Performance marketing strategy
The Brand-Stage Lens: Why Growth-Stage Brands Leak the Most
Where a brand is in its life cycle shapes where its marketing money goes and who controls it. Using a simple model built on audited fee data and industry fraud research, we estimate that early, growth and mature brands each lose a different share of their acquisition budget. Growth-stage brands, with half their spend outside the large platforms, lose the most per dollar.
One problem, three shapes
Discussions about ad-tech waste usually treat "advertisers" as one group. In practice, a seed-stage app founder, a fast-scaling consumer brand and a household name with 80 million users face very different versions of the same problem. They spend on different channels, they rely on different partners and they lose money in different places.
Looking at the leak by brand stage makes it far more actionable. It shows which brands are most exposed, and what each should fix first.
The three stages
The segments below reflect how marketing budgets typically evolve as consumer apps and digital brands grow. The channel splits come from our operating experience across the brands we work with. They're illustrative, not industry-wide statistics, and individual brands will differ.
Early stage
- Profile: fewer than 100,000 users. Seed-funded or bootstrapped, focused on getting users fast.
- Channel mix: around 70% of acquisition spend on self-reporting networks (SRNs) such as Meta, Google and Apple, and 30% elsewhere, with affiliates, DSPs, direct apps and OEMs.
- Who controls it: usually an agency or a single performance marketer. Founders rarely have deep marketing experience in-house, so they trust partners to run acquisition.
Growth stage
- Profile: 25 million or more users, or five-plus years in business, in expansion mode.
- Channel mix: roughly 50/50 between SRNs and non-SRN channels.
- Who controls it: an in-house performance team runs the large platforms. The non-SRN half is usually handed to one or two agencies or networks, which in turn bring dozens of sub-networks under them as billing partners.
Mature (saturated) stage
- Profile: 70–80 million users, 10+ years in operation. The focus shifts from acquisition to retention and customer lifetime management (CLM).
- Channel mix: SRNs fall below 30% of total spend. The rest is split between direct-app deals and CLM. Overall acquisition budgets shrink relative to earlier stages.
- Who controls it: often large holding-company agencies, sometimes bundled with deals to monetise the brand's own app inventory.
Why each stage leaks differently
The leak isn't the same everywhere. Two findings from independent research explain the pattern.
First, fees stack up most in the intermediated chain. The ISBA/PwC study found that only about 51% of spend reached publishers in the programmatic supply chain, with fees and untraceable spend taking the rest [1]. Buying directly from a large self-serve platform avoids most of those layers, though an agency fee (about 7% in the same study) may still apply [1].
Second, fraud concentrates outside the large platforms. AppsFlyer's 2026 fraud report found a 36x gap in fraud between affiliate traffic and SRN traffic [2]. The more budget a brand puts through affiliates and resold networks, the more it's exposed.
Combining these gives a simple model. Non-SRN spend loses about 55% (fees, untraceable spend and fraud). SRN spend loses only the agency fee, about 7%, where an agency is involved.
The model, stage by stage
Per $1 million of acquisition spend:
| Stage | SRN spend | Non-SRN spend | Calculation | Lost per $1M |
|---|---|---|---|---|
| Early | 70%, agency-run | 30% | 0.70 × 7% + 0.30 × 55% | ~$214K (21%) |
| Growth | 50%, in-house | 50% | 0.50 × 0% + 0.50 × 55% | ~$275K (28%) |
| Mature | 46%, agency-run | 54% direct apps | 0.46 × 7% + 0.54 × 22% | ~$151K (15%) |
A few notes on the method:
- Mature brands are measured on acquisition spend only, with CLM excluded.
- Their direct-app spend is modelled at 22%: the agency fee plus the unknown delta. Direct deals skip some programmatic layers but still lack price benchmarks and often run through agencies.
- Growth brands are modelled with no agency fee on SRN spend, because they typically run it in-house.
What the numbers mean
Early-stage brands: an expertise problem
Early-stage brands lose about a fifth of their acquisition budget, mostly through the smaller non-SRN share. But their deeper issue is dependence. Without in-house expertise, founders can't easily judge whether their agency is optimising for the business or for spend. Their biggest risk isn't one bad channel. It's not knowing which channels are bad.
What to fix first: get independent measurement early, and insist on reporting against revenue and retention, not installs.
Growth-stage brands: the black box
Growth-stage brands leak the most per dollar, about 28% in this model, because half their budget sits in the least transparent part of the market. Their in-house teams usually run the large platforms well. The non-SRN half is a different story: huge in volume, resold through many layers and handed off because no team can deal with every network directly. That hand-off is where fees, untraceable spend and fraud accumulate.
What to fix first: reduce layers in the non-SRN chain, demand supply-path transparency, and test incrementality channel by channel.
Mature brands: efficient per dollar, large in total
Mature brands lose the smallest share, about 15% of acquisition spend, because they rely less on the riskiest channels. But they spend the most in absolute terms, so a smaller percentage can still mean large sums. Their budgets are often controlled by holding-company agencies, and direct-app deals lack benchmarks: brands can pay very different prices for similar inventory with little basis for comparison.
What to fix first: benchmark direct-app pricing, and link agency compensation to incremental outcomes rather than spend.
The common thread
Across all three stages the same pattern holds. Money leaks fastest where:
- The chain is longest. More layers mean more fees and more untraceable spend [1].
- Oversight is weakest. Affiliate and resold traffic carries far higher fraud risk [2].
- Partners are paid on volume. When fees are a percentage of spend, nobody in the chain is rewarded for efficiency.
Limitations
- The channel splits are operator estimates, not published market data. Your own mix may be quite different.
- The leak rates are averages. Brands with strong direct contracts and strict supply paths can do much better.
- The model counts fees, untraceable spend and fraud only. It doesn't include softer waste such as poor targeting or creative fatigue, so real losses may be higher.
The takeaway
Every stage of brand growth leaks, but in different ways. Early brands leak through dependence, growth brands through the black box of non-SRN spend, and mature brands through unbenchmarked deals and agency-held budgets. Knowing which stage you're in tells you where to look first. For most fast-growing brands, it's the half of the budget nobody on the team can fully see.
References
[1] ISBA / PwC. Programmatic Supply Chain Transparency Study, executive summary. December 2020. https://www.isba.org.uk/system/files/media/documents/2020-12/executive-summary-programmatic-supply-chain-transparency-study.pdf
[2] AppsFlyer. State of Fraud for Marketers, 2026 edition. April 2026. https://www.appsflyer.com/resources/reports/state-fraud-marketers/