I'm calling it something else. The pricing wall just went up, and the SMB market (the one that built this category) got quietly abandoned on the way to Fortune 500 contracts.
Here's the part the press releases don't mention. Sierra's customer page reads like a board meeting at Davos: Rocket Mortgage, Sonos, ADT, SiriusXM, WeightWatchers, SoFi, Chime, Brex, and Ramp (Sierra, 2026). Forty percent of the Fortune 50 are now Sierra customers (TechCrunch, 2026). $150M ARR in eight quarters. That's a beautiful enterprise software story.
It's also a warning. Nobody raises $950M to serve a 12-person Shopify store. They raise it to chase $500K-minimum contracts at companies with 10,000 seats. And once a category leader plants its flag in enterprise, every other vendor (Decagon, Fin, Breeze) quietly follows. So who's left for the rest of us?
TL;DR: Sierra's $950M Series E at $15.8B (TechCrunch, 2026) cements its enterprise focus, with 40% of the Fortune 50 already on the customer list. The whole AI support category (Sierra, Decagon, Intercom Fin, HubSpot Breeze) is racing upmarket. SMBs (89% of US employer firms have under 20 people, per US Census SUSB) are being left without a home.
What did Sierra actually build?
Sierra sells outcome-priced AI agents to large enterprises, with no public pricing and a "talk to sales" gate (Sierra, 2026). Bret Taylor, co-founder, ex-Salesforce co-CEO, and ex-OpenAI chair, pitches it as the end of clicking buttons. He estimates global customer service spend at $400B per year (TechCrunch, 2026).
The product is good. I've watched the demos. The agent platform handles voice, chat, and complex workflows that route into Salesforce and ServiceNow. Sierra charges per resolved outcome, not per seat. That's the bet: pay only when the AI actually fixes something.
The customer list tells the whole story
Look at who's signing the contracts. Rocket Mortgage processed half a trillion in loans last year. ADT has 6 million customers. WeightWatchers and SiriusXM are public companies. Sonos sells $1.5B+ in audio hardware annually. The fintech roster — SoFi, Chime, Brex, Ramp — handles billions in processed volume. The D2C side runs eight-figure revenue books.
I pulled every named Sierra customer from their website on May 5, 2026. Median company size: ~25,000 employees. Smallest named customer: ~3,000 employees. There is not a single SMB on that list. Not one.
Sierra closed a $950M Series E at a $15.8B valuation on May 4, 2026, led by Tiger Global and GV (TechCrunch, 2026). The company reports $150M ARR in eight quarters and counts 40% of the Fortune 50 as customers.
Why $950M means SMBs got abandoned
Venture math is unforgiving. To return a $950M round at a $15.8B post, Sierra needs to credibly grow into a $50B+ outcome, which means winning Salesforce-sized deals at Salesforce-sized contract values (TechCrunch, 2026). You don't get there one $99/month customer at a time.
This is the part founders miss when they cheer competitor raises. A mega-round isn't a rising tide. It's a commitment device. Tiger Global didn't write $300M to watch Sierra serve dental practices.
The whole category is moving the same direction
Sierra isn't alone. The entire AI support category is being pulled upmarket by the same gravity:
- Decagon raised $131M Series C in 2026, anchored at mid-market and enterprise.
- Intercom Fin prices at $0.99 per resolution and aggressively targets mid-market accounts.
- HubSpot Breeze moved to $0.50 per resolution in April 2026 (HubSpot, 2026), which validates the outcome model but also signals where HubSpot's ICP now lives.
- Zendesk's AI add-on lands at $1.50-2.00 per resolution depending on tier.
Every category-defining vendor is now optimized for buyers with procurement teams. The pricing pages are gone. The "starting at" numbers are gone. The self-serve trials are getting nerfed in favor of "book a demo." This is what the late innings of a category look like, and SMBs are not invited.
Why per-resolution pricing breaks for small teams
Who actually serves teams under 50 people?
If Sierra is enterprise, Decagon is mid-market+, and Fin and Breeze are mid-market, the under-50-employee segment has roughly one and a half real options left, neither built for it. Eighty-nine percent of US employer firms have fewer than 20 employees (US Census SUSB, 2024). That's not a niche. That's the market.
Let me lay out the actual landscape with numbers.
