The new decade is off to the races, and the world of mobile marketing continues to evolve through an never-ending series of headwinds and tailwinds. Here are a few of the tectonic plates underlying trends in the mobile marketing industry that will present challenges or opportunities for strategic planning for marketing teams and their stakeholders.
The Big One: User Privacy
Potentially one of the biggest and most disruptive tectonic plates churning beneath our feet is that of privacy-first trends.
The ripples of the Cambridge Analytical scandal, years of headline-topping massive data leaks, and rising user concern over the privacy of their data continues to instigate waves of privacy regulatory frameworks in GDPR, COPA, and CCPA with more on the way. These will create more restrictions, costs, risk, and process to scaling and/or driving ROAS from digital advertising. Shutting off Europe was an answer that some advertisers took with GDPR, but the expansion of these frameworks makes this non-viable in 2020 and beyond (not to mention the need to tap into Europe to differentiate from saturated US prices).
Not only has the shifting of user privacy plates forced governments to act, but the strategic players in the industry as well. Immediately this led to Facebook’s nerfing of its audience insights tool, removing certain non-first party targeting abilities in addition to the most recent voluntary elimination of impression-based attribution (shared through Grow.co newsletter). Critically more momentous is what is brewing at Apple: a black swan-magnitude event caused by the prospective loss of the IDFA as a reaction of Apple to concerns over user privacy.
With the advent of limit ad tracking, some 30–40% or more of iOS users in the US and large English markets have adopted this privacy feature. To protect its position as the top ROAS channel Facebook rapidly and proactive began limited ad targeting on users with LAT ON. In Apple Search Ads savvy advertisers began playing a head-in-the-sand ostrich protection game in favor of fully attributable ROAS reporting by simply using demographic targeting to turn off LAT-ON user targeting; yet in the case of Apple Search Ads, advertisers have over time begun to feel the dual stings of:
- Missing out on a large and growing larger swaths of user inventory
- Worsening margins caused by the pricing wars pushing up auction prices as a result of advertisers competing in LAT-OFF waters becoming more and more crowded.
Returning to the black swan scenario: if Apple really were to get rid of the IDFA, a few of the significant outcomes would include:
- A breakdown in ad targeting, from the hyper-lucrative event and value-optimized systems, to lookalikes, to remarketing.
- A breakdown in analytics, preventing proper cross-session user attribution to acquisition sources/costs, causing a reliance on averages or probabilistic assumptions.
- A forced return to installing network SDKs to provide tracking performance numbers and the loss of the objective source of attribution truth that is/was the MMP (which could put even more dollars into the pockets of the oligarchy).
- A potential return to non-self-attributing, fingerprinting-supported networks (where fraud, scale and inventory quality may be more of an issue).
Ultimately the industry would figure out a new solution, but the immediate effect could be a re-leveling of the playing field as all players scrambled to figure it out. In this leveling out, the advantage would likely go to:
- Nimble new entrants who could learn how to do without high-fidelity, source-confident ROI tracking from day one
- Resource-rich players who could leverage their resources to develop or hone techniques like data-science powered incrementality modeling.
For exhaustive details on this subject from one of the industry’s most prolific strategists, turn to Eric Seufert’s recent article detailing the deprecation scenario of the IDFA.
The Blue Pill: Algorithm-Powered Marketing
On the other side, in the case that the black swan IDFA deprecation does not occur, the major trend will be the continued rise of algorithm-powered marketing. Algorithms like event and value optimized bidding have driven incredible ROI for advertisers so far and are sure to grow more powerful, if left undisrupted. Some ideas at where algorithms may move next:
- Multi-event bidding — for instance, optimizing for users who not only open the app three times within the first week, but who also complete three language lessons. This is currently being done advertiser-side via synthetic event optimization strategies.
- More powerful subscription optimization naturally through additional data points on who subscribes, and potentially via some new bidding model such as LTV-based bidding.
- More powerful in-app revenue bidding — optimizing for ad revenue whales (solutions like AppsFlyer, Iron Source and AppLovin are moving forward with solutions to address the conundrum of not being able to track ad revenue down to a user level, which will open up better pipelines for tracking user lifetime value).
- Expanded components of non-purchase event optimization — similar to value-based lookalikes, enabling some lever in event optimization beyond revenue that can differentiate user value by a secondary component of an event, beyond the simple fact that it occurred, such as the number of hours after install that the event occurred within.
Generally, algorithm-powered marketing will continue to get better at spending advertisers’ budget as these algorithms feed on more data connections, and moreover new types of targeting (e.g. Facebook trial or subscription events, Snapchat’s event-optimized bidding or Tik Tok’s soon-to-come event optimized bidding) will improve as they reach critical masses for each type of event, market and operating system.
This will also create defensible moats around big budgets which can afford to keep algorithms stocked with mounds of data and therefore offer them a competitive advantage in targeting than smaller budgets, which will continue to reduce the marketer’s ability to arbitrage targeting for better results and place more emphasis on the analyst/scientist’s role in powering campaigns by ensuring budget and bids are as high as possible. The price to access the top auction positions though will continue to be top-notch creative, as well as the ability to monetize users at a rate as close to their potential value as possible (segue to the next section).
Enter Greatness: Product/Marketing Sharing Goals/Outcomes
Increasingly, the limiting factor for marketers will be the monetization or retention of a product, irrespective of the outcome of Apple’s black swan event. Even if a marketing team can set up the proper data pipelines to power targeting and supply performant creative, the product will still be what determines how well the users can be monetized and thus how much the marketing team can pay per user, and therefore how many of the total addressable auctions the app’s ads can possibly participate in.
