Marketing Analytics Lenses: A framework for attribution

This week’s Newsletter-14 marks part of the extremely-necessary Prologue that will inform the application for how to navigate Measured.

If you want to learn more about this project, you can do so right here.

Let’s get right into it.

Marketing Analytics Lenses

One key factor to the ability to distil complex problems, isolate the question to be answered and navigate large sets of data, is the art of learning how to apply the best fit-for-purpose reporting lens. 

Reporting lenses isolate specific perspectives or viewpoints through which data can be analysed, interpreted and visualised. They can be shaped and influenced by the purpose of reporting and the intended audience that should view and make key decisions from the output of the report. 

For attribution reporting in particular, lenses offer a clear guide to reduce complexity and noise in marketing reporting by isolating a subset of data to focus on what will provide value to the stakeholder group that is asking. 

After 12 years as a data-driven marketing professional, I’ve distilled down the six key lenses every marketer should understand and use as a framework for knowing exactly where to look to solve their attribution and analytics challenges. 

The OPTICS Framework

The hard problem of attribution is the art of identifying and assigning value to the marketing touchpoints and interactions that contributed to generating a desired action. Most popularly, attribution focuses on the return on spend or return on investment made within a particular channel, but marketers want to assign value to the multiple channels they deploy in an omni-channel strategy to drive specific program outcomes. These extend beyond the outcome of spend to things like customers retained, customer lifetime value, return on campaign vs business-as-usual (BAU) messaging and most importantly, return on the strategy; whether it be brand, performance or loyalty focused. 

Due to the unpredictable and uncertain nature of consumer interactions, it isn’t possible to gain an exhaustive set of data from which to model a perfectly simulated environment from what actually occurred. As a result, marketers should measure the best set of sampled data they have, applying the best fit statistical techniques, all while understanding the limitations. The OPTICS framework provides six attribution lenses that can be used to slice data to provide confident answers to key marketing questions with the best data set available.

The OPTICS framework offers six simple reporting views to break down marketing’s most asked questions to assign value to marketing mix and omni-channel experiences.

The application of reporting lenses can vary based on the channels deployed in an omni-channel marketing strategy. The first step is to isolate the question to be answers so that the right data sample can be isolated to bring visibility to the effectiveness of strategies and tactics. Through using these lenses to understand where to look, time spent reporting can be efficient and support insight-led improvements to be made over time. 

The Latest in Marketing & Analytics

If you don’t get enough attribution chat from me, listen to this cracking talk from Tim Wilson on what to keep in mind when embarking on your multi-touch attribution journey. If you don’t already listen to the Analytics Power Hour podcast, Tim Wilson is consistently full of gold. Run don’t walk. I am diving into the depths of Retail Media Analytics at the moment. I thoroughly enjoyed this report from Mi-3 giving the view in the Australian landscape. One of my team members came across this resource via Johan Van de Werken who was kind enough to log some starter queries for recalculating key sessions and metrics out of BigQuery following your export from GA4. Thanks Johan!

Week in review.

I had a fantastic chat with a peer this week (who I met at Measurecamp earlier this year) and we had a great chat about “not MECE-ing your Marketing Funnel Reporting”. 

Beyond that, this discussion led to an ask to turn this concept into a practical example that illustrates how marketers can leverage their walled-garden reports alongside their touch attribution to better optimise activity.

So naturally, I jumped at the chance. Here’s where we landed. A practical breakdown of how we use Walled-Garden and Touch Attribution to measure the effectiveness of channels. 

Yes, this is where we hold two informative metrics side by side without feeling the need to add the numbers together. The metrics that best measure the effectiveness of display advertising at the top of the funnel can be VIEW-through while all of the non-linear, high intent conversions can be CLICK. 

“But I do get CLICK based conversions coming through from Facebook and Display in my Google Analytics Report?”

That is because the funnel isn’t linear. 

Your top of funnel efforts have been built largely for the purpose of driving awareness and nurturing non-aware customer cohorts through to become engaged prospects. If we so happened to intercept the exactly perfect person in our ripe market to buy – we’ll take ‘em – but it wasn’t the purpose of the activity. Therefore, viewing your GA4 DDA TOUCH based Model Reports (like these ones) is naturally going to focus a little more on bottom-of-funnel or performance activity.

Image with thanks to Krista Seiden from Google who you should absolutely follow if you don’t already. 

Notice how organic and paid search seem to always take out first place in these reports? I know what you’re thinking (because I get asked this all the time) and yes, there is some “bias” but it’s not the bias you think. Just like the examples above, Google can simply match more data for Google based channels (because it has all of it’s own Google data) so it has more data points to attribute. That being said, your bottom of funnel activities will take out first prize in your DDA touch models, pretty much always. Whether you’re running Facebook Lead Ads as pointy end tactic, or paid search. 

The rule to remember: 

Don’t send a ruler to do the job of a thermometer. 

The measurement approach should be fit-for-purpose for what needs to be measured. If you are driving reach and awareness, measure it that way by measuring who viewed, not who clicked. 

Want to submit a foundational marketing metric for me to deep dive into?  Reach out to me on LinkedIn or via email at kate@attributiongal.com All submissions will be attributed in the final manuscript.

That’s it for Week 14.

Stay curious folks.