What to Focus on in Marketing Attribution

What to Focus on in Marketing Attribution systems


Marketing attribution has never been more impactful, yet complicated as it is today. Thanks to the influx of marketing channels and devices, layers of complexity are added to what was, a few years ago, a simple funnel.

Concurrently, data is easier to track, and there is a hype around tracking as much data as possible.

On one hand Companies with legacy data systems scramble to drill into marketing analytics without first renovating disjoint systems, On the other hand others overcomplicate attribution and spend endless resources to yield little actionable insight and lose on what to focus on in Marketing Attribution.

Despite the complexity, majority of marketers cannot say with certainty exactly how valuable each channel or touchpoint is during each customer’s purchase process.

Marketing attribution is either missing from their analytics or overly complex. The result is the inability to optimize marketing budget. An effective attribution approach is what distinguishes an organization in the dark about its marketing spend from one that knows exactly how each channel is performing and allocates spend accordingly.

What to Focus on in Marketing Attribution systems


1.Select a model that fits the business model, and test it frequently

First click, last-click, U-shaped, time decay, Markov model, etc. are all types of models ranging in complexity that are used to attribute marketing value. However, none of them paints the exact picture, and every model, no matter how advanced, has its flaws. It is far more important to select a model that fits with the company’s business model as opposed to one that is simply the most advanced. To determine this fit, the organization needs to consider the nature of the customer journey.

Next, the selected model needs to be tested regularly, given the nature of the consumer journey

inevitably varies with time. The perfect attribution scenario could no longer apply in a different season, where customers come through a seasonal campaign directly instead of multiple clicks through other digital channel.

2.Be dynamic with industry changes

There is nothing more detrimental to an organization’s marketing health than to rely on systems and models that are antiquated and require significant manual stitching. Legacy data systems need to be renovated to be able to connect with digital site data. Analytics should ideally be performed via data platforms instead of disjointedly across countless Excel documents.

Conversely, as new models are created, organizations need to evaluate their fit and adopt if they fit the business model more closely than existing models.

3.Maintain alignment between teams

Another non-negotiable for attribution success is that all parties need to agree on the goals of the attribution program, the key metrics involved, and the models being tested. The goal of attribution is to produce more revenue by spending more effectively, but without clear communication, some teams may fear that they’ll lose budget — and therefore status — and get territorial.

Attribution isn’t just a marketing issue. Installing more complex models will involve tech teams, and at the higher levels, the finance teams should be in on the goals of the program from the beginning. It’s generally acknowledged that over-crediting the bottom of the funnel (e.g., Google or Bing) means that funds will be unnecessarily tight for upstream channels. But if you can use attribution to show the value in those upstream channels, the finance folks might be persuaded to loosen the purse strings a little — but those folks need to be included in the attribution talks from the onset to understand the significance of the data.

4.Focus on LTV, not just conversion

Many companies simply track conversion and traffic as performance metrics. However, the fulcrum of an effective attribution program is the lifetime value, or LTV, of a customer.

Among the metrics your teams need to understand, the most important, a good attribution program, is the lifetime value (LTV) of a customer. Certain channels may not run a high conversion, but brings in high value customers such as the average LTV of a customer ends up higher than other channels. Furthermore, if you rely only on single-conversion events as the ultimate goal, upstream clicks get undervalued. The traffic coming in at the top of the funnel looks a whole lot more important if the end result is measured over time, not one single purchase.

5.Know when enough is enough

That perfect attribution scenario mentioned above?

You might as well accept the fact that you’re never going to get there — or, at least, not for the foreseeable future (and beware the tech provider that says otherwise). Every model, no matter how advanced, has its flaws.

U-shaped, Game Theory, econometric, Time Decay, etc., are all nuanced and complex, and they give you more insights than simple last-click. But none of them paints an exact picture, so part of the secret of “mastering” attribution is knowing when good is good enough and when the pursuit of the ideal is just a dangerous rabbit hole.

WINDSOR.AI’s Marketing Attribution

Windsor.ai automates your data pipelines, integrates and brings all your data to life with advanced AI. We rely on a customized approach that adapts to the nature of your business and consumer journey. Don’t hesitate to request a free demo of Windsor’s Marketing Attribution Software, to see how it helps you get the best out of your marketing activities.

You Might also like to read:

Shopify CLV and net revenue attribution

Post-View and Post-Click in Attribution: An Overview

Data driven attribution models

What is Marketing Attribution? Definition and Types of Models