Building a virtuous cycle between product and marketing

One of the more fraught and contentious topics I see mobile advertisers grapple with is the alignment of product and marketing. Very progressive companies are coming to see the tight integration of marketing and product as a means of gaining maximum leverage from the primary drivers of app distribution: advertising on the Duopoly. In practice, “integrating marketing and product” means doing three things:

  1. Identifying the best existing users;
  2. Identifying the best potential users;
  3. Deciding which of those groups is most valuable and adapting the product to it.

For some conservative companies, especially those with product-centric leadership that rose to prominence before the current mobile paradigm, the “marketing” function takes a completed product and distributes it via advertising channels: the marketing team’s purview should be restricted to the advertising funnel and shouldn’t extend into the user’s interaction with the product.

It’s easy to understand and empathize with this mentality, since it is seems intuitive and is consistent with the way that non-digital industries operate. And there are other reasons that letting the marketing team influence product development might give pause:

  1. Product managers have deep domain expertise in how users interact with products and how various features can change a user’s trajectory with a product. Even seemingly small changes to the early-funnel experience (eg. the tutorial) can radically alter a user’s understanding of and desire to continue using a product. The idea of undermining that expertise with feedback from marketing campaigns is uncomfortable;
  2. Product development takes time — it’s much costlier and more time consuming to build a feature than it is to test an ad creative. The product team can’t ship updates as fast as the marketing team can ship campaigns, and this misalignment of cadences could cause endless thrash within the product team;
  3. The product team shouldn’t necessarily be reacting to fads and short-term trends. Companies should take a long-term view of their products and should be building features that are meant to serve users not just in the immediate wake of a marketing campaign but for years into the future.

All of these are valid concerns, but the reality is that algorithmic campaign optimization is a force that app developers need to accommodate: better to lean into it than try to fight it. A healthy balance should be struck between using marketing insights and product insights to guide product development. But a developer also has to ask itself, “Who am I building this product for?”: is it my existing user base, or is it my potential future user base?

This question may seem like an exercise in useless pontification, but it’s important. In Two fundamental problems with product A/B testing, I described why the composition of DAU is such a critical consideration when testing product features: a developer should seek to not only optimize the product for the people that are using it today, but also for the people that will use it in the future, who might look very different from the people that are using it today. “Looking different” tends to mean having been acquired from different sources: as products age, they tend to rely more on paid acquisition for new users, and so the composition of a product’s DAU base changes in that way over time.

I have seen very stark differences exist in the way that “paid” and “organic” users interact with a product. Users that were exposed to an ad for a product before installing it tend to have higher levels of information about and motivation for that product. What’s more, the information encoded into marketing data is invaluable, because it essentially tells the developer what it should build if it wants to acquire more of some paid audience. I call this the Who and What of product / marketing alignment:

Recognizing the Who and What of product / marketing alignment creates a powerful incentive for the developer to tailor the product to the most valuable users, which ultimately creates a feedback loop:

  • The product is tailored to accommodate the behaviors of the best users;
  • Those product updates seep into marketing assets, either consciously or subconsciously;
  • More of the most valuable users are acquired.

This process may seem like a natural evolution of an intuitive product development course — who wouldn’t do this? — but for it to take shape, a number of things are required. The first is a recognition of marketing audiences as they exist in the product: a “source” flag attached to all user profiles that allows the DAU-base to be segmented by origin. This is actually very rare to see in the wild; as I’ve noted ad nauseum in posts about “Growth” teams and A/B testing, the disconnect between product-led experimentation and marketing can be painful and often handicaps actual growth in cases where the product ages out of its previous acquisition dynamics (eg. when organic growth tapers off).

That “source” recognition cuts both ways: if marketing teams don’t consider the size of potential audiences as well as the down-stream metrics that contribute to value — which, ultimately, is ROAS — then they can’t enable the virtuous cycle of product / marketing alignment. Additionally, if marketing teams aren’t actually segmenting their targeted audiences in a meaningful way, which is accomplished through radical creative testing, but are instead relying solely on demographic features like geography to segment their campaigns, then they won’t be able to achieve the audience demarcation necessary to guide product updates and changes.

What all this means is that marketability — not just for users that look like the existing DAU-base but for the total addressable market of the product — should be, at the very least, an input into product development decisions. The opposite of a virtuous cycle is of course a vicious cycle: optimizing the product for a dwindling, low-value audience and alienating more valuable user segments in the process.

Photo by Mario Dobelmann on Unsplash