Web analysts and business decision-makers alike often don’t have confidence in what their web analytics solution tells them.
Every business using Adobe Analytics wants its implementation to capture the most valuable pieces of data possible to meet business goals and objectives. Unfortunately, even the most flawless implementations begin to devolve and decay over time—creating the need to conduct an audit.
This eBook provides a 21 point Adobe Analytics audit to ensure that you are capturing data you can trust and that you can confidently use to drive the right decisions in your organization.
The auditing strategy in this eBook includes:
- Best practices seen across the most successful companies
- Techniques to use regardless of site maturity or industry vertical
- Expert insight of how to create tangible business value from your analytics
Fill out the form to download your free copy.
Regardless of the size of your company, a periodic audit of your Adobe Analytics implementation is essential for verifying the accuracy of your data.
Don’t be scared off by the word audit. Though the term can suggest an adversarial relationship of sorts, an Adobe Analytics audit is something else entirely.
An Adobe Analytics audit is a proven method for companies to demonstrate they are generating substantial value from their investment in Adobe Analytics by verifying the technology is being used correctly and the analytics data is accurate.
The 21 points explored in this eBook are all designed to help you answer the most important question about your implementation: Can you trust your analytics data?
As you read through, use the following points to ensure you’re capturing the right data at the right times and using that data to drive the right decisions in your organization.
Conducting Your Audit
The points of this audit can be broken up into 4 categories, with some overlap:
- On-Page Audit
- Internal Data Quality Audit
- Conversion Measurement Audit
- Traffic Sources and Site Activity Review
Let’s get started.
On-Page Audit
An on-page audit involves going to your website and evaluating whether or not Adobe Analytics is correctly deployed and that no user experience issues inhibit your ability to collect meaningful data. The following 8 points address on-page audits.
1. Tag Presence
If tags are not present on an individual page, no data from that page can be collected.
Certainly any page interaction you want to track (for most companies that is every page) must have tags that are correctly installed and firing.
Ideally, 100% of a site’s pages will have Adobe Analytics tags and consistent tag versions deployed. You can verify tag presence manually page-by-page, or use ObservePoint’s Web Audits feature to scan a large batch of pages.
Feature Highlight: Web Audits
Companies like Adobe, Comcast, Johnson & Johnson, and Hewlett-Packard Enterprise use ObservePoint’s Web Audits to regularly scan their site and discover what technologies are gathering data. Using a web audit, you can easily verify that Adobe Analytics is installed on each page of your site.
2. Duplicate Tags
Just like missing tags, duplicate tags can muddle your data quality because they risk inflating your metrics.
In the worst-case scenario, a single page view may be counted multiple times, causing the page to appear much more popular than it is. Some other issues include:
- Conversion events being counted multiple times
- An old version of Adobe Analytics firing following a migration of tag management systems
And consider that inflated page views can cause conversion rates to look lower than they are in reality.
Beyond just messing with data quality, duplicate tagging at least double the analytics cost for every page where it occurs.
Consequently, you should ensure that you do not have duplicate analytics code deployed on the same page, and that if you do, all tags appear to be loading multiple requests in the manner you expect.
ObservePoint’s Duplicate Tags report can help you zero in on duplicate Adobe Analytics tags.
3. Vendor Compliance
Adobe dictates that tags must be configured within a certain syntax and that certain disclosures are made on the website, like a valid privacy policy. When these standards are not kept, Adobe Analytics may truncate or discard data.
For example, page names and many traffic variables must be less than 100 bytes. For English character sets, this equates to 100 characters. In a double-byte character set, like Japanese, it equates to only 50 characters. If your audit shows a large number of pages or tags do not conform to the vendor’s rules, your analytics data will be unreliable.
Using ObservePoint’s Custom Tag feature, you can automate vendor compliance by running custom JavaScript on each page of your site during a Web Audit to verify your privacy policy is available and accessible.
4. Page and Tag Load Times
Auditing your page and tag load times is important because load times impact the performance of your site in several ways, causing issues like:
Limited Analytics Ability. If a page is slow to load and users navigate away from a page before the page has finished loading, Adobe Analytics may not execute, reducing the real visitor counts on those pages.
Inaccurate Data Comparison. Further inaccuracies with data collection can occur when different sections of your site load tags in different locations (such as at the top or bottom of the page), in which case a page loading dramatically faster or slower than another may cause unequal measurement in your reports.
