and discover more problems. But of course you can't really afford to do too much "discovery," because as a web analyst, you've got a lot of other things to focus on.

So how can you, the busy web analyst, ensure that your web analytics data is consistent, reliable, and accurate without devoting huge amounts of time and budget to a QA team?

Automating Web Analytics QA with ObservePoint

ObservePoint's SiteAudit uses a web crawler to discover all of the pages on your website. It uses "engines" to load the page and all of its parts. While this is happening, it finds all analytics tags on the page, executes the JS code and can either complete or cancel the call to the vendor. So for every page on your site, SiteAudit will tell you where code is and isn't. Auditing your site requires just a few simple steps:

Setting up a New Domain

It only takes a few minutes to set up SiteAudit to find and alert you of all of your site's implementation issues.

First, enter the URL you'd like SiteAudit to scan from. This page will be used as a jumping off point, and SiteAudit will follow any links on this and resultant pages.

Next, enter a description for this audit. This helps you know what exactly you're looking at once the audit is finished. Also, there are some controls that allow you to limit the width and depth of your audit.

Finally, there are a few other details you should consider; for instance, how often to audit (more information on this in our getting started guide), to whom to send alerts, whether to share the audit data with other users in the account, filters to avoid off-site links, using other user agents, and which code is most important to you - more info on these options here.

Now that you've made your selections, click SAVE, and the audit will start. Go ahead and take a coffee break while you wait.

After some time, depending on the size of your audit and speed of your site, you will receive an alert to notify you that the audit has finished. When this happens, simply log back into your account.

On the home screen, you'll see a list of all of your audits. Click on the audit you want to look at.

Toward the bottom of the page, SiteAudit lists every tag with a unique account ID that was found during the audit. Clicking on the tag you want expands a table of variables.

What we're interested in first is which pages were found without tags? To find out, just click the number in the "missing / error" column in the first row of the table and voila - here's your list of pages missing tags. Click export and you've got what you need!

You just saved hours of sifting through your website, page-by-page, trying to ferret out tagging errors.

Are you spending too much time trying to find tagging problems on your site? Automate your web anlaytics QA with SiteAudit! You can even try it for free.