Having analytics data to rely upon when you have an ecommerce store is absolutely crucial. Accurate and timely data can uncover much more than just how many visitors come to your website or where they come from. You can in fact find out what marketing channels perform the best, where your visitors drop out or what content on your website is best performing. This Google Analytics audit checklist will help you to ensure your GA is setup properly.
Measuring conversions, time from first visit to conversion and segmenting visitors uncovers insights that can help you optimize your website. However, those insights are only as valuable as the data used to derive the insights is accurate and reliable. Ensuring analytical tool you use collects relevant, accurate and timely information is a critical precondition to actually analyzing traffic patterns on your website.
We’ve created a complimentary infographic which you can download here.
Before we embark further on the nuts and bolts of setting up analytics, I will first point out the need to define and identify measures that most accurately reflect the success and performance of the website itself. The process of measurement cannot be useful and accurate unless you know what is it that you want to measure.
No matter how accurate the measurement itself is, if you measure the wrong thing, you will not get the results that will be useful. By measuring a wrong performance indicator, not only will you get irrelevant data, but you will expand an effort to improve and optimize for things that do not make your website earn more money. This can be downright harmful to the ultimate goal of making visitors convert more.
To ensure you are measuring the right stuff, prior to setting up analytics itself is making what is usually called a measurement plan.
Measurement plan is a documentational foundation for setting up or auditing an analytics setup. The objective of a plan is to identify the primary measurements of performance and their interrelations. By making a plan you create a framework not only for an accurate measurement but to make a job of identifying the most important metrics.
Creating a measurement plan starts by defining the primary objective of a website. For example, if you operate an ecommerce website, your objective will be to sell as much of your products as possible. Thus, your priority metric will be a conversion rate on products purchase.
By observing how customers navigate the website, you can derive what individual events or content precedes conversion.
That way, you establish a series of metrics that each point towards higher level metric – for example a visitor may view a video of the product, then click the testimonials and read a few and then actually add the product in the cart. This sequence of events represents.a typical visitor journey and you can quantify the proportion of visitors that complete each of them.
So the first step in setting up analytics is to create a measurement plan.
To create a valid measurement plan start from the top.
Measurement plan setup checklist
|1.||Define the overall objective of a website|
|2.||Identify the measurement indicator directly contributing to this objective|
|3.||Establish the accurate mechanism to track the indicator|
|4.||Identify the supporting measurement indicators|
|5.||Identify the relationship and correlations and establish measurements|
|6.||Make the visualization of the plan to aid the process of implementing it|
When you create a measurement plan, you know what you want to measure and what metrics point to increased performance and success of your website. This is not enough, though. To be sure you are on the right track, you need to ensure analytical tool actually tracks what needs to be measured. The best measurement plan means nothing if your technical setup of analytics is not sound.
If a snippet is missing on any page of the website, it will not be possible to track that particular page with analytics. The other potential problem is if the snippet is included twice on a page. Both will result in distortion of the data by either tracking less visitors than actually were on the website or doubling their number.
By inspecting each page of the site, you should ensure the snippet implementation is done correctly. Once you ensure the reporting is correct – each visitor is reported only once – it means you can proceed further in the setup.
For more details related to account setup, read this article.
Account setup checklist
|1.||Check for presence of a snippet on every page|
|2.||Check if the snippet is at the proper position on the page|
|3.||Check for possible double snippets|
|4.||Check for proper account identification in a snippet|
|5.||Check if the customized snippet contains proper code|
Like many aspects of conversion optimization, there are certain best practices to follow in using Google Analytics. Unlike the other best practices, such as those for user interface design or doing experiments, best practices for Google Analytics are founded on exact principles.
One of the most important things that you should do is to create different views. Due to the nature of Google Analytics data collection mechanism, if you filter the data out, they will not be retrievable. Therefore, we (and everyone else) recommends having multiple views. Generally, most practitioners consider the three views to be minimum.