User Experience Issues. Page load time is obviously an important part of user experience, and a slow-load time is a major contributing factor to page abandonment. The web tags on your pages impact how quickly, or slowly, they load.
Your audit needs to check against the possibility of data loss through long load times or inconsistent tag placement.
ObservePoint’s Page Load Time and Tag Load Time reports can give you a quick-and-easy view into which pages/tags are taking a long time to load.
5. Internal Linking
Redirects and broken links frustrate users and affect search engine rankings. Though not directly tied to your analytics solution, broken internal linking on your site can potentially drive away users and limit your pool of data.
Best practices dictate that no website should have page-not-found (400-series) errors and use redirects (300-series) only minimally. Ideally, all links would land on a regular page.
If your audit reveals a significant number of redirects, you should review your organization’s redirect policy, and remove as many internal redirect links as possible. You may need to update the internal site linking to point directly to the terminal page and eliminate the need to pass through a redirect.
Another issue with redirects is losing the referral page. If a customer comes to your site through a banner ad and encounters one or more redirects, the final destination page will lose the original referrer and any conversions will not be attributed correctly. A strong tagging and data governance plan will ensure that redirects carry and reset the referrer variable on the final destination page.
ObservePoint’s Page Status Codes report can reveal which pages on your site have unnecessary redirects or 400-series errors.
6. JavaScript Errors
If you have JavaScript errors within any of your on-page code, then Adobe Analytics may not fire, which means no data for you. Best practice is to keep errors at a minimum. ObservePoint’s Console Logs report reveals any and all errors, warnings, and other logs that showed up on each page during an audit.
7. Tag Validation
Tag validation refers to the process of verifying your analytics implementation (including tags, eVars, props, the data layer, and other components) is collecting and transmitting the correct data to your Adobe Analytics solution.
An individual who is knowledgeable about the implementation design of your Adobe Analytics architecture should review pages, watching for data collection issues such as self-referral problems.
Many Adobe Analytics practitioners have some form of documentation (such as an SDR) specifying how variables should be configured on each page. Tag validation verifies that your implementation matches the expectations outlined in your documentation.
Examples of issues that you would pick up through tag validation include:
- Link tracking that fails to set an event declaration.
- Data being sent into the wrong report suite, such as a development report suite.
Validating your implementation can be very tedious and prone to human error, in which case automation will be your best friend.
Feature Highlight: Rules
Using Rules, ObservePoint users can define expected values for data collection on any given page or section of a site. If during a scan a rule doesn’t pass, the user immediately receives a notification.
For example, if you expect eVar 2 to always have a 5-digit number any time an Adobe Analytics pageview fires, you can set up a Rule that verifies just that.
8. Possible Data Leakage
As you audit for Adobe Analytics data collection, you will likely uncover additional, unauthorized tracking technologies on your site. If an unauthorized technology is sending sensitive data to third parties, that’s a red flag you need to address immediately.
Additionally, you may find there is proprietary data in clear text being passed to your analytics variables. Sometimes cost of goods sold or other sensitive data might be set in a variable.
Because any competitor can access the data by simply looking at the page source, you want to avoid having that data sent in clear text. One way to protect your sensitive data is to encode it on the page and decode it on the server side using Processing Rules or a VISTA rule.
Another point of concern is whether you are inadvertently passing personally identifiable information (PII) to Adobe Analytics or another technology. Verify that no such transmission is happening.
Using ObservePoint’s Rules with RegEx pattern matching, you can check for transmission of PII (such as emails and identification numbers), unauthorized technologies, or proprietary data, and receive alerts if any unauthorized data collection occurs.
Internal Data Quality Audit
This phase of your audit is a check up on the internal configurations within Adobe Analytics that can affect the quality of your data. The following 4 points are specific settings you should evaluate in Adobe Analytics.
9. IP Address Exclusion
Certain visitors who don’t accurately reflect your target audience, like company employees, may frequent your site and artificially inflate visit and page view counts. Adobe Analytics allows you to filter out these users by IP address (range).
During your audit, you should review the configured filters to ensure you are excluding appropriate traffic.
10. External Data Sources Enabled
When imported into Adobe Analytics, data from Google AdWords indicates how much of your website traffic or sales are driven by AdWords. Ensure your accounts are correct and have an active administrator.
External data imported into Adobe Analytics through Data Sources or Data Connectors require a record key if you need to tie the external data with the analytics data.
