AA vs CJA: Basic Metrics
29 Jun 2025 » Platform , Analytics Tips
I am currently working on a project where my client is in the process of transitioning from Adobe Analytics (AA) to Customer Journey Analytics (CJA). A few months ago, I helped another customer with their CJA implementation. While AA and CJA are similar, they are definitely not the same beast. Of the many differences that they have, I will start today with the basic metrics.
Why Metric Definitions Matter
If you are migrating to CJA, you may notice that your numbers do not match exactly with what you saw in AA. This is not a bug, it is a reflection of how the platforms are designed. You also need to make sure this knowledge is propagated, you do not want an executive calling you in alarm because the numbers do not match.
Let us examine the details.
Page View vs Hit
Do you remember those old website counters from the early days of the web? Tracking user activity has evolved considerably since then.
Adobe Analytics
The basic metric in web is page view. In a traditional website, every time a page (HTML) is rendered by a browser, we count it as a page view. With single-page applications (SPA), there is no page load, but the rule of thumb is that if the content changes by more than 50%, we count it as a page view.
Although AA was designed for the web, when mobile apps disrupted the market, it was not too far-fetched to equate app screens to web pages. I know that mobile implementations are very different from web applications, but we could continue to apply the same 50% rule as with SPAs. This meant that AA could also be used to track mobile apps.
Technically, either s.t() or MobileCore.trackState() needs to be invoked for a page view to be recorded in AA.
The other option that we have with AA are clicks (or any other interaction), which are tracked with s.tl() or MobileCore.trackAction(). These calls are also recorded but do not increment the page view counter.
Customer Journey Analytics
CJA, on the other hand, was created to process any type of event: page views, web clicks, phone calls, email opens, journey steps… So, while page views will be one case that we will often see, it is not the only one. The decision was then made that any event processed by CJA will be called a hit. In other words, hits just count the number of events. All page views are hits, but not all hits are page views.
If you still want to count page views, you can create a custom metric for events with eventType == web.webpagedetails.pageViews. However, this relies on the Web SDK and Mobile SDK implementations to use the recommended eventType values.
Visits vs Sessions
Marketers want to know how many times a user has interacted with their brand. Things get blurry with this ask: while we can detect when a user starts interacting with a brand (lands on the home page, opens the app…), we can never be sure when they stop. We have all seen friends and coworkers with dozens of tabs open. Or maybe you are one of them 😉. Does this mean that they are interacting with all tabs simultaneously? In other words, is keeping a browser tab open enough to count as “interaction”? If a user is browsing a website, attends a phone call for 45 minutes, and resumes browsing, should this count as one or two interactions?
Adobe Analytics
Given this uncertainty and to unify criteria, the industry came up with a standard definition: temporary interactions with a website were called visits. A visit starts with the first page view and ends after 30 minutes of inactivity (i.e., no tracking calls from the browser for 30 minutes).
To account for some edge cases, Adobe Analytics also imposes some additional restrictions to visits.
Virtual Report Suites (VRS) provided some flexibility to the rigidity of this definition, where you could define the visit according to your needs.
Customer Journey Analytics
CJA follows the same approach as VRS and allows you to configure the start and end of a session, which is the new name for this metric.
Do not get confused with the concept of a browser session, which is different from this CJA session. You have more details in my old post Of Visits and Sessions.
Visitor vs People
Finally, we want to know how many real people are behind those visits and page views.
Adobe Analytics
As with most digital marketing tools, AA identifies users through cookies. In order to calculate unique visitors over a period of time, AA just counts the total number of distinct ECID cookies it has seen during that period.
As I explained in the linked post, ECID cookies are unique per browser and can be deleted. The consequence is that a single user will end up having multiple ECIDs. Therefore, the visitor metric is always going to be inflated.
I would like to add a couple of clarifications:
- AA does not count the number of ECIDs generated, only those in the event calls received by its servers.
- The old ECID Service API has a method called
setCustomerIDs(), but it has no implication over the number of visitors.
Customer Journey Analytics
When you create a CJA connection, one of the parameters that you have to select is the Person ID. CJA uses this parameter to deduplicate visitors that are actually the same person. If there is no Person ID associated with an ECID, the ECID will be counted as a person, but when it can be linked to a Person ID, all ECIDs belonging to the same Person ID will be collapsed into one.
CJA calls the number of Person IDs people, which is the new metric that you will see in your reports.
This has two important consequences:
- The people count will always be lower than the number of visitors. In an AA to CJA transition, there will be no like-for-like comparison of these two metrics.
- The people count of past dates will decrease over time, as CJA gets more links between ECIDs and Person IDs. Multiple ECIDs will become just one person.
Key Differences at a Glance
Let me finish with a summary of these metrics in table format, for your convenience:
| AA | CJA | What is different? |
|---|---|---|
| Page view | Hit | CJA counts all events |
| Visit | Session | CJA sessions are flexible |
| Visitor | Person | CJA deduplicates users |
Photo by Mitchel Boot on Unsplash