Answer:
Stop investing in things that don’t work; invest more in things that do.
If only it were that easy. The reality for many marketing teams? Limited visibility on opportunity progression post sales handover. Fragmented and incomplete ad platform data. And countless hours pulling data together in Excel.
Optimizing media mix is fundamentally a data challenge. And so the solution, naturally, is to get the data fundamentals—capturing, integrating, automating, analyzing—right.
This is where things get tricky. How do you determine a platform’s efficacy when increasing spend could also raise the customer acquisition cost (CAC)? Who do you trust when the ad platform reports different sales outcomes compared to your CRM? And what happens when big gaps appear because ad platforms report activities differently than first-party data?
The short answer is you need a neutral, single source of truth for revenue performance—a purpose-built platform designed specifically to handle the wide-ranging data demands of a revenue (sales and marketing) organization. And if you take nothing else from this article, know that the media mix optimization challenge is solvable from a technical standpoint.
Here’s the longer answer:
First, you need to capture marketing activity data, which will allow you to have full visibility of every customer journey and attribute revenue and cost to each touchpoint. Despite tightening privacy regulations, you can capture nearly 100% of all online activity using first-party, cookieless, server-side tracking technology, which does not get intercepted by ad blockers. If requests are sent from your domain, these requests are classified as first-party data, and they will not get flagged by browsers or ad blockers, and they’re GDPR and CCPA compliant because they do not collect personally identifiable information (PII).
Second, connect marketing activities to revenue. This is the capability that most marketers lack. Getting platform-based cost data from Google Ads, Facebook, and other media platforms is relatively simple; connecting that data to actual user activity and revenue is hard. But it can be done. An ETL (extract, transform, load) solution can be used to extract, transform, and load data from many sources—CRM systems, media platforms, and other GTM databases—and integrate them into a single, unified data warehouse or database.
To make a homegrown solution work, it’s important to (1) link ad platform spend with UTM values that are present in systems that record customer and revenue data, (2) architect a hierarchy in which unified data can be connected and analyzed with a data visualization tool, and (3) ensure that the data is labeled consistently and completely. Still, this basic solution would not provide the most complete and accurate data due to tracking constraints introduced by ad blockers and restrictive privacy policies.
Third, attribute. Full-funnel reporting requires that you collect the entire user journey, not just tracking data from a single conversion event. If you’re collecting all user activity, then applying different attribution models to the data lets us select the right lens for analysis, and the best modeling method will correspond to the desired objective. For example, if our goal is to maximize reach, then it would make sense to use a first-touch attribution model to determine the channel that is most effective at generating first touches. If increasing conversion rate is the goal, then a last-touch or later-stage-weighted model would be more suitable. It’s crucial to recognize that one model is not necessarily better than another. Instead, it’s important to look at the data holistically and in the context of your goal.
The Sona platform delivers all three capabilities seamlessly in the form of a marketing cost dashboard. As a Sona user, this would be your jumping-off point to optimize media mix. The dashboard shows, in real time, the cost of every click, lead, and sale along with the milestone stages that are relevant to your business. You’d be able to compare how each channel performs under the same attribution model and then allocate the optimal budget to each.
To conclude, the optimal media mix isn’t a set-and-forget formula to figure out. It should be a regular, rebalancing process that includes pruning and adjusting how budgets are allocated across channels and campaigns. To do this effectively, it can only be enabled by establishing full visibility into user activity, marketing initiatives, revenue, and cost. It is an ongoing process that requires both skilled marketing expertise and the right accompanying data and insights.
--
What about marketing mix modeling?
Marketing mix modeling (MMM), as the name suggests, takes a broader view of marketing. Namely, it accounts for the 4Ps—product, price, place, and promotion—as well as other macroeconomic factors such as seasonality to formulate the optimal marketing plan using statistical analysis.
It is typically done once a year through a consultancy-driven engagement and can be quite costly, not to mention time consuming. This might make sense for large B2C businesses with big traditional media investments (TV, radio, outdoors) but not digital-native businesses that thrive on agility, innovation, and experimentation.
Some proponents of MMM may cite low attribution completion rates and a lack of visibility on cost as a reason to use it over revenue attribution modeling. But as we’ve explored above, this is not the case at all.
That said, we believe it is not appropriate to compare MMM to revenue attribution modeling because they’re fundamentally different. MMM concerns itself with the macro, GTM-strategy level questions, while revenue attribution focuses on channel and campaign level performance to optimize ad spend.