External data could come from an AdWords account, Cheetahmail integration, or a call center log. Often the transactionID variable is the common key.
There is no report in Adobe Analytics that shows data captured in transactionID; it can only be accessed in Data Warehouse. A transactionID might be used to link online sales with offline returns, for example, or online lead forms with follow up calls.
Use your SDR to identify when transactionIDs should be captured and where the data comes from, then build a business rule in your audits to verify that it is populating.
Information informs insights, which inform decisions, so do your research about what other data you can pull into Adobe Analytics.
11. Page Naming
Some analytics vendors recommend using the Page Title tag on each page to populate the PageName variable. However, this poses a risk to clean data because a Page Title tag is often bloated with additional words for search engine optimization or it might change.
A page name that changes is an enemy to analysts because the Adobe Analytics reports cannot combine old and new names into the same report. Every page needs an unchanging, unique name under 100 bytes in length. Various strategies can be used for a good page naming scheme, but it is wise to consult with the stakeholders before setting it in place.
12. Internal Search
Taking time to verify your internal search configuration during your Adobe Analytics audit is vital to understanding the findability of the content on your website.
Setting up internal search in Adobe Analytics allows you to evaluate where visitors search, what they search for, and the quality of the provided search results.
Internal search should record the actual search terms, the entire phrase, and number of search results returned. Most companies use both props and eVars for search terms so they can do pathing (s.props) and success attribution (eVars). Many websites make the mistake of recording the search term on every page of the search results, which inflates the number of times search terms are counted.
Understanding the effectiveness of your internal search is important because visitors who perform internal searches convert at a substantially higher rate.
Conversion Tracking Audit
Mature Adobe Analytics implementations measure conversion events to see if website visitors are successfully completing desired actions. As part of your audit, you will want to verify that conversion tracking is firing as expected.
13. Conversion Events
Conversion events are the activities you want your visitors to perform on the website (make a purchase, submit a lead form, apply for a loan, and so forth). Events, also called “success events,” are metrics which are used to evaluate how successful the website experience is for users.
Begin your audit by understanding and labeling the conversion events your website has. Your conversion events will likely have custom dimensions and metrics associated with them, so you will want to make sure that when Adobe Analytics logs a conversion event, the tracking includes those custom definitions.
A Word About Event Serialization
If your users ever bookmark conversion pages you run the risk of overcounting the conversion. For example, many customers may save a “Thank You” page because it confirms their order or contains information they want to refer back to. By default, the conversion is counted again when they later reload the page. Good analytics governance will address when an event should be serialized, or counted once.
Events might be counted once in a lifetime (purchase), once in a month (internal promotion), or once per session (contest registration), depending on the goals of the activity.
Purchases must also be serialized if there is a possibility of reloading the page with the Adobe Analytics tags intact. Purchase serialization uses s.purchaseID each time the purchase event is set.
Ultimately, your audit will require you to simulate your conversion events and verify the corresponding request to the Adobe Analytics server contains the appropriate data. You can approach this process manually or via automation with ObservePoint’s Web Audits and Web Journeys features.
During a scan of your site, ObservePoint can simulate user actions like form fills, selects, clicks, and submits, while also validating the analytics requests being sent off as a result of these actions.
14. eCommerce Tracking
If your website sells a product, having your eCommerce tracking enabled is critical.
By this measurement, your organization can evaluate the effectiveness of marketing investments and site optimizations. eCommerce events, such as Product View, Checkout, and Purchase must be set with the s.products variable. Price and quantity are only set with the Purchase event to prevent them from being inflated.
You will evaluate your eCommerce tracking the same way you do other conversion events, with a focus on the unique data points sent in an eCommerce event request.
You can also gauge the accuracy of your eCommerce data by comparing it against your company’s ordering system data.
Using Web Audits, Web Journeys, and Rules in tandem, ObservePoint users can verify that Adobe Analytics is capturing and transmitting eCommerce variables correctly.
15. Campaign Tracking
Campaign attribution is one of the primary goals of web analytics, particularly for eCommerce companies. Auditing your campaign tracking is perhaps the single most important part of a comprehensive audit.
Campaign tracking errors can occur as a result of one or both of the following misconfigurations:
- Campaign tracking codes (URLs) are formatted incorrectly
- Adobe Analytics tags on the campaign landing page don’t correctly capture the campaign parameters, such as a campaign ID
During your Adobe Analytics audit you should review both of the above campaign tracking components to ensure that your marketing efforts are receiving proper attribution.
