Adobe Analytics (also known as Site Catalyst, Omniture, and various other names both past and present) is a service that tracks and reports on how people use websites and apps. It’s one of the leading solutions for organisations who are interested in studying how people are actually using their digital offerings.
Studying real-world usage is often far more insightful, in my view, than surveying people before or after the fact. Competitors to Adobe Analytics would include Google Analytics and other such services that allow you to follow web traffic, and answer questions from those as simple as “how many people visited my website today?” up to “can we predict how many people from New York will sign up to my service after having clicked button x, watched my promo video and spent at least 10 minutes reading the terms and conditions?”
In their own words:
What is Adobe Analytics?
It’s the industry-leading solution for applying real-time analytics and detailed segmentation across all of your marketing channels. Use it to discover high-value audiences and power customer intelligence for your business.
I use it a lot, but until recently have always found that it suffers from a key problem. Please pardon my usage of the 4-letter “s word” but, here, at least, the Adobe digital data has always pretty much remained in a silo. Grrr!
There are various native solutions, some of which are helpful for certain use cases (take a look at the useful Excel addin or the – badly named in my opinion, and somewhat temperamental – “data warehouse” functionality for instance). We have also had various technology teams working on using native functionality to move data from Adobe into a more typical and accessible relational database, but that seems to be a time-consuming and resource-intensive operation to get in place.
So none of the above solutions yet really proved to meet my needs to extract reasonably large volumes of data quickly and easily on an adhoc basis for integration with other datasources in a refreshable manner. And without that, in this world that ever-increasingly moves towards digital interactions, it’s hard to get a true overall view of your customer’s engagement.
So, imagine how the sun shone and the angels sung in my world when I saw the Alteryx version 10.5 release announcement.
…Alteryx Analytics 10.5 introduces new connectors to Google Sheets, Adobe Analytics, and Salesforce – enhancing the scope of data available for analytic insights
I must admit that I had had high hopes that this would happen, insomuch as when looking at the detailed schedule agenda for this year’s Alteryx Inspire conference (see you there?) I noticed that there was mention of Adobe Analytics within a session called “How to process and visualise data in the cloud”. But yesterday it actually arrived!
It must be said that the setup is not 100% trivial, so below I have outlined the process I went through to get a successful connection, in case it proves useful for others to know.
Firstly, the Adobe Analytics data connector is not actually automatically installed, even when you install even the full, latest version of Alteryx. Don’t let this concern you. The trick is, after you have updated Alteryx to at least version 10.5, is to go and download the connector separately from the relevant page of the Alteryx Analytics gallery. It’s the blue “Adobe Analytics install” file you want to save to your computer, there’s no need to press the big “Run” button on the website itself.
(If you don’t already have one, you may have to create a Alteryx gallery user account first, but that’s easy to do and free of charge, even if you’re not an Alteryx customer. And whilst you’re there, why not browse through the manifold other goodies it hosts?).
You should end up with a small file called “AdobeAnalytics.yxi” on your computer. Double click that, Alteryx will load up, and you’ll go through a quick and simple install routine.
Once you’ve gone through that, check on your standard Alteryx “Connectors” ribbon and you should see a new tool called “Adobe Analytics”.
Just like any other Alteryx tool you can drag and drop that into your workflow and configure it in the Configuration pane. Once configured correctly, you can use it in a similar vein to the “Input data” tool.
The first thing you’ll need to configure is your sign-in method, so that Alteryx becomes authorised to access your Adobe Analytics account.
This isn’t necessarily as straightforward as with most other data connectors, because Adobe offers a plethora of different types of account or means of access, and it’s quite possible the one that you use is not directly supported. That was the case for me at least.
Alteryx have provided some instructions as to how to sort that out here. Rather than use my standard company login, instead I created a new Adobe ID (using my individual corporate email address), logged into marketing.adobe.com with it, and used the “Get access” section of the Adobe site to link my company Adobe Analytics login to my new Adobe ID.
That was much simpler than it sounds, and you may not need to do it if you already have a proper Adobe ID or a Developer login, but that’s the method I successfully used.
Then you can log in, via the tool’s configuration panel.
Once you’re happily logged in (using the “User login” option if you followed the same procedure as I did above), you get to the juicy configuration options to specify what data you want your connector to return from the Adobe Analytics offerings.
Now a lot of the content of what you’ll see here is very dependent on your Adobe setup, so you might want to work with the owner of your Adobe install if it’s not offering what you want, unless you’re also multitasking as the the Adobe admin.
In essence, you’re selecting a Report Suite, the metrics (and dimensions, aka “elements”) you’re interested in, the date range of significance and the granularity. If you’re at all familiar with the web Adobe Analytics interface, it’s all the same stuff with the same terminology (but, if it offers what you want, so much faster and more flexible).
Leave “Attempt to Parse Report” ticked, unless for some reason you prefer the raw JSON the Adobe API returns instead of a nice Alteryx table.
Once you’ve done that, then Alteryx will consider it as just another type of datasource. The output of that tool can then be fed into any other Alteryx tool – perhaps start with a Browse tool to see exactly what’s being returned from your request. And then you’re free to leverage the extensive Alteryx toolkit to process, combine, integrate, analyse and model your data from Adobe and elsewhere to gain extra insights into your digital world.
Want an update with new data next week? Just re-open your workflow and hit run, and see the latest data flow in. That’s a substantial time and sanity saving improvement on the old-style battle-via-Excel to de-silo this data, and perhaps even one worth buying Alteryx for alone if you do a lot of this!
Don’t forget that with the Alteryx output data tool, and the various enhanced output options including the in-database tools and Tableau output options from the latest version, you could also use Alteryx simply to move data from Adobe Analytics to some other system, whether for visualisation in Tableau or integration into a data warehouse or similar.
A use case might simply be to automatically push up web traffic data to a datasource hosted in Tableau Server for instance, so that any number of your licensed analysts can use it in their own work. You can probably find a way to do a simple version of this for free” using the native Adobe capabilities if you try hard enough, but anything that involves a semblance of transform or join, at least in our setup, seems far easier to do with external tools like Alteryx.
Pro-tip: frustrated that this tool, like most of the native ones, restricts you to pulling data from one Adobe Report Suite at a time? Not a problem – just copy and paste the workflow once for each report suite and use an Alteryx Union tool to combine the results into one long table.
Here’s screenshots of an example workflow and results (not from any real website…) to show that in action. Let’s answer a simple question: how many unique visitors have we had to 2 different websites, each represented by a different report suite, over the past week?
Performance: in my experience, although Adobe Analytics can contain a wealth of insightful information, I’ve found the speed of accessing it to be “non-optimal” at times. The data warehouse functionality for instance promises/threatens that:
Because of to the complexity of Data Warehouse reports, they are not immediately available, can take up to 72 hours to generate, and are accessible via email, FTP or API delivery mechanisms.
The data warehouse functionality surely allows complexity that’s an order of magnitude beyond what a simple workflow like this does, but just for reference, this workflow ran in about 20 seconds. Pulling equivalent data for 2 years took about 40 seconds. Not as fast as you’d expect a standard database to perform, but still far quicker than making a cup of tea.
Sidenote: the data returned from this connector appears to come in string format, even when it’s a column of a purely numeric measure. You might want to use a Select tool or other method in order to convert it to a more processable type if you’re using it in downstream tools.
Overall conclusion: HOORAY!