Within the next 18 months, Performance Marketing as we know it will undergo a radical change. Decisions that were previously informed by user level data at scale will soon be informed by a much more limited set of data. Such a shift will change performance marketing as we know it. This will be the result of three compounding factors:
- The removal of user level data by Google & Apple (IDFA, AID)
- Facebook & Google’s chokehold on customer acquisition costs
- The incessant tidal wave of hypercasual games
In this article, I will outline how GDPR (General Data Protection Regulation) & CCPA (California Consumer Privacy Act), which at their core are healthy & very much needed privacy regulations designed to protect consumers, are driving this. I will also cover why this will result in less data, higher costs, fewer alternatives to Facebook & Google and what marketers can do about it. Finally I will cover why I believe this will happen sooner than most people expect.
1. The removal of user level data by Google and Apple
As an ever-dwindling number of us remember, the introduction of performance marketing in 2011 was chaos. Standardized attribution models or out-of-the-box server integrations with advertising networks did not exist, leaving decision makers with only aggregate and approximate data, it was bare bones.
This changed with the introduction of the UDID — the Unique Device Identifier -and with it, the ability to understand KPIs and behaviors at the user level. It eliminated the need for deduplication, generated better insights, and birthed the entire attribution industry.
Jump forward to May 1, 2013. Apple began rejecting apps submitted that used UDID. This was because the UDID was problematic in a number of ways, all driven by the fact that it was permanent and unchangeable. Marketers who had understood their value, began amassing UDIDs to build profiles that lead to better contextual advertising (i.e. never show a fried chicken advert to a vegan), paving the way for performance marketing as we know it.
Apple’s successor to the UDID was the IDFA, which debuted in 2013. The biggest difference between the two was that they could reset this identifier at any time to sever the links to the profiles that marketers were building. The reality was, however, that few users knew this, or did it.
Despite the UDID moment of 2013, performance marketing’s growth has not slowed, and in fact has actually grown at a seemingly unstoppable rate in the face of declining spend on other platforms. Digital advertising spending worldwide — which includes both desktop and mobile devices — stood at an estimated $194.6 billion in 2016.
However, change is in the air, with multiple signals now pointing to Apple retiring/removing IDFA by the end of next year. This would force Google to follow suit by removing AID. The first signal that this is simply a matter of time comes from Facebook’s recent VTA changes:
If we take a step back to look at Big Tech’s policies, imminent change comes into sharper focus. Apple is inconsistent, not allowing third party cookies in Safari to prevent web traffic tracking, but still allowing IDFA for tracking on mobile. Google has already announced it is removing third-party cookies from Chrome by 2022. Such inconsistency cannot last.
Apple will eventually get to parity on both web and mobile. In one of the most recent operating system updates (Mojave), Apple prepared for the removal of third-party cookies on the web in the adframework library. Their most recent release here contains these changes, there will be no more third party cookies, Period. My hypothesis is that they are going to update the adframework for iOS (maybe as early as iOS14 in June this year, although now unlikely given the current, uncertain, environment) to follow suit.
The result of the removal of user-level tracking will be nothing short of seismic for the industry. Marketers have got very comfortable to accessing user-level to make marketing more efficient, more targeted, and more contextual. A shift to aggregate data removes the ability to behave in ways that performance marketers have taken for granted. The main areas that will be affected are:
- Look-a-likes & audience targeting. The process of creating a campaign based on a set of user identifiers that have the attributes you are looking for. For example, an advertiser can take all user IDs that have paid in the last 30 days and ask Facebook to find other users on their network that show the same characteristics. When marketers cannot upload a list of user identifiers at all, this can obviously no longer work.
- Retargeting. The process of trying to communicate with customers who have lapsed from your product, by uploading their identifier to a system like Braze for Push Notifications, or Facebook to talk to them directly in a news feed advertisement. This will not be possible for the same reasons as above.
- Estimating and Monitoring ROI . The deployment of user-level data to identify the source and cost of a user evaluate their Lifetime Value (LTV) down to its root causes from retention to purchase conversion to purchase frequency to purchase amounts. Imagine a cohort of 100 users, generating $100; instead of knowing that one person spent $100 and the other 99 did not contribute (and so should be ruled out in future campaigns), the advertiser will now only know that the average value of all the users is $1. This level of information will send marketers back to the dark ages.
While these are the main aspects that will suffer, there will certainly be others as the full impact of this change hits the industry. We will see a reversion to a world that looks something like 2012, where many small-walled gardens of users forming a long tail, with the short tail dominated by Facebook and Google.
2. Rising customer acquisition costs on Facebook & Google
Compounding this removal of user level data, advertisers will face ever-rising costs as they get pushed toward Facebook and Google. Given that they are for all intents and purposes the judge, jury, and executioner, why are costs rising?
Facebook and Google were built to be advertising platforms. In fiscal year 2018, 85% of Google’s revenue was from advertisement. Facebook is a similar story, reporting $16.9 billion in total revenue for the final quarter of last year, up 30 percent year-over-year, 98% of which came from advertising.
The market valuation of a publicly-traded company is invariably closely tied to their earnings. Facebook and Google are no different. Their focus is on maintaining or growing their revenue in the face of user base growth that has slowed appreciably. The result is a market-wide increase in eCPMs, because of more demand than supply, and the fact that targeting on those platforms is getting harder and harder due to existing user-level approaches becoming defunct, leading to much higher inefficiencies in spend. Facebook’s CFO is already highlighting the impact of user level data and privacy implications in its most recent earnings report:
“First, the recent regulatory initiatives like GDPR and now CCPA have impacted, and we expect they’ll continue to impact, our ability to use such signals. Secondly, mobile operating systems and browser providers such as Apple and Google have announced product changes and future plans that will limit our ability to use those signals. And then, finally, we have made our own product changes that gives users the ability to limit our use of such data signals to improve ads and other experiences. And there, I’d point to something like the rollout of Facebook activity controls, and that’s at 100% today.