One view should be an untampered one, showing the raw data, without any filter whatsoever applied. Second view is to be used for testing newly introduced filters. Before you apply any new filter you create to the raw data, it should be tested for its effect in the test view.
Finally, the third view should contain the data with filters you verified work as intended. This is called the Master or Data view and it is the most important view, as it will invariable be where majority of analysis is conducted.
Second best practice is to establish and maintain a naming convention for any customizable element you use (UTM parameters, channel and content grouping, events, notifications etc. Establish a clear guidelines and ensure everyone with access to the account follows them.
Third recommended practice is to establish alerts. Google Analytics features the alerts you can tie to certain events on the site. For example, a spike in traffic or any other anomaly will be instantly reported via email so you will be able to check what is going on.
As the fourth best practice, check if the demographics, advanced attribution and user metrics features of Google Analytics are enabled. Although these should be enabled by default, verify that this is the case. Demographic report gives you data about the age, gender etc, using third party cookies, advanced attribution makes more data on links that bring visitors to the site available and user metrics allow you to analyze reports on per user basis.
And fifth important best practice is to link all the Google services accounts, such as AdSense, AdWords, etc. This is important to allow better analysis of marketing efforts, especially if you use AdWords for traffic acquisition. By using this feature, Analytics will be able to estimate ROI of the marketing.
Account management setup checklist
|1.||Create and maintain three views|
|2.||Establish naming convention and guidelines|
|3.||Create alerts for instant notifications|
|4.||Enable Demographics report, Advanced link attribution and User metric reporting|
|5.||Link other Google Services you may use|
Ensuring data integrity
When we set up analytics correctly, we still have not ensured the data we record is accurate. The next step we need to take is to make sure that our tracking reports only the real visitors. The problem is not negligible, as it can be seen from this screenshot (show a spam infested site).
What is spam? A spam traffic is a stream of false visitors – in fact a fake reports. To fight those, Google Analytics has a filter feature. Filter enables you to make it impossible for any other traffic than the real one to be reported. Configuring filters should be taken with care as it is possible to filter legitimate traffic too.
For this reason, it is necessary to create views in Google Analytics – a feature that enables you to track the same website in different ways. You want to do this due to the fact that it is impossible to access filtered out data once you introduce a filter. This means that if any filter happens to filter out legitimate data, this data will be lost forever.
Prior to introducing filter, try it out in another view to see its effects on legitimate traffic. Filters can be used for different purposes in addition to testing. For example you may filter out visitors from specific countries to observe the performance of localized website in the country it is localized for.
Checking out the filters and making sure they work properly is an important step in setting up the analytics. For more on filters, read our comprehensive article.
Data integrity setup checklist
|1.||Check traffic sources for anomalies|
|2.||Verify that there is no existing filters that block out traffic|
|3.||Introduce a filter to include only the traffic that comes from your hostname|
|4.||Leave a filter on in a test view, to prevent it affecting the data (master) view|
|5.||When you verify the filter is working and only blocks spam traffic, move it to master view|
|6.||Remain vigilant and check for spam in regular intervals|
Tracking subdomain traffic
Many websites maintain different subdomains for example blogs, forums, technical support. Tracking the traffic to subdomains is necessary to make sure you can see what your visitors do across the subdomains. Making sure you track subdomains traffic along with your main domain is critical. Sometimes your visitors will navigate across the subdomains and convert. Therefore, it is necessary to make sure subdomain traffic is tracked and attributed to the same user as he moves through different subdomains of your website.
For example, your website may include a blog subdomain. Your visitors may approach it multiple times and read articles. Finally, they may convert following a link from your subdomain. To establish the sequence of events and optimize a visitor journey, you need to setup domain tracking.
Tracking subdomain traffic setup checklist
|1.||Identify subdomains you want to track|
|2.||Check if the cookie domain is set to auto|
|3.||Remove subdomains from referral exclusion filters if they are there|
Internal traffic exclusion
While this seems to be obvious, it is an issue that is frequently neglected. Most websites are frequently updated, edited or otherwise approached by personnel of the company that owns it or the agency that maintains the site. Filtering out this activity is necessary as it may create a false impression of traffic that is not really customers.
