When analytics disappoints (and what to do about it) – Technical aspects

The diagnostics for analytics can take many forms, but the challenge lies in where to start troubleshooting first. One approach is to simplify the source of the problem – errors can lie either in the data collection or in the report settings of your analytics solution. Let’s look at a few typical occurrences.

Checking data collection and settings

Addressing data collection issues first is usually a good place to start. They are symptomatic of technical issues with an analytics code installation. These technical issues usually appear as sudden changes of data trends in analytics reports – Spikes or steep drops – or implemented changes in tag details, such as mistyped regular expressions.

First, eliminate the basic analytics installation concerns. Verify that the site pages contain the latest supported version of the web analytics solutions being used. If a site has not been updated in a while, it can be easy to have features or settings that are no longer useful in place. And analytics providers sunset features, so for the latest information, follow a well-supported resource from the solution provider.

Next, troubleshoot to confirm that the analytics page tag is installed correctly. Manually you can always look at the source code of a webpage just to to see if its there. No code means no data regarding the page is being reported in the analytics solution. Developers typically add the main analytics code at the top or bottom of a webpage.

A manual search of a page for an analytics is acceptable for simple sites with a page or two; For sites with a large number of pages, a few tools are available to make verification easier. One easy tool is the notification alert. Located in the upper right of the screen, the notification alert reflects what should be corrected in the settings (You can learn more about each indicator here in this Zimana post)

A world of debuggers and tag checkers

Given the steady onrush of JavaScript, coupled with design configurations that increasingly account for mobile, a bevy of debuggers are available to manager issues related to code. Among the most popular is WASP, a debugger created by Stephane Hamel of Cardinal Path, is an example, considwred the gold standard for checkers. It was launched on 2006 and has steadily been refined over the years; a WASP Chrome plugin is now available.

There is also Tag Assistant, a Google Chrome plugin which works with Google Tag manager and other enterprise-level tagging as well as Google Analytics.

You can also use a web development JavaScript checker like JSLint to check if there is an interference with another JavaScript code being called on an inspected page. This requires pasting the code into the checker.


Other tools that can help for issues regarding tags and cookies include Firebug (which works mainly with Firefox; Firebug Lite, a version for Google Chrome, is available) and Google Chrome DevTool. These can focus on potential issues that impact how the website code is being interpreted in the browser.

Tools are helpful for verifying how tags and fired when webpages are rendered. Packet sniffers come in handy for webpage render inspections, revealing, for example, where bots crawled without executing a Javascript call. Fiddler and Charles are the best known packet sniffers. In general if the data volume seems lowers than usual, inspect any design changes that impact data collection.

You can also record how data is being collected – Fiddler and Charles have recording features, which show a number of server calls for site elements when a webpage is rendered during a session.

Having a review of what tags have been added to a site can help highlight what tags should be in place at a website’s launch, and what has impacted the collected data from a site.

Confirm Report Settings

Some errors are associated to misinformation in the report settings. The result for analytic users is viewing trends that do not make intuitive sense. The errors are usually consistent – examining over different time periods or reviewing dimensions can reveal hints about what may be going on.

1. View if data is appearing in the report

Use real time features for quick confirm basic functionality for receiving data. Most analytics reports takes up to 24 hours to display results online, but many sub features can be available within a few hours depending on site visit volume.

2. Note misalignment of goals

Know how your site elements are related to conversion goals. It’s the core for what should be provided – apps. information, and the social media that compliments it.

3. Review environment in which reporting is issued.

This connects to number 2 – review what conditions are being created. Is there a period in which a spike occurred? What marketing occurred? Most of the details may not tell you why a spike happened, but you can better focus on how to deploy some possible next steps to correct the problem

4. Examine the syntax of filters to make sure they are functioning.

Filters sort our data that is reported through the use of regular expressions – text characters that filter characters in a sequence. These are tricky because it takes an imagination of the text pattern being filtered. Planning for regular expressions means accounting for ongoing anomalies. Thus user error or a sequence overlook can occur.

There are a few tools available for verifying the validity of regular expressions – RegExpal is an open source checker, though you may have to check if the platform is supported.

Also this is a good reminder to keep an untouched profile. For example, Google suggest maintaining an unfiltered analytics profile alongside one that contains filtered. Doing so permits quick identification of data anomalies and discrepancies.

5. Journal any changes made by admins

Mark each update or setting change with Google Annotation. Doing so can help everyone understand the history of what was added to a website or if a setting was changed. This can eliminate choices that have been tried before, and maintain focus on what can be corrected.

6. Note data changes greater than 10% as a starting point for diagnostics

This is more of a judgment call, but swings in data more than 10% can indicate a problem in data collection or a difference in a trend. This guide can help zero in on areas in which questions can be asked and action taken. Results in E-commerce and Adwords reports, for example, usually are based on traffic that clicked tagged URLs – like a paid search ad URL or a landing page for a commerce offer. Because these clicks rely on multiple tags or a complex site structure, like a subdirectory, click losses greater than 10% can be indicative of a tag firing concern.

7. Conduct periodic team review of report status and concerns

The best diagnostics step in maintaining the report settings is plan an occasional review with the team together and see how the reporting is used. This breaks silos – imaginary limitations of personal knowledge of what is occurring on a website. The end result is the right result, getting team members to communicate.

I know what analytics is – but how can I use it better for my business?