ObservePoint’s CampaignAssurance can help protect against both of the above issues, allowing you to confidently rely on your campaign attribution data.
Traffic Sources and Site Activity Review
This portion of your audit is an evaluation of some of the most important traffic sources, paths, and pages on your site. As you build out a repeatable auditing process (ideally through automation), the below assets will be the most important for you to audit to verify accurate data collection.
16. Best Performing Traffic Sources
One of the key benefits of using Adobe Analytics in online marketing is being able to compare the effectiveness of traffic sources. You can evaluate how much of your traffic comes from Google searches, direct site access, AdWords, Bing/organic and email campaigns, among others.
As a part of your audit, ensure that the Marketing Channels Rules and any Processing Rules are still working as designed and that no new scenarios have crept in over time which the rules may not account for.
17. Best Performing Keywords
Understanding how organic search visitors are finding your website is an important part of evaluating marketing efforts.
Consider adding a profile filter to replace Google keyword reports that read as “Not Provided” with a landing page URL or some other more meaningful information. “Not Provided” is typically simply a result of visitors who were already logged into their Google account when performing their search.
18. Critical Visitor and Conversion Paths
Understanding how visitors use your website is an important element in optimizing the site. During your audit, consider which pages users most often access after coming to your home page. Are these the pages you expect? What insight does this reveal about user behavior?
If your site has a high-traffic conversion path, such as a booking path, you will want to make sure each page and feature in that booking path is consistently available to users. ObservePoint’s Web Journeys can simulate user traffic, moving through these conversion paths on a set schedule (e.g. every 15 minutes) to make sure they are functioning properly.
Feature Highlight: Web Journeys
ObservePoint’s Web Journeys make it possible for companies to test their most important web experiences and conversion paths. Web Journeys replicate your site’s user journeys, such as shopping carts or user logins, from start to finish, and tell you if anything prevents the path from completing or if the analytics are not tracking the activity.
19. Top Ten Pages
Identify which pages on your site consistently have the most traffic. Doing so will help you prioritize future auditing efforts. Make sure Adobe Analytics is properly installed on these top pages, either through a manual on-page audit or via automation.
An ObservePoint Web Audit can scan your top pages on a set schedule (e.g. every hour, every day, etc.) to verify that Adobe Analytics is properly installed on each page.
An important tracking limitation to consider is whether the influx of traffic recorded on these pages will exceed your account’s data collection limit.
For example, a “Low-Traffic” entry will appear in your reports when more than half a million unique page names have been recorded for the month. The page name will still be recorded, but unless it is a highly trafficked page, it will only be accessible in Data Warehouse, not in the reports.
In your audits, review how many unique page names you are likely to record in a given month. If they exceed 1 million, consider broadening the page name granularity to reduce the number of unique pages recorded.
20. Landing Pages Review
Optimizing landing pages is an important step in ensuring marketing spends are being used effectively. As part of your audit, look at the highest trafficked landing pages and verify whether they meet your expectations. If a landing page is over- or under-performing your expectations, conduct an on-page audit manually or via automation to verify that Adobe Analytics is installed correctly on that page.
An ObservePoint Web Audit can scan your landing pages at regular intervals or during planned campaign activity to ensure they are available and tracking correctly.
21. Worst Performing Pages
Normally you might not pay attention to the worst pages on your site, but taking a moment to consider which pages have the lowest performance can help you understand if visitors appear to be using your site as designed, as well as help you eliminate clutter if the pages are no longer relevant.
Sometimes pages that look like they’re not performing well may just have broken analytics tracking. Periodically check your worst pages to make sure analytics tracking is working properly.
Scheduled audits of your site with ObservePoint can help you catch tracking errors early so you can minimize data loss.
Expedite and Automate Your Audit with ObservePoint
Clearly, a thorough Adobe Analytics audit is a significant undertaking, and one with a short shelf life at that. Remember, you can trust the quality of your data only as far back as validated by your last audit.
Of course, the problem is that for most enterprise websites, a page-by-page audit is nearly impossible to execute manually, which is why automation is so important. Ongoing, automated audits help restore trust in your data quality, and can reduce the burden on the teams responsible for verifying the accuracy of their analytics implementation.
ObservePoint’s automated tag governance and analytics testing solution can help. Schedule a demo with an ObservePoint rep to learn more.
About the Author
LinkedIn More Content by Clint Eagar
