The Platform Always Takes Its Cut; Social Distribution and the Slow Erosion of Brand Value.
Brands have spent fifteen years building recognition on infrastructure they don't own. The algorithm was never neutral. And the debt is becoming due.
Brands have spent fifteen years building recognition on infrastructure they don't own. The algorithm was never neutral. And the debt is becoming due.
In the spring of 2011, a former construction engineer called Peter Dering launched a Kickstarter campaign for a camera clip. He raised $364,000 from 5,258 backers in a matter of days. He had built something people wanted — and he had built it with them. A community that had pre-funded the idea, followed its development, and received it as participants rather than consumers.
That distinction, between a community and an audience, is the argument this piece is making. It is not an easy argument to make, because it runs counter to a decade of consensus about how brands are built. That consensus holds that distribution is reach, and reach is recognition, and recognition is brand. With the rise of content creators and personal brands, we're seeing a clear narrative around this distinction, and the platforms on which most brands now operate have an obvious commercial interest in that story being true.
It is not true. And the cost of believing it is beginning to show up in enterprise value.
The Grand Bargain
When Facebook, then Instagram, then TikTok extended the promise of mass reach to any brand willing to post consistently and engage authentically, the economics looked extraordinary. Free distribution. A direct line to consumers. The removal of the traditional gatekeeper — the retailer, the media buyer, the print run — in favour of an algorithm that was, at least in principle, neutral.
Brands accepted these terms without much scrutiny. The early numbers were remarkable. Pages accumulated millions of followers. A single post could reach an audience that would have cost hundreds of thousands of pounds to address through conventional media. The economics appeared structurally transformative.
What was not scrutinised carefully enough was the nature of the bargain. Platforms offered reach. They never offered ownership. Reach can be rented. Recognition has to be owned. And there is a difference — structural, compounding, and now financially material — between the two.
What the Algorithm Actually Is
The algorithm is routinely described as a neutral signal-processor: a system that surfaces what people want to see based on what they have previously engaged with. This framing is convenient for platforms and has been largely unchallenged by the brands that depend on them.
It is also wrong, in the way that matters most.
An algorithm is a revenue mechanism. It is calibrated, at its core, to convert attention into advertising yield. Every design decision — what content is surfaced, how frequently, to whom, in what format — is made in the service of that objective. User experience is a means to that end. Brand building is a side effect, not a purpose.
This does not make the algorithm malicious. It makes it indifferent. And indifference, at scale, is structurally more dangerous than hostility. A hostile counterparty can be argued with. An indifferent one simply changes the rules when its commercial interests require it.
The numbers are unambiguous. Instagram's average post reached roughly 12% of a page's followers in 2019. By 2024, that figure had collapsed to 4.0% — an 18% year-on-year decline, confirmed by Hootsuite's industry benchmarking.1 Facebook now delivers organic reach of approximately 1.37% of page followers, with a median engagement rate of 0.2%.2 Some reports place it lower still.
Sources: Hootsuite Social Media Trends 2025; Social Status / Social Media Examiner; Statista via Martech Zone.1,2,3
"Reach is rented. Recognition is owned. The platform has always known the difference — even when you didn't."
The counterpart to this collapse is equally unambiguous. Global social media advertising spend reached $219.8 billion in 2024 — nearly 30% of all digital advertising expenditure worldwide.3 That is the aggregate cost of brands paying to reach the audiences they spent years building. It is not a marketing efficiency story. It is a structural transfer of value from brands to platforms, delivered gradually enough that it has never provoked the reaction it deserves.
The Structural Confusion
The error most brands have made — and it is an error that Entry 001 of this publication was written to address — is conflating reach with recognition. One is rented. The other is owned. They are not the same asset.
Entry 001 defined brand as the market's recognition of a pattern. Not what a business claims to be, but what it repeatedly demonstrates itself to be. Owned recognition is the product of consistent behaviour, compounded over time. It is what allows a business to be chosen rather than merely noticed. It supports pricing power, commands loyalty, and — critically — it is what acquirers are valuing when they pay a premium for a business rather than its assets.
Rented reach is something different. It is visibility at a moment in time, on terms set by someone else. It can be purchased, boosted, or borrowed. It does not compound. When the conditions that produce it change — when the algorithm shifts, when the platform's commercial priorities evolve, when a new fee structure is introduced — reach diminishes or disappears. The recognition it appeared to be building does not transfer with it.
Brands have been building rented reach and calling it owned recognition. The algorithm has been happy to let them.
