WeWork was a real estate business.
By the most standard read of its operations (leasing long-term commercial space, subdividing it, reletting on shorter terms, capturing the spread), it was a real estate arbitrage with hospitality services attached.
Real estate businesses, in 2018, traded at 6-10x EBITDA.
WeWork raised at valuations implying 20-30x revenue.
The gap between those two numbers is not a story about deception. It is a story about how multiples are set, and what positioning does to the input variables that drive them.
What positioning actually does to a multiple
A multiple is not a measurement. It is a category-conditional output.
The same financial profile produces different multiples depending on which comparable set the market applies to it. A 30% gross margin business benchmarked against logistics trades differently than the same business benchmarked against software, because the implied terminal growth rate, capital intensity, and recurring-revenue assumption are different in each frame.
Category assignment is therefore not a downstream consequence of the business model. It is an upstream input to the valuation.
Whoever defines the category first (the company, the analyst, the comp set) defines the multiplier that gets applied.
The illegitimacy discount
This is not a marketing observation. It is a structural one, and it has been documented in the organisational sociology literature for over two decades.
Zuckerman's work on what he called the illegitimacy discount showed that firms whose category membership is unclear to the analysts who cover them get systematically discounted by the market. Not because their financials are worse, but because the audience cannot place them.
Hsu and her colleagues extended this into a more general finding: organisations that span categories pay a penalty in audience evaluation, because category-spanning creates interpretive friction.
The corollary, and the part with operational consequences, is that the inverse also holds. Firms whose category assignment is clearly favourable benefit from the multiplier attached to that category, whether or not their underlying operations match the category in mechanism.
Why the arbitrage held for nine years
The discipline that made WeWork's positioning credible was not the language. It was the operating choices that made the language defensible long enough to compound capital at the elevated multiple.
The pitch
The pitch was never "we lease offices." The pitch was: "we are a technology platform for the future of work, with network effects across a global member base."
Every element of that sentence was constructed to move the comp set:
- Technology (not real estate)
- Platform (not service)
- Network effects (justifies escalating multiple)
- Future of work (justifies long-duration TAM)
- Global member base (justifies recurring-revenue treatment of underlying lease arbitrage)
The category-signal moves
Branded design language across every location. Mobile app as the primary operating surface. Member events, member directory, member pricing tiers. Acquisitions of software companies (Meetup, Flatiron School) that had no obvious operating fit but added category signal. A revenue line item called "membership" rather than "rent." A founder who spoke in technology vernacular and was photographed at technology conferences.
None of this changed the underlying economics. All of it changed the comparable set the market reached for. And the comparable set was the multiplier.
Why VCs underwrote it and public markets didn't
There is a further mechanism here that explains why the arbitrage was funded so generously by venture capital but eventually rejected by public markets.
Pontikes, in her work on ambiguous classification, drew a distinction between two types of audience.
Market-takers use categories to identify and evaluate goods. They find ambiguous category membership confusing and unattractive.
Market-makers are interested in defining new market structures. They find ambiguity appealing: it signals a firm that might constitute a new category rather than fit an existing one.
Pontikes' empirical work on software firms showed that venture capitalists rewarded category ambiguity while consumers penalised it. The same asymmetry applies to private versus public capital.
Private investors operating on a market-maker logic underwrote WeWork's category ambiguity for years. Public market analysts, operating on a market-taker logic, did not.
For nine years, the arbitrage funded the next round of capital, which funded the next round of expansion, which extended the runway during which the category had to remain defensible.
The model was structurally fragile. It depended on continuous capital and continuous category coherence. But while it held, it raised more capital at higher multiples than any pure real estate business of equivalent scale could have.
What broke it, and what didn't
The S-1 process broke it. Public market diligence forced a category re-evaluation that private capital had not forced for a decade.
Once the comp set shifted from technology platforms to commercial real estate, the multiplier collapsed and the model could not survive at the new valuation.
What did not break it was the mechanism. Category arbitrage as a structural feature of how multiples are set is alive and well, in every deal where the company's self-positioning sits on the boundary between two categories with materially different multiples:
- SaaS-flavoured services businesses.
- Platform-flavoured marketplaces.
- AI-flavoured workflow tools.
The logic is identical. The discipline that determines whether the arbitrage holds or breaks is whether the operating reality eventually catches up to the category language, or whether the gap widens until external diligence forces the re-rating.
What this means for PE operations
Two implications, both load-bearing.
First: in deal evaluation
The category the target is presenting itself in is an active variable, not a label.
The diligence question is not "is the company in this category."
It is: does the operating reality support the category, or is the category running ahead of the operations.
A company whose category language is two steps ahead of its operating model is being valued on a multiple it cannot defend through the next external diligence cycle.
Second: in portco operation
Positioning is a multiples lever, not a marketing exercise.
The portco's category language at exit determines the comp set the buyer reaches for, which determines the multiple the deal closes at.
Pontikes' later work on category strategy frames this directly: categories are not external constraints to which firms adapt. They are also instruments that firms can shape to align the market with the firm.
A portco that has spent four years operating at the boundary of two categories, and credibly executing into the higher-multiple one, sells differently than one whose category has remained where it was at entry.
The work is not in the language. The work is in the operating choices that make the language defensible long enough to matter at exit.
Further reading
- Hsu, G. (2006). Jacks of all trades and masters of none: Audiences' reactions to spanning genres in feature film production. Administrative Science Quarterly, 51(3), 420–450.
- Hsu, G., Hannan, M. T., & Koçak, Ö. (2009). Multiple category memberships in markets: An integrative theory and two empirical tests. American Sociological Review, 74(1), 150–169.
- Pontikes, E. G. (2012). Two sides of the same coin: How ambiguous classification affects multiple audiences' evaluations. Administrative Science Quarterly, 57(1), 81–118.
- Pontikes, E. G. (2018). Category strategy for firm advantage. Strategy Science, 3(4), 620–631.
- Zuckerman, E. W. (1999). The categorical imperative: Securities analysts and the illegitimacy discount. American Journal of Sociology, 104(5), 1398–1438.