$1.36 billion, all cash. That is what Bending Spoons paid for Miro — a company that last raised at a $17.5 billion post-money valuation in January 2022. If both figures hold, the transaction prices a category leader at roughly a 92% discount to its final private round.
I have spent four cycles watching narrative premiums collapse into cash-flow reality, and this is the cleanest specimen I have seen outside crypto. The signal is not the buyer. It is the sellers — a syndicate of tier-one venture funds — choosing certain cash over an uncertain listing. That choice is macro, and it maps almost one-to-one onto what is quietly repricing across on-chain markets right now.
Start with the mechanism, not the headline. Miro is a visual collaboration platform — the infinite-canvas category leader. Founded in 2011, global by default, freemium by design. Its revenue model is seat-based subscription; its growth engine was the 2020-2021 remote-work surge. Both facts determine the outcome.
Bending Spoons is not a strategic acquirer in the classic sense. The Milan-based firm buys mature, cash-generating software, strips cost, raises price, and harvests the margin. It does not buy growth. It buys the residue of growth — the installed base that remains after the hype has been repriced.
That is the first structural fact: a category leader with a real product and real revenue was worth more to a cash-flow harvester than to the public market. The second structural fact explains why. Miro's moat was never deep. Digital whiteboarding is a feature, not a product. FigJam bundles into Figma. Microsoft Whiteboard bundles into M365. Atlassian embeds whiteboarding into Confluence. None of those competitors need the whiteboard to be profitable. They can give it away. This is the classic bundle-versus-point-solution war. The point solution loses not because it is worse, but because it is alone.
Because the identical war is being fought on-chain, and most participants have not named it.
Consider the Layer-2 landscape. The prevailing debate is OP Stack versus ZK Stack — optimistic versus zero-knowledge, fraud proofs versus validity proofs. That framing is a distraction. The real competition between OP Stack and ZK Stack is not cryptographic; it is distributional — which stack convinces more applications to deploy chains under its banner first. Base, built on the OP Stack, did not win because optimistic rollups are technically superior. It won because it was bundled into Coinbase's user surface. That is the FigJam dynamic, transplanted. The stack with the best technology loses to the stack with the best bundle.
Watch the liquidity. When I mapped stablecoin net flows across chains in the last two quarters, the pattern was not 'best technology attracts capital.' It was 'capital follows the surface where the user already holds an account.' The chain is a feature of the wallet. The wallet is a feature of the exchange. The exchange is a feature of the fiat rail. Each layer absorbs the one beneath it — precisely how Microsoft Whiteboard absorbed an entire standalone category without writing a line of novel collaboration code. This is the architecture of value hidden beneath the hype.
The second parallel is more uncomfortable. Miro's valuation collapsed because its growth narrative — seat expansion, net revenue retention above 120% — could not survive the return to normal demand. In crypto, the equivalent of seat-based expansion is emission-driven TVL, and it repriced on the same schedule. A protocol that pays users to deposit capital has not acquired a customer; it has rented one. The moment the emission curve steepens past the yield threshold, the rented capital leaves. That is not a liquidity metric. It is a retention metric wearing a liquidity costume.
I learned to distrust these numbers the hard way. In 2020, I built a Python tool to track capital efficiency across six major DeFi protocols during the Compound emission wars. The finding that stuck was not the 15% cross-protocol arbitrage. It was that the protocols with the highest headline TVL had the lowest organic retention. The dashboard said growth. The cohort data said churn. The same gap exists between Miro's 2021 headline valuation and its 2025 transaction price.
There is a third layer, and it concerns pricing itself. The interest rate models inside Aave and Compound are not price discovery; they are governance-selected step functions dressed as market equilibria. When utilization crosses a threshold, the borrow rate does not emerge from supply and demand. It jumps because a parameter was set in a forum vote. The curve is arbitrary, and the market prices it as if it were discovered. That is a Miro seat count with more decimals.
One more structural note. A mature asset with real cash flow and a shallow moat is not rescued by the growth market. It is acquired by the cash-flow market — Bending Spoons in software, consolidators in crypto. And the historical record on cross-chain infrastructure suggests crypto's consolidators are buying damaged goods. Cross-chain bridges have been drained for more than $2.5 billion cumulatively, yet interoperability still depends on them. That is not a bug in one bridge. It is a structural fact about a category that bundles risk while claiming to bundle liquidity — a harvest waiting to be recognized as such.
The consensus rebuttal is decoupling. Crypto natives argue that on-chain assets trade on their own liquidity cycle, insulated from the software repricing. I find this comforting and wrong. The mechanism is identical: a narrative premium built in a zero-rate, abundant-capital regime, repriced in a higher-rate, capital-scarce one, with the marginal buyer withdrawing first. Miro and a mid-cap altcoin are not different asset classes in this respect. They are the same trade, expressed in different ledgers.
The second rebuttal is AI. The market believes AI plus crypto is the escape hatch — the second curve that re-rates the whole sector. I would caution that AI is not a curve; it is a bundle. Whichever wallet, exchange, or model coordinator absorbs AI first will absorb the point solutions beneath it. The frontier model is the operating surface. Everything else is a feature. Predicting the pivot before the pivot is printed means recognizing that the escape hatch is, structurally, another version of the trap Miro walked into.
The transaction price is not the story. The direction of capital is. When a category leader sells for 8% of its last mark to a buyer that does not believe in growth, the market is stating what it thinks duration is worth. Crypto has not escaped that repricing. It has simply not yet printed its own Bending Spoons moment. Silence the noise, listen to the block height — and ask which of your positions is a point solution standing alone.