How To Build An Attribution Model

Knowing how to build an attribution model — or at least knowing how to run attribution models — is critical to increasing revenue and ROI. There will be no going back to the ways of yesteryear; not while more and more transactions take place online rather than in line at a brick-and-mortar store.

Agencies and their clients must know how to build attribution models in order to make the most of their ad spend. It’s also crucial in order to inform future advertising and marketing spends that are based on data harvested in real-time and analyzed by experts.

For many advertisers — especially smaller businesses or larger companies who have yet to harness their marketing efforts to powerful marketing mix models —  this may sound intimidating.

But if you’re a business owner wondering how to build an attribution model, the proper course of action is quite simple.

Learning how to build an attribution model depends on one’s answer to another question: Who is your audience? Once we learn this metric, then building an attribution model follows a prescribed course of action.

Attribution Models: How To Build

We’ve entered a cookieless world. We must find new ways of engaging with customers. To do so, we need to map the customer journey.

Consider how many touch points there were the most recent time you purchased a product or service. Chances are it wasn’t a straight line; there were likely multiple stops along the way. 

These touchpoints indicate to marketers how customers discover a brand; how they engage with it; and how and why they move their way down a sales funnel.

This is why attribution models are a necessity: They illuminate these paths and help guide a business’ future marketing decisions based on what the paths reveal about consumer behavior.

To provide a broad overview of how to build an attribution model, we can break it down into a few main segments.

First, we’ll want to identify funnel stages and then set trackable goals. Is our client looking to grow revenue or simply raise brand awareness?.

Next, we’ll need to trace the customer’s journey to see precisely which marketing channels they navigated to get to you and, hopefully, a conversion.

We’ll be capturing data throughout this process, of course. And running in the background is an awareness of cost-per-acquisition (CPA). No calculation of ROI is complete — or even possible — without it. We must know how much each conversion costs the business so as to make more informed decisions moving forward.

Finally, we test. Testing helps ensure we’ve chosen the right attribution models that will deliver insights into customer behavior that will continue to inform future ad spend.

As we’ve written, “marketing attribution is the search for and identification of user processes that lead to website sales. It is the analysis of a series of actions, known as user touchpoints, that lead to one of any number of desired business outcomes.”

A customer’s journey may include searching for a product or service online, engaging with the brand on social media, and then ultimately making a purchase. 

If a potential customer bails on the process and jumps out of the sales funnel, a business should want to know why so that it can course-correct, re-engage with that consumer, and encourage them to complete their transaction — i.e., buy something.

What’s more, as we analyze the data, we begin to learn which touchpoints create conversions for which consumers. That makes it easier to incorporate CTAs (calls-to-action) into the sales funnel in order to capture more of these consumers.

Thus, knowing how to build an attribution model means knowing your audience and assigning credit to marketing touchpoints.

Leavened: Speed, Transparency, Value

At Leavened, we use a market mix modeling (MMM) approach. This helps our clients understand the total impact of their marketing channels on incremental sales.

Unlike other methods, MMM understands the value of online as well as offline media channels. Plus, MMM takes into account external influences, such as price, distribution, inventory, and other market factors. In short, MMM is a trusted and valued marketing method that’s used by the largest global brands today.

The difference between pre-Leavened and post-Leavened approaches to building an attribution model comes down to three things: speed, cost, and transparency.

The reason this method of modeling was until very recently the domain of bigger, more profitable companies was because models cost a lot of money to run and took a long time to deliver results. Smaller companies often had neither the time nor the budget to run these sophisticated attribution models.

But now is not then. Today is not yesterday. The future of MMM and attribution models is now in the hands of companies of just about any size and marketing budget. With our powerful suite of tools, we can help you find the connections between online activity and offline sales, among many other pathways.

Any brand with daily or weekly levels of spend, impressions, and/or GRPs can now analyze the impact of ad spend on aggregated unit sales, revenue, and brand metrics.

Leavened quantifies incremental lift in revenue over time for online and offline marketing tactics.

In short, our Marketing Mix Modeling is the most efficient, affordable, and transparent optimization tool there is.

Leavened knows how to build an attribution model that works for each of the agencies and companies with whom we partner. Digital marketing campaigns and custom marketing attribution models have never been more accessible — or more powerful.

From paid search to creating a custom attribution model to pinpoint conversion paths, we’ve moved beyond multi-touch attribution models and singular conversion credits into a new realm of time- and position-based marketing attribution. We know more than ever how users interact with our brands thanks to these new and powerful data-driven tools of analysis.

If you’re ready to learn to run it yourself, get in touch with Leavened today.