The pricing gap, side by side
| Vendor | Pricing Model | Effective monthly cost (1,000 conversations) | ICP |
|---|---|---|---|
| Sierra | Outcome, talk-to-sales | $5,000+ (estimated, no public pricing) | Fortune 500 |
| Decagon | Outcome, custom | $3,000+ | Mid-market+ |
| Intercom Fin | $0.99/resolution + seats | ~$1,500+ | Mid-market |
| HubSpot Breeze | $0.50/resolution + Hub | ~$1,000+ | Mid-market |
| Zendesk AI | $1.50-2.00/resolution + seats | ~$2,500+ | Mid-market+ |
| Corebee | $99 flat | $99 | SMB / under 50 |
The numbers above are conservative. They assume you can get on these platforms at all. Sierra and Decagon will not return your call if you have 12 employees. Intercom and HubSpot will, but the contract starts above $1K/month before AI usage.
Why "outcome pricing" hides the real problem
Outcome pricing sounds fair. You only pay when the AI actually resolves something. Gartner predicts 40% of SaaS spend shifts to outcome-based models by 2030 (Gartner, 2026). On paper, this aligns vendor and buyer.
In practice, the per-resolution number is a lever the vendor controls and you don't. What counts as a "resolution"? The vendor decides. When prices change, you find out at renewal. And the lower-bound math always assumes you stay small.
The real math on Intercom Fin's $0.99
What outcome pricing does to a growing team
Here's the part nobody talks about. We've onboarded over 200 SMB customers at Corebee, and the pattern is identical: their support volume is bursty. Black Friday triples their tickets. A product launch quadruples them. A bug in production sends a 10x spike for 48 hours.
On flat pricing, none of that matters. Your bill is your bill. On per-resolution pricing, your worst week becomes your most expensive week. The day your business breaks out is the day your software bill starts to bleed.
A real SMB cost model
Imagine a 12-person Shopify store doing 800 conversations a month. AI handles 70%, so ~560 resolutions.
- On Intercom Fin at $0.99: $554/month in AI fees, plus seats, call it $800-1,000 all in.
- On HubSpot Breeze at $0.50: $280/month in AI fees, plus Hub, $500-700 all in.
- Black Friday week: 3x volume = ~1,680 resolutions that month. Fin: $1,663. Breeze: $840.
- A bad month with a bug: 5x volume. Fin: $2,772. Breeze: $1,400.
For comparison, flat pricing at $99 stays at $99. The shape of your bill matches the shape of your business: predictable.
This is why outcome pricing is built for enterprises with procurement teams who model worst-case scenarios. It's a terrible fit for an 8-person team that just wants to know what they'll spend next month.
The opportunity nobody is funding
Here's the contrarian take. Sierra's raise didn't validate AI customer support for SMBs. It abandoned them. The category leaders are now financially incentivized to ignore anyone who can't sign a $50K contract.
That gap (89% of US employer firms per US Census SUSB 2024, every Shopify store, every two-person SaaS, every local services business) is the largest unserved customer support market in software. And the people building for it are getting smaller, not larger, as the big vendors race upmarket.
The pattern is familiar if you've watched a SaaS category mature. Every winning vertical started by serving who the incumbents ignored.
What history says about abandoned segments
When Salesforce focused on the enterprise, HubSpot built a $30B company by serving the SMBs Salesforce wouldn't return calls from. When Workday focused on F500 HR, Gusto built a $10B+ company on small-business payroll. When Oracle ignored the dev tooling market, Atlassian built a $40B company.
The script is consistent. Category leader raises a mega-round. Mega-round forces upmarket motion. Upmarket motion creates an obvious gap below. A scrappy team builds for the gap. Ten years later, the gap company is bigger than anyone predicted.
Sierra raising $950M is the cleanest signal yet that we're at step two of that cycle in AI customer support. Decagon, Fin, and Breeze are reinforcing it.
HubSpot moved Breeze to $0.50 per resolution on April 14, 2026 (HubSpot, 2026), validating outcome pricing while raising the floor. Combined with Intercom Fin at $0.99 and Zendesk AI at $1.50-2.00, the entire category is now structured around per-resolution economics that punish unpredictable SMB volume.
So what should an SMB founder do?
Don't buy the vendor that's most loved by Fortune 500 procurement. Buy the vendor whose ICP includes you.
Corebee is built for the rest of us. $99 flat. No per-resolution surprises. No "talk to sales" wall. We exist because the founder team has run small businesses, and the math on per-resolution pricing never made sense at our scale. If you're a team under 50 people doing real customer support volume, this is the math that fits your business.
The goal is not to compete with Sierra. Sierra is winning Rocket Mortgage and SoFi. Good for them. The goal is to be the obvious answer for the 89% of companies Sierra will never call back.