Every audit should make sure that analytics does not show your own traffic or traffic from the agency that maintains the website or does your CRO. If this sort of traffic is not excluded, you will end up with inflated number of visitors (especially if the website is frequently changed or updated) and skew your visitor behavior data. To avoid this make sure you exclude internal traffic, using options provided in Admin.
To achieve this use the filters (same feature used to filter out spam traffic) and filter out IP addresses that your employees or agency use to access the website. You can filter out the entire subdomain or just individual address.
Internal traffic exclusion setup checklist
|1.||Identify sources of internal traffic – IP addresses or domains|
|2.||Create a new filter and exclude the traffic from these sources|
|3.||Verify the filter works and put it in your master view|
To make your data better and more accessible, it is recommended to group the pages in order to make the reports more readable. The problem of many ecommerce websites, is that the pages are generated dynamically for different product variation – colors, size etc. The result is that you may have reports containing hundreds of different pages and it would be very hard to differentiate between different products.
The way to overcome this problem is to group similar pages in one and so you can see the amount of people purchasing or being interested in different versions of products.
Content grouping setup checklist
|1.||Examine your content reports and see if dynamically created pages create multiple reports of the same page|
|2.||Consolidate each under the full name of the page|
|3.||Verify the reports are working and no default page name is in the all pages report|
In order to have a clear view of how your customers navigate and purchase the products on your website, you need to set up a conversion funnel. Properly setup and configured conversion funnel enables you to visualize clearly how a visitor becomes a customer.
Funnel begins with a customer placing a product into a cart and then proceeding to confirm a purchase. Each of these steps presents a customer with an obstacle. Observing how they negotiate those steps reveals what parts of funnel may be flowed.
Designing and structuring a funnel journey is an important part of properly functioning analytics. Google Analytics allows visualization of a distance goal so configuring a goal results in having a funnel visualization.
Funnel setup checklist
|1.||Check if the goals exist|
|2.||Review the conversion process by user testing it directly|
|3.||List all the positive outcomes for the website|
|4.||Create a destination goals with intermediary steps that enable visualization of customer journey|
|5.||Verify the funnel works and represents reality|
To track behavior and engagement of visitors on a website, it is not enough to rely on default features of the Analytics. To have a better view of a visitor behavior, it is necessary to setup tracking of actions of visitors as they navigate the website. Some actions are tracked by default, such as link clicks and similar. However, if you have a video on your product page, or a file that can be downloaded those actions will not be tracked by default.
To track them you need to use events, or a custom code that sends the event tracking reports to Google Analytics. To be useful, events need to be tied in with a measurement plan. For example, if a measurement plan identifies a blog posts as the content helping convert visitors, then there should be an event that triggers when a visitor reads a post.
To ensure that event report accurately you should use a Google Analytics diagnostics. This useful tool enables you to immediately check the results of event reports. Using this handy tool and Analytics real time reports you can ensure that your event is reported immediately and correct the code or other issues if it is not.
Event setup checklist
|1.||Using Measurement plan as a reference make a list of events that contribute to conversions|
|2.||Go through the web pages and add the necessary code to track those events (or use Tag Manager to do it)|
|3.||Assign a relative value for each event, reflecting its significance in conversion path or funnel to make comparison easy|
|4.||Create event goals for easier tracking for more important events (preferably the ones preceding the actual macro conversions)|
|5.||Verify the events are reported properly|
To make it easier to track acquisition channels, you need to group them. Similar to pages, different acquisition channels may receive different designations in your reports. For example, Twitter may appear in forms of:
Both represent the same source, so the traffic acquired from it should be consolidated under one label. Channel grouping is important and should be carefully configured. Each source/medium should be consolidated so that it shows most data in least possible space.
