Olo is not a SaaS company anymore. Here's the multiple the math actually works at.

It’s a Friday morning, the kind where the espresso machine in the corner is louder than the trading desks, and I’m three browser tabs deep into the Olo App announcement that hit on Tuesday. My desk has the usual artifacts: a half-eaten croissant from the place downstairs that still hasn’t figured out QR ordering, two printed S-1s, and the Beyond4 Phoenix agenda that someone left here on purpose. I am supposed to be writing about a hotel REIT spin. Instead I keep going back to one paragraph in the Restaurant Business piece, the one where Olo describes a single consumer app that lets diners order across hundreds of its brand customers.

I read it again. Then I open the take-private deck from last summer, the one Thoma Bravo’s team walked LPs through. And I write a one-line note to myself in the margin of a printout: Olo is not a SaaS company anymore. Then, because I’m me and the second espresso has landed, I write a second line: and that is the entire point of the take-private.

That’s the thesis. Let me show you the math.

What actually got announced

The short version: on March 3, Olo announced a consumer-facing ordering app — a single, branded surface where a guest can place an order at any restaurant in Olo’s network. The FSR Magazine writeup frames it as a logical extension of Borderless, Olo’s 40-million-account guest-identity layer. The narrative restaurants are being sold internally is: you keep the brand, you keep the data, you keep the margin, and Olo gives you reach you didn’t have before. The narrative diners get is simpler: one app, every restaurant you actually like, no DoorDash markup.

Olo today sits on a ~750-brand network on the enterprise side. That’s the install base. The 40M Borderless users are the demand-side asset that, until now, was monetized only indirectly — as a conversion-rate uplift Olo’s brand customers got “for free” inside their own apps. The App turns that 40M into a directly-addressable consumer audience.

This is not a SaaS feature. This is a marketplace.

I’ll come back to that.

The Thoma Bravo entry: what ~6x sales actually bought

The take-private closed at roughly $2B against trailing revenue that the public filings put in the high-$200M range. Call it ~6x sales — generous for a category that the public markets had been rerating down toward 4x for two years, and consistent with what Holden Spaht has been telling LPs about software being mispriced versus AI optionality. (His augment.market interview is the cleanest articulation of the thesis: the market is wrong about durable software, and the firm intends to be the buyer of record while it figures that out.)

Six times sales for a vertical SaaS asset with mid-teens growth and improving FCF margins is not “cheap.” But it isn’t the headline. The headline is what 6x sales gets you inside Thoma Bravo’s portfolio system: the ability to do balance-sheet things you cannot do as a public company without getting your multiple cratered the quarter you announce them. Spinning up a consumer app that depresses near-term EBITDA. Cross-pollinating with another portco. Building a marketplace inside what the sell-side models as a SaaS line item.

If you take Spaht at his word, the bet at entry was: pay a SaaS multiple for an asset that we are going to re-architect into something the public market would have re-rated lower in the short run and higher in the long run, and absorb the J-curve in private where nobody has to mark us to market every 90 days.

The App announcement is the first visible artifact of that re-architecture. Which means it is also the first thing that lets an outsider try to draw the sum-of-parts.

The valuation walk (and the part you should flag as interpretation)

Here is where I have to be honest about which numbers are reported and which are mine. The reported facts: ~$2B take-private, ~6x trailing sales, 750 brands, 40M Borderless accounts, App launch confirmed for 2026. The rest is interpretation, and I’m going to mark it as such.

Step one: size the marketplace opportunity.

Olo’s brand customers process digital orders in the tens of billions of dollars in GMV annually across the network. Olo doesn’t disclose the exact figure cleanly, but the order-volume math the company has hinted at in past investor materials, combined with what its brand mix would suggest, puts brand-network digital order GMV plausibly in the $25–35B/year range by 2026. (Interpretation. Use a range. Don’t anchor.)

Assume the Olo App captures 5% of that GMV by full-year 2027. Why 5%? Because that’s roughly the share a credible “second-place” consumer surface tends to reach inside a network that’s already actively pushing it through its own brand apps and emails — and Borderless gives Olo a warm-start audience that a cold-launch consumer app would not have. (Interpretation. The 5% is mine.)

5% of $25–35B = $1.25–1.75B of App GMV.

Step two: assume a take-rate.

I’m going to use 4%. Not because Olo has disclosed it — they haven’t — but because that’s the rough zone for a marketplace that is positioning itself as the anti-DoorDash: low take, brand-friendly, the data stays with the operator. DoorDash takes ~15%. A 4% number lets Olo tell the operator “we are a third of the cost of the aggregator and you keep the customer.” (Interpretation. The 4% is mine. If it’s actually 2%, halve the next number. If it’s 6%, add 50%.)

4% × $1.25–1.75B = $50–70M of marketplace revenue in year one of meaningful App scale. Push the capture rate to 7–8% by 2028 and you’re at $80–120M. That’s the number that matters.

Step three: apply multiples.

