Why AI Makes This the Right Moment to Fix Vendor Spend

After leaving Affirm, I spoke with 75+ companies of all sizes trying to understand whether this was solvable. Nearly all of them were managing vendor spend, contracts, and renewals in spreadsheets. Not because they liked it, but because nothing better existed for them. Enterprise solutions were built for Fortune 500 procurement teams. Everything else was either too narrow (just invoices, or just approvals) or too expensive to justify.

This is a core financial workflow. And the world was managing it in spreadsheets.

I kept thinking about what things looked like before Salesforce. Before Carta. CRM in a spreadsheet. Cap table in a spreadsheet. The problem wasn't that companies didn't care. Nobody had built the right system of record yet. That's what we set out to build with Stackpack.

Every Tech Wave Reshapes Finance

Cloud gave finance teams access. Big data gave them visibility. AI is giving them something harder to name: call it judgment.

For a long time, finance was fundamentally a reporting function. You closed the books. You produced the deck. You answered questions after the fact. Big data started shifting that, and suddenly there were more signals than ever. But most finance teams are still drowning in data they can't act on fast enough.

AI closes that gap by collapsing the distance between information and action. The shift is from "here's what happened" to "here's what you should do," and that's a fundamentally different role for finance.

Vendor and SaaS spend sits right at the center of this. It's one of the largest and fastest-moving line items on the P&L. It touches every department. And it's notoriously hard to understand. Imagine trying to manage revenue without a CRM: no pipeline, no attribution, no way to forecast. That's where most companies are with vendor spend right now.

Three Problems That Compound on Each Other

When I describe what Stackpack solves, I always start here, because the problems aren't separate. They stack.

• Time. Finance and ops teams are chasing information that should already be in front of them: contract terms, renewal dates, who actually owns a vendor relationship. Hours and days per quarter, gone.
• Money. Unused seats nobody's touched in months. Duplicate tools doing the same job. Auto-renewals that fired before anyone had a chance to push back. Shadow SaaS that was never in the budget. We're seeing 20-30% savings opportunities across our customer base for companies without strong prior visibility. Across $3.7B in spend on the platform, that adds up fast.
• Decisions. You can have all the data in the world and still make the wrong call because you're missing context. Negotiating a renewal without usage data. Approving a budget without knowing what's already being spent off the books. The data was technically available. The insight wasn't.

And there's a fourth problem emerging in real time: AI agent spend. No attribution, no governance, no ownership. It's 2012 SaaS sprawl happening again, but faster and at higher cost. We're building the infrastructure to manage it before it becomes the next uncontrolled line item on every P&L.

Visibility Isn't the Same as Action

Companies have never had more dashboards, more integrations, more real-time data. And the decisions aren't getting proportionally better.

Data without context is noise. A CFO staring at a $4M LLM spend number doesn't know what to do with it until they know which of that is waste, which is underleveraged, and which is actually driving revenue. The tools proliferated faster than the insight did. So you end up with a very expensive, very well-instrumented version of the same guessing game.

Visibility tells you what's there. It doesn't tell you so what, or what next.

Most tools stop at the first question. You can see the spend, the contract, the renewal date. But knowing a renewal is in 30 days doesn't move the needle if you still have to manually pull the contract, chase down usage data, figure out your negotiating position, and draft the first email yourself. That's where the decision dies: not from lack of information, but from the work required to act on it.

That's the gap Stackpack closes with agents. Not agents that surface the insight, but agents that do the work. A contract retrieval agent that finds and summarizes agreements before you even ask. A renewal agent that flags what's coming, reads the usage data, and kicks off the process automatically. A negotiation agent that identifies where you have leverage and helps you open the conversation. The last mile from insight to action used to fall entirely on the finance team. Now it's infrastructure.

Day One vs. In the Trenches

The CFO use case looks different depending on where you are in the journey.

New in the seat: the first 90 days are about establishing credibility fast. SaaS spend is one of the best places to do it because it's visible and actionable, and most organizations have no real picture of what they're running. A new CFO who can walk into a board meeting within weeks and report exactly what the company is spending, where the savings opportunities are, and what's renewing in the next quarter looks like they have command of the business. Stackpack gives you that picture faster than anything else. You're not waiting on a vendor audit or hunting for an IT spreadsheet.

Deep in it: the dynamic shifts. Boards are pushing for efficiency. Every line item is under scrutiny. The question moves from "what do we have" to "what can we cut, what do we need to negotiate, and where are we actually getting value?" It's also essential during audits and due diligence, whether you're raising capital or in a transaction.

Day one, you need control. In the trenches, you need leverage. Both come down to the same thing: a real-time system of record that surfaces the right information at the right moment.

Why Now

95% of the time, when we're talking to a prospect, we're not competing with another software vendor. We're competing with a spreadsheet.

That's a sign the category hasn't been properly built yet. That's exactly what CRM looked like before Salesforce. What cap table management looked like before Carta. The market was there. The behavior was there. The right software just didn't exist yet.

AI made this the right moment to build Stackpack: not because AI is a feature, but because it's what finally makes the last mile possible. Surfacing insight is table stakes now. Acting on it automatically, before the renewal slips, before the budget is set, before the contract auto-fires, that's the product.

Every company I talk to is dealing with this problem today. The question is just whether you want to get ahead of it or clean it up later.