TL;DR
How modern marketing became better at acquiring customers, and more dependent on infrastructure it doesn't control
Most marketing leaders can tell you how efficiently they are acquiring customers. CAC by channel, marginal returns, how much more budget Meta or Google can absorb before performance starts moving in the wrong direction. We have become very good at answering those questions.
There is another one we rarely hear in the same conversation: how much of that acquisition does the company actually control?
This has come up repeatedly in conversations we have with growth teams. A company may have a sophisticated acquisition operation, strong unit economics and several channels working at scale. Yet when we start looking at what sits underneath that growth, a surprisingly large part of it can depend on a handful of external systems: their auctions, their algorithms, their data and, ultimately, their rules.
Nothing has to be going wrong for this to matter. In fact, the interesting cases are often the companies where acquisition is working extremely well.
Platforms have become exceptionally good at removing friction from growth. Meta and Google provide access to enormous pools of demand. Apple and Google give software companies global distribution, payments and trusted storefronts. Marketplaces have done much the same for commerce. Companies accepted dependence because the alternative was building capabilities that were slower, more expensive, or simply impossible to reproduce at the same scale.
That was a rational trade.
The question is what happens when the trade becomes so embedded in the growth model that nobody thinks about it as a trade anymore.
At REPLUG, this has changed one of the questions we ask when looking at acquisition. Alongside performance, we increasingly want to understand what the company is building while it buys growth. Does each year of investment leave the business with more ways to reach customers again, or does next year's growth still require purchasing essentially the same access all over again?
That distinction is becoming more important as marketing itself changes.
Look at what has happened inside paid acquisition over the past few years. Audience selection, bidding, placement decisions, and increasingly the combination and delivery of creative have moved deeper into the platforms. Google's Smart Bidding uses Google AI to optimize bids at auction time using contextual signals. Performance Max goes further across bidding, budget optimization, audiences, creative and attribution, working from the objectives, budgets, conversion data and assets supplied by the advertiser.
For most teams, using those systems is the sensible decision. They can process more signals and react faster than a human media buyer. Keeping a person in control of every lever because that used to be how campaigns were run would be a strange definition of progress.
But it changes the job.
A strong acquisition team increasingly needs to be excellent at the inputs: economics, measurement, creative, strategy, and deciding what the system should optimize toward. More of the execution connecting those inputs to an outcome happens inside technology the company cannot reproduce itself.
We see the consequence when teams talk about diversification. Adding another channel is often treated as reducing dependency, and in one sense it does. Moving from two major platforms to five reduces concentration. But if all five relationships work in roughly the same way- the company provides capital, data, creative, and an objective while an external system determines much of how those resources reach customers - the company has diversified its suppliers without necessarily changing the underlying model.
This is the paradox worth discussing more than the usual discussion about rising CAC: efficiency and dependency can grow together.
The better the platforms become, the more rational it is to use them. And the more useful it becomes to know which capabilities the company still wants to develop itself.
The advertising auction makes this easy to understand. A team can improve its creative, measurement, and customer targeting and still pay more for the same customer because competition for the inventory has changed.
The same dynamic now reaches much further into the customer relationship.
Apple's evolving approach to alternative payments in the European Union is a good example. In August 2026, Apple announced another change to its EU business terms. From October, alternative payment options can sit alongside Apple In-App Purchase, while Apple continues to apply commissions to digital goods and services under the new structure; apps distributed outside the App Store are subject to a Core Technology Commission on digital transactions. These terms replace a fee structure Apple had introduced previously.
The exact percentages will probably continue to receive most of the attention, especially from developers directly affected by them. The interesting part, however, is the mechanism. A company can acquire the customer, build the product, and create an alternative transaction journey, while another company still has enough control over the surrounding infrastructure to change the economics of that relationship.
This happens in less visible ways too. Attribution rules change. Access to data changes. Distribution rules change. Features that marketers built strategies around disappear or get replaced. Sometimes those changes make the ecosystem better. Sometimes they create a commercial problem. Either way, businesses adapt to decisions made somewhere else.
Public companies are unusually explicit about this because they have to describe the risk to investors. Duolingo says its mobile apps are accessed almost exclusively through Apple and Google and warns that changes to fees, advertising practices, data availability and attribution can affect how it markets and monetizes its products. Match Group describes similar exposure around app-store payments, user data, CRM and paid marketing efficiency.
These are highly capable companies. Their disclosures are useful precisely because this is not a story about bad execution. You can execute extremely well inside an ecosystem and still have limited influence over how that ecosystem evolves.
This is usually where the conversation moves to “owned channels.” I don't think that language is particularly helpful.
Very little digital distribution is genuinely owned. Organic search depends on Google's algorithms. An app depends on operating systems and app stores. Email and CRM rely on infrastructure providers and on customers continuing to give you permission to contact them. A website still sits inside a stack of browsers, devices, payments, and other technology the company does not control.
There are, however, meaningful differences in how much access has to be purchased again.
A customer searching specifically for your brand is economically different from someone you can reach only by winning another advertising auction. The same applies to someone who opens your app every week, responds to CRM, or returns directly to your website. Third-party infrastructure is still involved, but the business does not need to recreate the entire relationship from zero every time it wants another interaction.
Paid acquisition can be very good at creating this kind of access. Someone acquired through Meta today may become an app user, join a loyalty program, subscribe to communications, return through branded search, and eventually recommend the product to somebody else. The important question is what happens after the first paid interaction.
This is also why I think companies often underestimate the strategic role of their mobile app. We still see businesses treating the app largely as another conversion destination or another channel to optimize. For a company with a recurring customer relationship, it can do something much more valuable: give customers a reason to come back directly, transact, discover products, and continue the relationship without every interaction beginning with another media purchase.