This confusion has a specific financial consequence. Madden, Fehle, and Fournier demonstrated in 2006, using the Fama-French methodology standard in finance, that firms with strong brands outperformed market benchmarks and did so with lower risk.4 Bahadir, Bharadwaj, and Srivastava confirmed in 2008 — analysing 133 actual M&A transactions through SEC filings — that brand equity accounts for a significant and measurable share of acquisition transaction value.5 Brand equity as enterprise value is not an assertion. It is an empirically demonstrated finding.
Acquirers price platform dependency accordingly. Market data from 2024–2026 shows that e-commerce businesses with over 70% of revenue from a single platform receive a material valuation discount.6
| Channel profile | EBITDA multiple | What acquirers are pricing |
|---|---|---|
| 90%+ single platform (e.g. Amazon)Platform controls demand, pricing, and customer relationship simultaneously | 3.0 – 3.5× | Single point of failure — ranking change or account suspension can destroy the business |
| 50–70% platform, some diversificationSome owned channels; demonstrated ability to build beyond one platform | 4.5 – 5.5× | Lower concentration risk; growth levers beginning to appear |
| Balanced omnichannel + owned DTCMultiple acquisition channels; owned customer relationships and data | 5.5 – 6.5× | Defensible — owned recognition held in customer relationships, not platform accounts |
Sources: ClearlyAcquired, E-commerce Valuation: Current EBITDA Multiples 2024–2026; Canopy Management, Why Amazon-Dependent Brands Sell for 40% Less (2025).6 Note: multiples vary by deal size, sector, and buyer; ranges reflect market central tendency.
On a business generating £2 million in EBITDA, the spread between a 3.5× and 6.5× multiple is a £6 million difference in enterprise value. That is not a rounding error. It is the financial cost of rented reach mistaken for owned recognition.
A Tale of Two Slings: Peak Design and the Anatomy of Owned Recognition
How a San Francisco bag company survived Amazon copying its top-selling product — not because of the viral video, but because of what the video was able to invoke.
In October 2020, Amazon's private label division released a product called the Amazon Basics Everyday Sling, priced at $32.99. The shape, the aesthetic, the pocket configuration, and the placement of the logo patch were strikingly similar to the Peak Design Everyday Sling — first launched in 2017, retailing at $99.95.7
Peak Design's CEO, Peter Dering, described the moment plainly: "somewhere in Amazon there's a bell that goes off, which says 'OK, this one's going to be worth our time to go make the knockoff.'" Peak Design had been the top-selling premium camera bag on the platform by a significant margin. That data made them a target.8
The company considered legal action. It decided on something more interesting: a 90-second video called A Tale of Two Slings, in which Dering narrated the differences between his product and Amazon's. Years of R&D. Recycled, BlueSign-approved materials. A lifetime warranty. Fairly paid factory workers. Carbon neutrality. The video's dramatisation of Amazon's product development — a goggle-eyed executive telling a subordinate to "keep combing that data" and "Basic this bad boy" — achieved viral reach without requiring legal standing.
The Conventional Reading
The conventional reading of this story is that Peak Design won the PR battle. That is not the interesting reading.
The interesting reading is structural. Peak Design survived the encounter with Amazon not because of the video, but because of what the video was able to invoke: a community of backers, customers, and advocates whose relationship with the brand predated and outlasted any platform. The company had raised over $60 million across fourteen Kickstarter campaigns by 2025, building a customer base that had, in many cases, funded products before they existed.9
Sources: Shopify/Dering (2025); ECDB; Kickstarter (2025).9,10,11
Its average customer owns more than seven products — a figure Dering cites as the measure of what the company has actually built.10 Not transactions. A relationship that compounds.
What Amazon Could and Could Not Copy
Analysis based on CNBC reporting (2021) and Shopify/Dering interview (2025).7,9
Peak Design's $99 Everyday Sling continues to command its price in a category where a functionally similar substitute exists for $33. That premium — roughly 3× the alternative — is not explained by materials or features alone. It is the financial expression of fifteen years of owned recognition. Brand equity denominated in dollars at the point of consumer decision.
And none of it lives on Amazon's servers.
The Invisible Fee
The transfer of value from brands to platforms does not arrive as a single event. It arrives as a series of adjustments, each individually manageable, collectively corrosive.
It begins with declining organic reach, managed by increasing posting frequency and content investment. It continues with paid promotion — initially modest, increasingly non-optional. It proceeds through algorithm updates that deprecate previously high-performing formats: the static post, the long-form caption, the link in bio. It arrives, finally, at the point where a brand is paying to reach an audience it built, in a format the platform has chosen, through a system the platform can alter without consultation.