Public marketplaces — the asset-light, take-rate variety — trade at roughly 3x forward sales today. Vertical SaaS infrastructure in restaurants trades closer to 4–5x, with the better operators (Toast, Olo pre-take-private) historically getting 5–7x. Inside Thoma Bravo’s hold, you can argue Olo deserves the higher end of the SaaS band. Let’s hold the SaaS book at 6x to be charitable.

• SaaS book: ~$300M of SaaS revenue × 6x = $1.8B
• Marketplace book: ~$100M of marketplace revenue × 3x = $0.3B
• Blended enterprise value: ~$2.1B
• Blended multiple on ~$400M of total revenue: ~5.25x

That is not multiple expansion. That is multiple compression — from 6x at entry to ~5.25x blended on the way out, even though the absolute dollar value is higher.

This is the part the bull case usually skips. Adding a lower-multiple revenue stream to a higher-multiple business mechanically drags the blended number down. The only way the trade works for Thoma Bravo is if (a) absolute EV grows faster than the multiple compresses, or (b) the SaaS book itself gets re-rated higher because the App makes the SaaS contracts stickier, or (c) the marketplace eventually re-rates toward the SaaS multiple because the network effect is durable enough to deserve it.

I think (a) and (b) are real. I’m skeptical of (c) on a 3–5 year horizon. The public marketplace comp set has been remarkably stable at 3x for a long time.

Why the App had to ship anyway: the DoorDash question

You can argue the math above and still conclude Olo had to launch the App. I do.

The reason is what’s happening adjacent to them. DoorDash has spent two years building out a commerce stack that goes well beyond delivery — the white-label storefront product, the ads business, the SevenRooms acquisition at $1.2B that I’ll get into when that post runs, and the broader commerce-platform repositioning my colleague has been tracking. The strategic question for Olo became: if a guest’s primary digital touchpoint with restaurants ends up being a DoorDash-owned surface — even when it’s a brand’s own logo on the front — then Olo’s enterprise software gets reduced to plumbing behind someone else’s brand layer. Plumbing trades at 3–4x. Identity trades at 8x.

So the App is, at minimum, a defensive move to keep Olo’s seat at the table marked consumer surface that the operator controls. Whether it ever scales to the $100M marketplace revenue line is almost beside the point — even at half that, it changes the conversation about what Olo is.

And on the buy-side, that re-narration is what justifies holding through the J-curve.

The bet, and what would change my mind

Here’s the bet I’d actually make today, with appropriate position-sizing for an asset I cannot trade.

Base case (50%): App reaches 3–5% capture of brand-network GMV by end of 2027. Marketplace line scales to ~$60–80M at a 3–4% effective take-rate. SaaS book grows 12–15% organically. Blended business is a $500–550M revenue company by 2028 at a ~5x blended multiple = ~$2.5B EV. Thoma Bravo books a modest IRR, mostly from operating leverage and debt paydown, and exits via secondary or strategic rather than IPO.

Upside case (25%): The App becomes the second-most-used non-aggregator ordering surface in casual/fast-casual. 7–8% capture by 2028, marketplace at $100–120M, SaaS rerated higher because per-brand ARPU rises as the App drives demonstrable incremental order volume. $3.5–4B exit. This is the case Spaht is selling to LPs. (See the augment.market interview — the framing of “the market is wrong about software” only works if you can point to a thing inside the portfolio that the public market would mispriced this badly.)

Downside case (25%): The App fails to differentiate against the brand-own-app experience, Borderless doesn’t convert, and the SaaS book is now carrying the cost of a consumer product nobody wanted. The blended multiple compresses to 4x. Thoma Bravo eats the J-curve. $1.6–1.8B exit — a loss, before fees.

The thing I’d want to see in the next 18 months to move probability from base to upside: weekly active App users north of 8M, with at least 30% of orders being to a restaurant the user hadn’t ordered from on the operator’s native app in the prior 90 days. That second number is the only one that proves the App is additive to brand-network GMV rather than cannibalizing it.

If it’s cannibalizing — if the App is just relocating orders Olo was already going to process — then the marketplace revenue isn’t new revenue, it’s a tax on the SaaS book. That is the worst possible outcome and it’s the one the case against the AI-premium thesis is going to apply to a lot of these “platform repositioning” stories over the next year. My colleague’s piece on that is forthcoming and worth your time when it runs.

For now, the working line on my desk stays the same: Olo is not a SaaS company anymore. Whether that’s a good thing depends entirely on what happens inside Borderless between now and the back half of 2027. The 4% take-rate, the 5% capture rate, the 3x marketplace multiple — those are my numbers, not Olo’s, and you should price them accordingly.

But the direction of the math is harder to argue with than the levels. Adding a marketplace to a SaaS business compresses the blended multiple. That trade only works if the absolute number gets big enough fast enough. Thoma Bravo paid 6x to find out in private. The rest of us get to watch.

— Marcus edits The Bottom Line for TableTransfers. Tips: [email protected].