The app is obviously not independent infrastructure. Apple and Google still sit underneath it. But a customer opening an app because they already have a relationship with the brand is a very different distribution position from a company that needs to purchase the next point of contact.
I find it more useful to think about temporary access and accumulated access. Some marketing spend gives you access to demand while you are paying for it. Other investments can leave behind brand demand, a direct relationship, repeat behavior, organic discovery, or an audience you can reach again.
The difference is what remains after the investment has been made.
There are companies already making versions of this trade, although they rarely describe it in these terms.
Trivago increased advertising spend by 21% in 2025, primarily because it put more money into brand marketing intended to increase direct traffic over the long term. Its global ROAS fell from 132.1% to 128.4%, with the company attributing much of that decline to the continued brand investment even as performance-marketing efficiency improved.
I like this example because it is messy in exactly the way real capital allocation is messy. Trivago did not discover that performance marketing was bad and switch it off. It accepted pressure on a metric it could measure immediately because it believed changing how people reached the business had value over a longer period.
Cars.com makes the same idea visible from another angle. It reported that roughly 60% of its 2024 audience arrived organically and explicitly described that as reducing its reliance on search-engine marketing. Reformation's 2026 IPO filing says roughly three-quarters of its direct-to-consumer new customers in 2025 came through unpaid marketing channels, which it links largely to brand awareness.
These numbers are not benchmarks. A startup entering a new market should not look at them and conclude that 60% organic traffic is the target. The economics and maturity of the business matter enormously.
What they show is that the structure of demand itself can have value.
We see the reverse situation surprisingly often with established brands. They may already have substantial awareness, millions of customers and large media budgets, yet the mobile relationship is underdeveloped. The company owns a lot of consumer attention in the everyday sense of the word, but has done relatively little to turn that existing demand into direct app usage, repeat behavior, or a customer relationship it can activate again. In those cases, the opportunity is not necessarily finding another acquisition channel. It can be making more of an asset the company already has.
That is a very different growth problem from the one faced by a new startup that simply needs reach.
There is a practical problem with all of this. The investments that change the structure of demand are often harder for marketers to defend than the next dollar of paid acquisition.
If Meta is producing customers below the CAC target, the argument for another €500,000 is easy to understand. A request to improve organic App Store visibility, build CRM capability, strengthen direct app engagement, or invest in brand demand can be much harder to compare because the return may appear across several places and over a longer period.
Marketing tends to respond by explaining the channel: why ASO matters, why CRM matters, why brand matters. That is rarely the strongest argument to take into a leadership meeting.
The business question is more useful. How concentrated is our growth in platforms whose pricing and rules we don't control? How often do we have to pay to reach customers we have already acquired? If we make this investment, does it give us meaningfully more influence over how customers find and return to us two years from now?
Those questions don't magically make long-term investments good. “Building the brand” and “owning the customer relationship” can become convenient excuses for work that produces little measurable value. Marketing leaders still have to demonstrate progress and connect investment to business outcomes.
But the measurement horizon does not have to end with the return from a single campaign.
We have had conversations with teams where almost every growth discussion starts with how much additional budget the current channels can absorb. That is an important operating question. It becomes more interesting when you add another: if those channels stopped scaling efficiently, what have we built that would give us another way to grow?
Every marketing investment should produce an outcome. A customer, a purchase, a reactivated user, demand that eventually turns into revenue. Long-term arguments about control do not make weak unit economics disappear.
I would simply add a second question to the evaluation: what remains afterwards?
Sometimes the answer is a profitable customer. That's enough. In other cases, the same investment can also contribute to brand demand, a direct customer relationship, repeat behavior, organic discovery, or an audience the company can reach again.
This is why I don't think the goal is independence from platforms. It would be unrealistic for most companies and, given how effective the platforms are, probably undesirable. Nor is there a sensible universal ratio between paid and organic, rented and owned, or performance and brand.
A young company may need to rent most of its distribution for years. An established consumer brand may have the opposite problem: plenty of demand already exists, but too little of it has been translated into direct digital or mobile relationships. The right mix follows the economics and maturity of the business.
What should change as a company grows is its understanding of the trade. Which parts of the growth engine are external because that is economically advantageous? Which capabilities are worth developing internally? And is today's marketing activity giving the company more options for tomorrow, or simply requiring it to buy a larger version of the same access again?
Two companies can report very similar CAC and ROAS while having very different answers.
So the question I would put alongside the performance report is this:
If the economics or rules of our most important distribution platforms changed tomorrow, which parts of our growth engine would we still be able to influence ourselves?
The answer doesn't need to be everything.
But marketing leaders should know what it is.
--
Sources
[1] Google Ads Help, “About Smart Bidding.” https://support.google.com/google-ads/answer/7065882
[2] Google Ads Help, “About Performance Max campaigns.” https://support.google.com/google-ads/answer/10724817
[3] Apple Developer, “Changes for apps in the European Union,” updated August 18, 2026. https://developer.apple.com/support/apps-in-the-eu
[4] trivago N.V., 2025 Annual Report / Form 20-F. https://www.sec.gov/Archives/edgar/data/1683825/000168382526000006/trvg-20251231.htm
[5] Cars.com Inc., 2024 Form 10-K. https://www.sec.gov/Archives/edgar/data/1683606/000095017025029023/cars-20241231.htm
[6] Reformation, Form S-1, 2026. https://www.sec.gov/Archives/edgar/data/1787117/000110465926077832/tm2513004-7_s1.htm
[7] Duolingo, Inc., 2025 Annual Report. https://www.sec.gov/Archives/edgar/data/1562088/000162828026025737/duolingodecember312025an.htm
[8] Match Group, Inc., 2025 Annual Report. https://www.sec.gov/Archives/edgar/data/891103/000089110326000068/matchgroup2025annualreport.pdf