First, the Data Correction
Luma Partners published a report this week with a headline that runs counter to what we’ve been saying for the last two episodes: deal activity in digital media and marketing technology has lost momentum, weighed down by geopolitical tensions, declining year over year.
We read the actual report. The headline and the charts don’t quite speak the same language.
Here’s what the data actually shows:
Ad tech sub-$100M deals: Q4 2024 — 19 deals. Q4 2025 — 22 deals. That’s up. What’s down is $100M+ deals — 6 in Q4 2024, 4 in Q4 2025, 1 in Q1 2026. The big deals are getting harder. The lower middle market is holding.
Martech sub-$100M deals: 42 in Q4 2024, 40 in Q4 2025, 41 in Q1 2026. That’s flat. Not a collapse.
Digital content sub-$100M deals: 36 → 34 → 26. This one is genuinely declining — and the reason isn’t macro. It’s AI. If you’re a sub-$100M digital content business, the first question any buyer asks right now is whether you’re disruptible by AI. A lot of them are. That’s showing up in the deal count.
The category that’s missing from Luma’s data: sub-$50M. Both of us are busiest in the $20-50M EV range and it is moving. There has to be a data set that captures what’s happening at that level, because from where we sit, the wild west of lower middle market deals is very much alive.
The real story: large deals are harder. Small deals are fine. Digital content is structurally challenged. Everything else is roughly holding.
Deal #1: Amex + Hyper (HyperCard)
On April 16th, American Express announced an agreement to acquire Hyper — an agentic AI expense management startup founded in 2022, backed by Sam Altman, former MasterCard CEO Gene Lockhart, Netflix co-founder Mark Randolph, and One Medical founder Tom Lee.
Financial terms weren’t disclosed. This was almost certainly a sub-$20M deal. The framing here isn’t valuation. It’s sequencing.
This was a partner-first deal two years in the making. Hyper and Amex announced a co-branded card product with AI agents embedded in Amex’s platform back in late 2024. Amex has been running Hyper’s technology in production through a live card product since then. They knew the team. They knew the product. There was no competitive auction.
The strategic logic is clean: in March 2025, Amex acquired Center — an expense management software company, rumored at around $600M — which gave them the expense management workflow infrastructure. Hyper gives them the AI agent layer that sits on top of it. Together, they’re building a direct competitive platform aimed at Concur, Ramp, and whatever Brex and Capital One are assembling.
Amex did $72B in revenue in 2025, overwhelmingly from the card. What they’re building now is an adjacent software business on top of the card franchise — card plus expense platform plus agentic AI for corporate spend. The middle layer of expense management bloat — the manual approvals, the reconciliation friction, the paper receipts — is exactly what Amex has been quietly building toward eliminating since their 2019 acquisition that became AmericanExpress 1AP, through Nipendo in 2023, through Centur, and now through Hyper.
Christian’s read: this was an acqui-hire. Great team, real product, get them inside the building and go fix corporate expense management. Clean out the cap table, call it a day.
Deal #2: Viant + TVision Insights ($40M)
On April 15th, Viant Technology announced a definitive agreement to acquire TVision Insights for $40M — $22.5M in cash and $17.5M in Class A common stock, closing Q2 2026.
The headline number matters less than the sequence.
This is the third leg of a deliberate, multi-year build:
2024 — Iris TV: Content layer. CTV inventory classification at the content level — genre, emotion, brand safety — not just at the app or show level. Still broadly licensed and platform-agnostic.
March 2025 — Locker: Identity layer. Publisher first-party data, data activation, alternative ID management.
April 2026 — TVision: Attention layer. Panel-based, second-by-second eyes-on-screen measurement for linear TV, CTV, and walled gardens. Are viewers actually watching? How many? For how long?
Viant CEO Tim Vanderhook called it the trifecta: identity, context, attention — all feeding into their intelligence layer. He told Ad Exchanger that the TVision deal simply wouldn’t have been possible without the Locker acquisition. This wasn’t opportunistic. It was sequenced.
The detail that didn’t make the press release but matters: Maverick analyst Maria Ripps asked Vanderhook directly whether TVision’s data would remain available to other measurement providers. The answer was no. Existing contracts get honored, but as they expire, the data goes exclusive to Viant’s platform. That’s a meaningfully different integration playbook than Iris TV, which stayed broadly licensed. Same acquirer, two different strategic postures depending on what the asset does for competitive differentiation.
The economics are clean. TVision did approximately $10M in 2025 revenue. $40M is exactly 4x — reasonable for a strategic capability acquisition in ad tech. Viant’s balance sheet at end of 2025: $191M in cash, zero long-term debt, $75M untapped credit facility. They spent $22.5M in cash — roughly 12% of their cash pile — in a quarter where Luma is reporting ad tech multiples compressed 21%. They’re deploying capital with conviction against their own roadmap regardless of what the index is doing.
One complication worth noting: TVision had raised $16M in VC at an $80M valuation. The $40M exit is a haircut on preference. There was almost certainly some back-room work done to make sure the founders and team got a reasonable outcome inside that structure. On a multiple of revenue it’s fine. On a multiple of last valuation, less so. But for a 12-year-old company with real revenue and a genuine strategic acquirer at the table, this is a good outcome in a difficult market.
Props to Eric Stearns, Viant’s Head of Corp Dev, who joined from RBC Capital Markets in August 2025. This appears to be his first deal in seat. First one through the door and it’s a clean, sequenced strategic buy with solid economics. Good start.
Next week: Christian may or may not spill what he heard at the JEGI-Leonis conference in New York (Chatham House rules, technically). Ayelet just spoke at the Own It Women’s Agency Ownership Summit in front of 400 people and has a lot to say.
Both will be at Possible in Miami. Come find us.
Subscribe to In/Organic for weekly M&A coverage. Deal Review Fridays live every week on LinkedIn and YouTube.