Pierre DeBois Chicago Small Business Expo
Pierre DeBois speaking about analytics at the 2014 Small Business Expo in Chicago

Most businesses have heard the definition of analytics at some point since 2006 (with Zimana being in business since 2009).  Here is the definition, for those just joining the internet.  Analytics is about the measurement of a system. With respect to small businesses, the interest is in web analytics. Web analytics is the measurement of website performance.

The origins of web analytics is linked with web development.  Originally it was meant to track and help account for webpages being served – in essence, a diagnostic tool.  But with the advent of paid search, combined with newly introduced Javascript tags, marketers realized that web analytics could provide a means to manage marketing campaigns, thus raising results and reducing costs.

The next advancement of importance  are smartphones and tablets. They changed the ways people review websites before making a purchase, because sites are accessed while visitors on the go instead of just at home.  Moreover, websites are increasingly accessed by mobile devices, so the sites have to be mobile-friendly.

If you are starting your business, there are a few ways to start using analytics to confirm how your website enhances your business digital strategy.  Planning for data collection can become complex based on the amount of marketing you have planned, but a dedicated planning pays off in organizing necessary steps, reducing effort needed over time.

If you are unsure where to start, here are a few great starting points if you are just planning to launching a website.

  • Identify how your web presence will be laid out.  Web designers use wireframes to lay our each page navigation, but if you are technically challenged you can do so on a piece of paper. This will also help assess costs for launching and maintaining associated components such as a microsite for an app or an e-commerce cart.
  • Choose your downloadable material for the site – these can be tagged in analytics through features such as event tracking, a minor code typically added to the HTML links for the documents. Tag managers are also available to make tag management manageable.
  • Install an analytics tag is on each page. Most analytic solutions use a Javascript code added into the HTML coding of the page (for Google Analytics, there are several checkers available).  Be prepared to modify the code for e-commerce, CRM, and other  features that are not included with standard reporting (a web developer or analytics specialist should help). Also consider a tag manager to manage a number of tags easily, particularly for complex set ups with cross-domains and e-commerce shopping carts.
  • Consider names for each webpages that includes a highly-trafficked keyword if feasible. You’ll have to keep page name lengths simple for user convenience (and it may depend on the URL you select). Three words in a subdirectory URL is a good rule of thumb to maintain.  So a URL can appear like this example below:


In this example, three words can be selected to the subdomain, kept relatively short in case a customer is typing the URL or speaking the words into a search window on a smartphone.  Each word can be an opportunity to use a keyword – say “shoe repair” in this instance – or a useful high volume description like “Chicago” to help in local search.


There’s no one combination that will guarantee success, but the key is develop the most descriptive page names that will aid discovery in a general search engine.

  • Make sure each page title and description are filled out, since those come in handy not only for search engine but in sharing content on Facebook, Google +, Pinterest, and other social media platforms.
  •  Identify a goal page from your website. This will aid reporting on how well people are navigating on your site.
  • Plan your marketing in addition to executing a search strategy.  Search traffic is helpful, but you may need to select marketing for an additional reach to your audience.

Growing A Pinterest Audience – Infographic via Who Is Hosting This

This infograph from the site Who Is Hosting This takes a comprehensive look at how a Pinterest audience is developed.  Doubling followers is no simple task, and no one task will trigger a flood of followers.  But organizing the activity in managing a Pinterest account can create additional followers.

Review the tips below and see how this infographic can refine your social media strategy for long-term success.

Double Your Pinterest Followers in Just Five Minutes a Day - Via Who Is Hosting This: The Blog

Source: WhoIsHostingThis.com

SEO How-To: Quickly Highlight Bad Links with Google Analytics

Link building can be an elaborate process, depending on the website, but sometimes you need a simple way to identify bad links. Google Analytics has a way to help identify

Go into the reporting of  Google Analytics account, and navigate to the Landing Pages report. The Landing Page report is a sub-bullet under the Content report under the Behavior segment (Behavior > Content > Landing Page).

Google Analytics Landing Page

Note: There is another landing page menu item under Acquisition > Search Engine Optimization. It does NOT contain the same second dimension menu as the one pictured, so do NOT use it for this link analysis.

Next, click on the drop down menu of the secondary source and add the following as a secondary dimension: Acquisition > Source.

Google Analytics Landing Page Report

You’ll see the source that generated the visit alongside the page on which the visitor landed to enter your site. You will see those sources that may lead to a quality link – you’ll see Small Business Trends as a favorite for Zimana, for example – and spammy links which should be questioned and removed. You can also see some 404s if that is the case.

Now examine links, using different time periods (30 days, 60 days, 90 days) to see what links appear in each period. The info indicates what people are consistently encountering.

To quickly highlight pages that appear in a long list (over 25 dimensions), use the find bar, a search feature in the Firefox and Chrome browsers, to discover the pages in the dimensions column. Click on the three bar icon in the upper right of the browser window. A Finder window will appear at the top (This may vary by browsers; Firefox displays the finder at the bottom) . Type in the word you want, and the Find Bar will reveal every appearance of that word throughout the page. You can even toggle to the specific instance in which the word appears.

After some evaluation, decide if you want to use disavowal links if you see more than a few spammy back links that appear in the reports. To implement a disavowal link, use the tools provided in either the Bing Webmaster or Google Webmaster tools.

You can also add this link report setting as a shortcut in the Dashboard. Adding the shortcut offers the convenience of not having to manually add the secondary dimension each time you check.