Stitched customer paths using AI to see the full customer journey  

During any given day at Corvidae, someone will mention “rebuilding” data at some point. This process of rebuilding data by probabilistically stitching together touchpoints of a customer’s path is habitually understood here in the office, but it may not be familiar to the casual onlooker, so allow me to explain. 

The Objective 

We all track the usage of our websites to better understand how well our marketing efforts are speaking to our customers, which messages are resonating, and which are not, and we do this with the ultimate goal of improving our bottom lines. 

The Status Quo (and the problem) 

To that end we have historically employed tracking solutions which rely on a cookie to map our customers’ interactions with our marketing, our websites, apps, and the frequency of their visits. The problem with cookies is they are short-lived, ephemeral little snippets of data which exist inside a single browser, on a single device, for a short period of time. The moment the user switches devices, clears cookies, deletes their history, or moves from your website to your app the old cookie is lost and another is born. In the eyes of your tracking software, they have just become a completely new human. 

The unfortunate side effect of this limited lifespan is that when we track a conversion – be it a sale, subscription, or registration – we can only see interactions which have happened on the customer’s current device, in their current browser, within the lifespan of the living cookie. The result is journeys which tend to contain only the last interaction or two before the user became a customer, completely ignoring all the earlier and probably influential interactions that led to their decision to convert. 

So when you’ve seen customers click on a single ad then commit to your offering immediately and thought “that can’t be it,” you’re right. That probably wasn’t it. 

This all adds up to a skewed view of the effectiveness of the advertising in the earlier parts of our marketing funnels, as we’re generating awareness and persuading consideration. It is instead shifted in favour of marketing activities which are designed to convert only those nearing the end of their decision journey and those who already have a high purchase intent. This audience segment is highly competitive and very expensive as a result. 

The Solution: Stitching 

The only real solution to this problem is to bridge the void between broken sessions to create customer journeys that include all stages of the marketing funnel, tying awareness, consideration, and conversion together. This is what we mean by “rebuilding” data: creating a dataset that tells the truth about the path your users take before buying into your offering. 

At Corvidae, we approach this in two distinct ways: Deterministic Stitching, and then Probabilistic Stitching. 

Your web and app data in its raw form is a bundle of unrelated pageviews and conversions. Deterministically we use a series of first-party cookieless IDs stored at different scopes to stitch together the events we are certain are related; functionally similar to a cookie-based process but fully compliant, with greater coverage, and improved capability. 

Once the “easy” joins are made, we already have more complete journeys than you could expect from a previous-generation web tracking suite, we then move to probabilism. 

Using location, device, and behavioural markers in clickstream data, we determine the likelihood of two seemingly separate users actually being individuals or individuals from within a single household who have simply switched devices or changed location. When the likelihood is statistically high enough that two users are actually one, we blend the sessions from both paths together and iterate. We do that again and again recurrently until we have constructed an end-to-end view of all customers’ journeys through your marketing mix.  

These complete and well-formed journeys make it possible to do a few things: 

• Capture customers earlier, before they are exposed to your competitors or reach a saturated, expensive to compete in the marketplace. 

• Remove wasteful activity which isn’t driving your customers to action. 

• Identify impactful touchpoints which were invisible under more short-sighted lenses. 

• Capitalise on performant activity which has headroom for greater spend. 

• Shortcut users to conversion with targeted activity at pivotal points in the customers’ decision-making process. 

Explore Corvidae 

If you’re interested in exploring what Corvidae can do for your organisation, book a demo with one of our team! 

Our Corvidae experts are able to walk you through the solution, explore its suitability, answer any questions you might have and give you an overview of what you can expect from the onboarding process.