Enhanced ecommerce features
To facilitate better reporting of ecommerce transactions, Google Analytics contains Enhanced eCommerce reporting feature. To be useful, this feature needs to be activated to function.
While Enhanced ecommerce functionality can be enabled in Analytics admin interface, to work, you need to make a few adjustments to the snippet code and on the page. There is a number of good guides for this, such as the one on Google support pages. Besides, a number of ecommerce platforms, such as WordPress or Shopify, have plugins that enable ecommerce reporting.
Sometimes, to capture data Google Analytics does not show by default, you will use custom dimensions or metrics. These are pieces of code you add to a snippet and that Google Analytics reports using custom form. An example of this is a custom dimension you set to track active time visitor spends on a website, using custom event script, implemented using Google Tag Manager.
In order to ensure these are implemented correctly and do their job you need to check it out. If you manage several Analytics account and implemented same custom dimensions in all of them, you can use an automated method to audit them. This tool is an extension on Google Sheet that fetches the data from all accounts you have access to and shows you if the custom metrics and dimensions actually track and report the data you want them to track.
You can download this tool at GitHub and it was developed by Simo Ahava, one of the best known Google Analytics and Tag Manager gurus.
After checking for the major issues, there are a few other issues you should check. While admittedly minor, they still may skew your analytical data and result in missing an important insight.
- Check the time zone Google Analytics is using. Default time zone for Google Analytics is PST – Pacific standard time. Unless your company operates in Western US coast, you’d want to adjust this to reflect your home time zone. This is important to get right in order to make sense of when your audience visits your website – is it during the work hours, after work, etc.
- Another issue that you may encounter is the lack of UTM parameters in your campaigns. While Google AdWords, AdSense and other marketing products will provide the identification of campaigns automatically, using other providers, such as Twitter, Facebook, email or other similar channels and mediums will frequently result in not knowing which campaign brought visitors to the site. The solution is to create specific URLs for each campaign and link to the products or content you promote. Google offers a UTM builder greatly easing this task.
- Check for account structure and only allow admin privileges to people you are sure will know what they are doing. Use the audit process to check out for this and make sure that all who need to have privileged access have it and revoke it for the staff members who will not need to use it in their daily activities. Beside, keep the admin privilege limited to a small group (ideally a couple of people) so that not everyone can introduce changes to the settings.
- Make sure every change to account, property or view settings is followed with a proper notification. This makes account management easier.
- Check out the dashboard and make sure you create one providing an instant access to the most important metrics, in accordance with a measurement plan.
- Check that Exclude traffic from known bots and spammers check box is checked in View settings. This automatically applies a filter for all known sources of spam traffic.
For additional tips&tricks, read our extensive guide.
Audit must be done with care and deliberately. Omitting any step can prove to be a critical error that has potential to slow down or even invalidate any quantitative research you make. Since there is no way to see the data retroactively or eliminate filters (for example) and see the data filtered out, you need to take great care in audit process to do it right and ensure the full integrity of your Analytics account.
Automated tools, such as Supermetrics Audit template are useful, but can never be a replacement for a deliberate and thorough hands on audit. First off, they do not solve the issues and secondly, without enough data to begin with, you will not be able to deduce if something is wrong. For established and running analytics account though, these tools can be irreplaceable.
Conducting an audit is a critical activity as the data provided by Analytics can be of critical importance for many aspects of an ecommerce website. Not only does it provide a clue on how many visitors there are and how popular website is, but it enables you to deduce which content is the most popular. The item of primary importance is ensuring that attribution analysis works accurately.
This feature can enable you to save a lot of money in advertising by identifying most efficient channels and concentrating on them or fixing what is wrong with less performing ones. To properly set up it you need to have goals, acquisition channels and funnel properly setup and only audit process can ensure all of these features are properly implemented.
We have created an infographic as part of the Google Analytics audit series. Here is the direct link to the infographic if you want to post on your website – http://www.objeqt.com/wp-content/uploads/google-analytics-audit-infographic.png