The 4+1 Types of Digital Analytics Solutions
If you’ve been following this blog for a while (in which case: thank you!) you already know how much I love Adobe’s analytics solutions, like Adobe Analytics and Customer Journey Analytics. And while I can’t shut up about my excitement over every new feature, I rarely talk about the strategic and positioning aspects. The most I’ve written about that is when I was comparing Adobe Analytics and Google Analytics and gave my take on where people should go after the UA sunset.
In some of my recent conference talks and my Adobe Analytics Masterclass, that strategic component got a lot more attention. For the Masterclass, I even dedicate a whole half-day (25% of the training!) to strategic decisions and incrementality considerations. So today, I want to share some of my takes on how I think about different solutions and what makes them special.
To do that, I’ll first cover how I prefer to structure the large amount of available solutions. After that, I’ll go through each category and talk about the tools within those categories. So, without much more ado, let’s get started!
Evaluating a Digital Analytics Solution
Alright, so how do I think about digital analytics solutions? Just like everyone else, I came up with a mental model that helps me classify the countless available tools and solutions on the market. For me, my categories come from evaluating two critical qualities of any solution:
- The Out-of-the-box (OOTB) value. This dimension tries to capture the (professional, but still subjective) feeling when you first open up a new tool. Does it make you feel like you are a small child walking into a toy store? Or does it feel… empty? Do you go in and are immediately inspired to click on things, try stuff out, and marvel at all of the newly discovered options? Or do you want to ask “where’s all my stuff”? Of course, this dimension is somewhat subjective, but something I’ve been able to apply to every tool I’ve seen.
- Customizability. Different from the first dimension, this second one tries to describe what you can “teach” the tool. No single solution on the market can ever offer all the functionality a company needs OOTB, so being able to teach it some new tricks and make it truly work for you and your needs is essential. While some solutions are built from the ground to be customized, others are built with a “take it or leave it” mindset that doesn’t allow any customization.
With those two dimensions, we can build a neat visual representation through a coordinate system. If we separate the high and low performers for each dimension, we end up with those four quadrants:
How handy! With those four quadrants, we can easily sort any available solution into a category that helps us compare it to others in the same or another cluster. But to do that, we first need some examples and names for the categories, so let’s get to those! Spoiler warning: There even is a secret fifth category waiting at the end!
Category 1: Reporting tools
First, we have the category in the top left corner. Tools from this category are high in out-of-the-box capabilities, but rank rather low on customizability. The best example for a tool from this category is the old Google Analytics 3/Universal Analytics.
Through almost two decades, GA was the de-facto standard in digital analytics. For many, it still defines how an analytics solution should work, look, and feel like. Its many builtin reports and dashboards have inspired many other tools, including which reports we expect and metrics like our beloved Bounce Rate. If someone opened GA for the first time, they immediately got a feeling of “I’m an analyst! Look at me, mom, I’m doing the analytics!” because of the slightly nerdy but still accessible real time landing page.
On the flip side, the amount of customization in GA or other tools of this category is very limited. Sure, we can define custom metrics and dimension, but those are always treated as something completely separate from the builtin capabilities. Dashboards are a joke, and the options to collaborate with others in your company are practically non-existent. That’s why many companies started with GA but quickly outgrew it.
In our categorization, this is where I would put GA/UA:
With reporting tools as the first category, let’s now look at the opposite extreme!
Category 2: BI Tools
In the bottom right corner, our second cluster holds what I would describe as BI tools. As BI tools are designed to be agnostic and ingest any data imaginable, they are naturally high in customizability. Of course, that comes with the price of very low OOTB value. They are, quite literally, empty until you start creating content in them.
Some good examples for this second category are tools like Power BI or Tableau. Being positioned purely as visualization tools, they are far less opinionated about what you should do with them, especially when compared with tools like GA3/UA. While you can build amazing, visually stunning and interactive reports, you need to teach them every small bit of business logic before you can even think of analyzing bounce rates and Unique Visitors. On top of that, tools like this always require a technical expert to build reports with the right logic.
A bit different from those visualization tools but still in the same category are tools like Big Query. While those usually don’t come with a graphical interface, their SQL or SQL-like capabilities make them very flexible. While it is super impressive what some wizards can get out of a data warehouse (DWH) with SQL, the command line-like interface makes it again something only experts can handle confidently.
Those two types of tools, visualization tools and SQL DWHs, are usually combined into some variation of a “modern data stack” or “BI stack”, requiring large teams with dozens of experts to provide even the most basic analytics capabilities to the company. With those come long processes and large delays in decision making, which is especially unfortunate in the agile, short-lived digital space.