If one of twenty users from an event optimized campaign should convert based on user quality, yet the app only converts one of fifty, then the marketing team will be forced to throttle bids and budgets, which will lead to limited growth. Not only that, but the inability to convert purchases robs the algorithms of data points vital to targeting, which further degrades the overall effectiveness of the marketing campaigns.
The most successful growth teams in 2020 will be those that enjoin marketing, product and analytics in order to align all parties around a single shared goal of profitable growth. Marketing and analytics need to collaborate to determine the optimal CAC to pay for users without either over-paying and causing a medium-term panicked pullback OR defaulting to conservative assumptions (again, think big budgets & bids will win and establish the moats in 2020).
Marketing and product need to align on the product roadmap and understand when new changes are expected to affect monetization trends without gaps in this critical knowledge sharing. Product and analytics need to collaborate to rapidly study how the forecasts and proper CACs should evolve as the product evolves.
The more these three functional areas continue to operating in siloes or with disparate goals, the more profit or growth will be left on the table, and the more an app will be susceptible to being bullied by more aligned, less siloed growth programs and the more the app will risk falling behind in the algorithm-powered arms race to either stagnate or die.
Additionally, teams that can figure out how to incorporate the process of liveops or customized UX onboarding based on a user’s signals of potential value will enter a category of greatness. This will unlock incremental scale and profit from every program and represents a forward-looking ability to arbitrage or hack growth independently of creative droughts, rising auction competitiveness, lack of data budgets, or product monetization weakness.
For instance, offering more early-stage freebies or perks (or deeper discounts if initial efforts fail) for users coming from a value-optimized campaign or those exhibiting strong natural LTV signs could reduce the risk of losing these whales. Or, turning up the push/in-app marketing, ad monetization, or experimental UX testing rates for an installs-sourced user or churn-likely user may help notch up retention and ROI without unnecessarily risking retention.
Show Me the Data: Unlocking Growth Via Analytics
The tectonic plate of analytics is sure to be ever-more essential to powering growth programs than ever before, also irrespective of the outcome of Apple’s black swan event.
To begin, we have touched on the need to understand the directly attributable LTV of advertising-acquired users to power bidding and reporting; but this is really the table stakes requirement that all teams should be doing. The teams that can go beyond simple blended CPIs and master the science of truly re-attributing credit from the organic catch-all bucket towards marketing efforts will be rewarded with the ability to invest more confidently and actually to grow the organic bucket faster.
The industry perception of organic downloads as sacrosanct and believing that organic is truly organic is a confounding one. After all: where do your first users come from, who spread the brand awareness and bring more users into your organic bucket? From your first press release, your friends and family marketing and your initial advertising campaigns! Only after that does organic start to take on a life of its own, so to speak; but compounding credit to “pre-organic” sources will continue to build.
The nature of organic is that it is a return on investment from marketing efforts from the start of the company to the present day. In attribution terms, organic is largely the last-click attribution of other efforts and is an outcome of other efforts such as organic keyword rankings, PR initiatives, advertising, user referrals, and more. Therefore, to assume that organic is 100% independent of paid marketing efforts is silly.
What is not silly and quite thorny is not the fact that organic has roots in inorganic activities, but rather the challenge of quantifying the true source of organic. This is the true issue standing in the way truly valuing inorganic marketing efforts; yet this is a challenge to be solved, rather than to be ignored. To learn how to re-attribute the portion of organic activity that is influenced by different marketing efforts is to establish the ability to power the source of organic and grow organic even more.
The flip side of this is incrementality. While paid marketing can and should be credited with some percentage of organic-bucketed activity, paid marketing can and should also be discounted in order to reflect the fact that some people would have downloaded an app, uninfluenced on their own, in which case the paid marketing download is cannibalistic to growth (or perhaps cannibalistic of an earlier paid marketing effort).
Brand search downloads in particular are another confounding industry inefficiency. While necessary to ward off conquesting to maximize the ROI of marketing efforts should a searcher be tempted not to download the exact thing they searched, brand ads should be discounted for the inevitable level of cannibalization that they produce. Impression-based attribution can also produce more cannibalization than click-based attribution.
Analytics teams that can build and create the support with stakeholders to trust custom, not solely last-click source attribution or incrementality models such as the following will be able to unlock additional growth via a smarter, more sustainable program.
Re-allocating credit from organic to paid should be done for some of the following marketing efforts:
- Tracking gaps caused by Apple’s LAT on users, Google’s expanded iOS search inventory, which is untrackable outside of the network.
- Tracking gaps caused by non-trackable advertising interactions such as traditional media, connected TV, influencer, podcasts/audio ads and more.
- Marketing credit mis-attributed to organic but due to K-factor/organic uplift.
Re-allocating credit from paid to organic based on cannibalization of advertising. This is most manifest in ads placed in stores (i.e. Apple Search Ads and Google UAC); although in the latter case the recent industry shift led by Facebook to eschew impression-based attribution may mitigate the concern of non-store ad incrementality.
The methodologies for re-allocating credit for these types of advertising may leave something to be desired by stakeholders used to the precision of deterministic reporting, but a few popular options include:
- Pre-post lift studies, especially as powered by a statistical model such as Facebook’s prophet model.
- Market or audience holdout study. Ad networks have realized the need for incrementality studies and begun to offer support for these, such as Facebook’s test and learn.
That’s all for today! Thanks for reading and stay tuned for more posts breaking down mobile marketing concepts.
Be sure to bookmark our blog, sign up to our email newsletter for new post updates and reach out if you’re interested in working with us to optimize your app’s mobile advertising or ASO strategy.
Incipia is a mobile marketing consultancy that markets apps for companies, with a specialty in mobile growth marketing, analytics & business intelligence, and creative production. For post topics, feedback or business inquiries please contact us, or send an inquiry to email@example.com