By 2022, Marketplace Pulse calculated that Amazon was taking more than 50 cents of every dollar generated by third-party sellers on its platform, when referral fees (6–45% depending on category), FBA fulfilment, storage, and advertising were aggregated.12 That figure has since been substantiated by Fortune's analysis of Amazon's own SEC filings: seller fees grew from 22% to over 29% of Amazon's non-AWS revenue between 2019 and 2024, a rate of growth that outpaced the increase in the share of goods sold by third parties.13
"A brand does not receive an invoice for the reach it has lost. The cost is invisible in the way that matters most: it does not appear as a line item, so it is rarely interrogated as a strategic problem."
These are not hidden costs. They are structural conditions. The platform's commercial interests and the brand's commercial interests are aligned at the point of distribution and diverge at the point of value capture. Brands have spent fifteen years mistaking the former for evidence of the latter.
What Brand Built on Owned Ground Looks Like
The distinction Entry 001 drew — between brand as expression and brand as pattern — maps precisely onto the distinction between rented reach and owned recognition. A brand built on algorithmic reach is rented. Optimised for the conditions the platform currently values: the right format, the right cadence, the right engagement signals. When those conditions change, the brand must change with them or lose visibility. A business that must continually convince rather than one that is simply chosen.
What made Peak Design's business defensible was not its marketing. It was the conditions beneath it. A viable model: investor-free, profitable from its first campaign, margins maintained because the business does not depend on platform economics to generate demand.14 Behavioural alignment: a B-Corp certification, a lifetime warranty, a manufacturing code of conduct, and a founder who has not encountered a question about his business he will not answer fully.15 Constraint: Peak Design makes carry solutions. Continuity: fifteen years of consistent identity, quality, and community — none of it contingent on an algorithm's preferences.
These are the preconditions Entry 001 described. And the preconditions of owned recognition — enterprise value that does not evaporate when a platform changes its rules.
The brands beginning to rebuild around this logic share a characteristic: they are measuring things the algorithm does not reward. Customer lifetime value rather than follower count. Direct channel revenue as a proportion of total revenue. The cost of acquiring a customer versus the cost of retaining one. These are the metrics of owned recognition — a business whose value is held in relationships, not in rented visibility.
The Strategic Implication for Leaders
For founders, the platform dependency question is fundamentally a question about what you are building. If the primary beneficiary of your brand-building activity is the platform — if what you are accumulating is rented reach rather than owned recognition — then you are investing in someone else's asset, on terms someone else controls.
The question is not which platform to prioritise. It is what you are building that would survive the platform disappearing tomorrow. Peak Design's $53 million DTC channel, its community of Kickstarter backers, its seven-plus products per average customer: none of these belong to Amazon. They are owned recognition. They belong to the brand.11
For CMOs, the risk is different. Brand equity that resides primarily in social following and organic reach is rented. The counterparty is large, commercially sophisticated, and under no obligation to maintain the conditions that made your following valuable. When a platform shifts its algorithm, it is not doing something to you. It is adjusting its system in its own interest, as it has always been entitled to do.
The strategic implication is not to abandon platforms. Peak Design sells on Amazon. The implication is to use platforms as distribution without mistaking them for recognition. When Amazon copied Peak Design's bag, it could replicate everything the platform could measure: the product specification, the category positioning, the price anchor. It could not replicate what the platform could not measure: fifteen years of consistent behaviour that had made a portion of the market unwilling to accept a substitute.
"That portion of the market is Peak Design's enterprise value. It was never Amazon's to take."
The war economy narrative, explored in Entry 002 of this publication, showed how brands with genuine story can command margin that mere cost arithmetic cannot explain. The platform economy reveals the same truth from the other direction: brands without genuine recognition — brands whose equity lives in follower counts and organic reach rather than in consistent, compounded behaviour — are holding an asset they do not own, on terms they did not set, subject to revision without notice.
The platform gave you reach. Rented, on their terms, subject to revision without notice.
It was never going to give you recognition. That has to be built — in behaviour, in consistency, in the relationship between what you promise and what you repeatedly deliver.
One compounds. One doesn't. Only one of them is yours.
Drawing on Madden, T.J., Fehle, F. & Fournier, S. (2006), Brands matter: An empirical demonstration of the creation of shareholder value through branding, Journal of the Academy of Marketing Science, 34(2), 224–235; and Bahadir, S.C., Bharadwaj, S.G. & Srivastava, R.K. (2008), Financial value of brands in mergers and acquisitions: Is value in the eye of the beholder?, Journal of Marketing, 72(6), 49–64. The empirical case that brand equity is enterprise value — not marketing overhead — is developed across both works.