In our coordinate system, here’s where I would put this category of solutions:
Now that we have our first two clusters, let’s look at the next category, which is…
Category 3: Analytics Tools
In this category, we have finally arrived in the most exciting category. Analytics tools combine a high out-of-the-box value with a high level of customizability. For tools of this category, there is very little difference between the capabilities that everyone gets the features and workflows that are exclusive to your company. With a tool from this category, we are invited to start using it immediately or take some time to make it truly our own, increasing value even further. Before you read on: Take a guess which examples I would put into this category.
You are completely right, this is Adobe’s category. Adobe Analytics brings everything you expect from a modern digital analytics solution and some truly unique features. While its unparalleled accessibility for advanced attribution or analysis features are impressive already, it manages to combine these OOTB capabilities with many, many options for customization. Any builtin metric or dimension can be used with any custom event or dimension, pathing works across all types, and attribution can be used for everything. Correctly implemented, there is very little it can’t do.
In this logic, Adobe’s Customer Journey Analytics ranks a bit lower on OOTB value but a bit higher on customizability. Adobe is already working on providing some easier access to basic digital analytics functionality, yet it can feel a bit empty when compared to Adobe Analytics today. It more than compensates for that with the exciting new features it brings (like Derived Fields) so that it absolutely deserves to be ranked in this category.
Our much more complete coordinate system now looks like this:
With three categories defined, it’s time we look at the unfortunate category of…
Category 4: Time Wasters
As you can imagine, this category of tools doesn’t come with a lot of OOTB functionality and sadly can’t be customized in any meaningful way to provide more value. Tools from this cluster often try to differentiate themselves by appearing as especially easy to learn or privacy friendly, usually on a more than questionable basis.
Matomo is a good example for this list of low-lights. Besides being infamously slow and old-looking, the low number of builtin features with the practically non-existing customizability makes it a big waste of time and effort. In the same sense, one-page-analytics tools like Fathom or Plausible try to appear as easy to learn but are not even worth deploying, given that it takes only one analysis session to outgrow the capabilities they provide. As I’ve written before, even log file analysis tools are more capable.
Another sad entry in this category is Google Analytics 4. Even its most vocal advocates have gone on record complaining about how immature and devoid of features it is. While it is pretty obvious that Google tries to turn GA4 into a data-collection tool without any analysis features, some still try to make it work to analyse data, with growing frustration.
In our quadrant logic, here’s now the full picture:
And there you have it: All four categories of tools with examples for each category! But Frederik, you might say, what about the others? What about Amplitude, Heap, Mixpanel, and all the rest? I’m glad you asked! For those, I have one more category in stock.
Category 5 (bonus): Challengers
Of course, there are a lot of tools that I would place somewhere in the middle of all the others. While those have a somewhat-mature reporting or analysis interface, they can also do some neat custom things. That’s why I put them into their own bonus category, right in the center.
Amplitude is a good example. It can do an impressive number of things already and has a super flexible data model. And that’s great! However, I always struggle to find coherent UI and UX concepts within the tool. From the outside, it looks like a typical startup company, where a lot of features are built in isolation without anyone really caring about the large picture. This results in duplicate work, incoherent user experience, and quickly changing interfaces.
In a couple of years, maybe some of those tools will have moved into another category. Depending on what they prioritize for future feature development, they could very well cater towards some different user base that is looking for a specific set of features. Right now, they try to be too many things at once for my liking, making it hard for me to recommend them.
Here is now the complete picture for all categories and tools:
Conclusion
So, that’s how I currently view all of the available digital analytics solutions and how I structure them in my mind. With those 4+1 categories, I’m quickly able to put any new tool into a respective cluster. For example, whenever I see a tool similar to Fathom or Plausible, I can now quickly discard it and put it into the “not worth my time” corner. Of course, there is a lot of personal opinion, experience, and strategy involved into any categorization, so everyone is very welcome to share their own perspective.
You might have noticed that I sneakily included one more tool in the final chart without mentioning it in the text. Right at the top, above the challengers, I’ve put the amazing Piwik PRO. I went on record before in praising it as the ideal replacement for GA3/UA, which is still what I believe today! If you’ve never heard about it, you should definitely check it out.
That’s all I have for this post. Enjoy the rest of your day!
German Analyst and Data Scientist working in and writing about (Web) Analytics and Online Marketing Tech.
2021 – current Adobe Analytics Champion
EMEA Adobe Analytics User Group Lead
Adobe Analytics Community Advisor