The AI Sales Tool You Can Own Is Already Public

A company called Futuri released a framework this week for how sales leaders should use AI. Automate the Grind, elevate the Genius. It sounds good. It probably is.

Futuri is a private company based in Austin. It's not on a stock exchange. You can't buy shares in it.

That's not the point though. The point is what companies like Futuri reveal about a pattern that is happening right now in public markets — and it matters a lot more if you own Salesforce.

Salesforce released earnings on Wednesday. Revenue hit $11.35 billion, up 11% from a year ago. The stock jumped 11% after the bell. The headline was AI.

The actual number that carried the move was this: Agentforce reached $1.5 billion ARR, up 240%. Combined with Data 360, the total was nearly $3.9 billion, up over 210%.

$1.5 billion in recurring revenue is not a demo. It's not a pilot. It's a business.

Salesforce has a $213 billion market cap. Agentforce represents roughly 3% of total annual revenue right now. That's still a small slice. But 240% growth on $450 million a year ago — compounding like that, it stops being a slice and starts being the whole pie, eventually.

The more interesting question is not whether Agentforce will grow. It's who would have built Agentforce if Salesforce hadn't.

The answer used to be Gong.

Gong started in 2015 as a conversation intelligence platform. It recorded sales calls, analyzed them with AI, and told managers what was working. It was genuinely the first thing of its kind. In 2021, at the peak of the AI and SaaS frenzy, Gong was valued at $7.25 billion — a 72x revenue multiple on $100 million in annual recurring revenue.

Today Gong has crossed $500 million in ARR with 55% growth. It has 4,500 customers including half the Fortune 10.

And it's worth $4.5 billion.

That's a 38% drop in valuation despite three times the revenue. The company can't go public without a down-round. Founders whose options were priced at the $7.25 billion level are holding underwater paper. Gong CEO Mark Raz said publicly that the company is in no rush to IPO, which is probably true — you're never in a rush to do something that destroys your previous investors.

The reason for the collapse has nothing to do with whether Gong's product works. The product works. Customers use it. Growth is accelerating.

The reason is that the thing Gong built — AI that sits on top of your sales process and tells you what to do — is now a feature inside Salesforce, HubSpot, Microsoft, and ZoomInfo. The platform that owns your customer data and your pipeline also owns the AI layer. Gong is selling to companies that already have a good enough version built into the system they pay millions for.

You can be a great product and still lose. If the people who own your customer's core workflow decide to build your product inside theirs, you're no longer a company. You're a feature they decided not to ship.

This is the structural dynamic that separates investable AI stories from ones that sound convincing but aren't.

Every week a new private company announces an AI sales framework, or an AI revenue platform, or an AI agent for pipeline management. The names change. The architecture changes. The pattern is the same.

These companies need distribution. Distribution means convincing sales leaders to adopt a new tool, train their teams on it, integrate it with their CRM, and trust it with their most important customer relationships.

Now Salesforce needs to do the same thing. But Salesforce is already the CRM. The reps already log in there every day. The managers already review deals there. The data already lives there. Salesforce doesn't need to convince anyone to adopt Agentforce — it just needs to turn it on.

That is not a marginal advantage. It's the only advantage that matters in software.

The numbers from the latest earnings tell the story. Agentforce delivered 3.2 billion work units in one quarter — up 97% from the prior quarter. Seven billion total across the platform. That's usage. That's behavior. Not pipeline, not intent, not a press release about what customers are exploring. Work units that were completed.

Meanwhile Salesforce generated $15.15 billion in free cash flow over the trailing twelve months, on a 34% free cash flow margin. The company trades at a P/E of 22 and a PEG ratio of 0.39. It's growing revenue at 11%, growing free cash flow at 21%, and it just raised full-year guidance by $200 million. It has $8.3 billion in cash against $71 billion in total debt — yes, that's a loaded balance sheet, but $15 billion in annual cash flow service that debt comfortably.

You're not paying a premium to own Salesforce's AI bet. You're buying a profitable machine at a reasonable multiple that happens to be growing its most interesting product at 240%.

Gong, by contrast, needs to convince buyers that a standalone AI sales tool is worth a separate seat, separate budget, and separate training — when the thing they already pay $500,000 a year to can now do the same thing inside the system.

What about the companies that aren't Salesforce? What about HubSpot, which also embeds AI across its platform? Or smaller players that focus on verticals Salesforce doesn't touch?

These are real questions. The pattern doesn't mean every specialized player dies. It means the bar gets much higher. The specialized player needs to do something the platform genuinely cannot — deeper analysis, different data sources, a workflow so specific that a generalist can't replicate it without bloating their product.

Most of them can't. Most of them don't. They build a better version of the thing the platform already does, and hope the platform is slow to follow. The platform is never slow enough to matter.

Gong spent five years being the best at conversation intelligence. Then four major platforms shipped conversation intelligence and their valuation dropped 38%. Being first and being best are both irrelevant when the winner is whoever owns the login screen.

Here's how to think about this when you hear about the next AI sales company.

Ask one question: who already owns the path to this buyer?

If the answer is a public company you can already buy — Salesforce, HubSpot, maybe Microsoft or Oracle depending on the vertical — then the private company is competing against a feature that costs its buyer almost nothing to add. The public company gets to cross-sell to millions of existing customers. The private company gets to cold-call.

If the answer is nobody — the market is fragmented, there's no dominant platform, or the workflow is genuinely new and hasn't been absorbed yet — then the private company might have real optionality. But that's a smaller set of cases than you'd think.

The AI-for-everything cycle is going to produce a lot of companies that build great products nobody pays for, because the people who already own the relationship built a good-enough version for free. The companies that own the relationship are already public. You can already own them.

The test is simple. When the next earnings season hits, look at whether Agentforce ARR is still growing at 200%+. Look at whether cRPO — the backlog of committed contracts — keeps accelerating at 14%. Look at whether the next quarterly guidance raise comes from organic AI growth instead of acquisitions.

If those numbers hold, the story isn't changing. It's accelerating.