So each of these factors limits our ability to target and measure the effectiveness of ads on our platform, and that can negatively impact our advertising revenue growth. Both Mark and Sheryl talked about the importance of ad targeting for small businesses, and I think it’s important to note that the regulatory and platform changes will have a proportionate impact on the ability of small businesses to use ads to grow and thrive.”
So — advertisers are increasingly pushed toward the short tail of Facebook and Google to drive app discovery, but these platforms are getting steadily more expensive over time. This phenomenon is not new, and the general trend has been obvious and consistent for years now. However, in years past this price increase would have been a challenging, though manageable in reality since there were always alternatives outside Facebook and Google to spread marketing dollars. That has now changed.
3. The incessant tidal wave of hypercasual games
Historically, the most appealing place to put any performance marketing budget outside of the short tail of Facebook and Google has been larger ad networks that have a large SDK footprint in lots of apps, or demand-side platforms (DSPs). These sources hold the next largest number of potential advertising impressions. Examples of these networks are companies like Applovin or Ironsource.
This has essentially vanished as a strategy because these networks have almost entirely specialized into the business of hypercasual games. Hypercasual games have been in the marketplace ever since Apple opened the app store but started to multiply in 2018. They are best characterized being easy to make and easy to play. Low production costs allow their developers, in contrast to traditional mobile game developers, who typically spend a year or more on developing larger, more complex games with much higher LTVs and much longer life cycles. These simple games are often built to have mass appeal and aggressively monetize with in-game display ads. This has allowed the developers that build successfully-scaled hypercasual games to accumulate high returns, as well as large amounts of user-level information — critical to a rotating universe of games with short life cycles but large reach.
The proportion of players that have any form of payment means (ie, credit cards) is very low. Hypercasual essentially created a large melting pot of low value, low attention span impressions, whose only true value is to other games that fit the hypercasual profile (future games by the same developer, or by other developers). So marketers, lacking the user-level data they used to have, and faced with rising costs in the short tail, are now feeling the compound effect of declining value in the other large pools of potential advertising impressions due to the glut of hypercasual games.
So what’s next?
In summary, the three compounding changes to the performance marketing ecosystem have resulted in less data, higher and rising costs, and less alternatives to Facebook & Google. This is, of course, until a new disruption happens, but that has yet to happen at any meaningful or enduring scale (yes, I know… TikTok). If discoverability was hard before, it will only get harder.
What can you do about it?
The first and worst answer is nothing, and hope that by petitioning the platforms we can change policy and avoid the removal of user-level tracking. Assume that the rising eCPMs will plateau and that the performance marketing ecosystem will rebalance away from Hypercasual.
The best answer is to focus on creative & placement. I said at the beginning that we will move from a competition between who has the most user level data to a competition of who can make the most informed decision in the quickest reaction cycle with the aggregate data available. This means having world-class, scaled creative production coupled with actionable insights at a fast speed of iteration. It also means understanding your placement strategy, knowing that different messages appeal to people in different ways in different places. This is essentially a pendulum swing back to somewhere between performance and product marketing, which has not been required for the past years because of the reliance on data.
The forcing functions I have outlined will cause a behavioral shift that should actually benefit the industry as a whole, as advertising creative quality & contextual relevance will increase, albeit painfully. Performance marketers will need to start thinking about how their strategy differs when talking to people via Snapchat versus Twitter, instead of lazily assuming the data will tell them what works and optimizing from the lagging indicators of KPIs. All of this while making decisions on aggregate, not user-level data.
Behavior alone will not be enough, though. That behavior must be accompanied by technology. Automation will be key, as the decisions made will be on less data and more often, but with less granularity. Accurately and dynamically predicting the value of a customer from small data sets will ensure smart decision making on the performance marketing portfolio. The other aspect will be creative: marrying the creative, copy, and placement together in ways that decrease eCPI and maximize IPM. Being able to do so while fighting creative fatigue across multiple partners and locations, and potentially hundreds of campaigns will be the difference between winning and losing. Leveraging machine learning will be a huge unlock that is still to really make an impact in the world of performance marketing, which is surprising, given the number of repeatable tasks.
The companies that come out of this moment victorious will be the ones that were proactive, decisive, and intelligent when thinking through the ramifications of the three compounding factors listed above. They will ultimately have built up the people, technologies, products to turn the challenges into opportunities.
At N3TWORK, we created the N3twork Scale Platform (NSP) specifically to combat the challenges we are all about to face. It is hard and getting harder, and we want to help. The NSP is a set of tools, technologies & expertise for game developers, that help turn great games into huge businesses. We offer capital and access to the same battle-tested automated mobile user acquisition technology used in all N3TWORK games, including the global hit puzzle-RPG, Legendary: Game of Heroes. NSP has enabled the deployment of hundreds of millions of dollars against mobile marketing at scale without degrading ad performance or in-game user economics. NSP allows game makers to focus on making and running an exceptional game by automating the creation and optimization of mobile advertising, including:
- Aggregation of marketing data and calculation of segmented, projected lifetime customer value metrics
- Creation of interactive ad creatives
- Deployment and management of ad campaigns
- Analysis and performance measurement required to efficiently deploy marketing budget at scale
Find out more about how you can build your winning moment by heading over to devnsp.wpengine.com