Four people. Under 70 days. A new product category.
Two developers (one part-time), a part-time designer, and one product lead took a new product category from concept to public launch in a single quarter. The AI helped. The process is what did it.
<70 days
Concept to public launch
4
People on the build, two part-time
1
Platform, reusable across the product line
PerformLine is a leading marketing-compliance firm, with more than a decade monitoring web, social, email, and call channels for banks, credit card companies, fintechs, and mortgage lenders. By 2026 a new channel had shown up that none of its clients controlled and none of its competitors monitored: AI assistants. ChatGPT, Gemini, Claude, and Copilot were answering consumers’ questions about financial products directly (rates, fees, disclosures). And when an answer was wrong, the regulatory exposure landed on the financial institution. The AI provider carried none of it, which is wild. The bank didn’t write the sentence and the bank owns the risk. Whoever answered that first would define the category.
- Industry
- RegTech · financial services compliance
- Product
- AI Response Monitor, launched May 4, 2026
- Team
- 2 developers (one part-time) · part-time designer · fractional product lead
- Engagement
- Fractional product lead, a few hours a week
“A new product category needs a staffed-up initiative: a department, a quarter of planning, a roadmap committee.”
“Four people running a startup discovery loop (research, rapid iteration, clients reacting to working software) shipped it in under 70 days, with AI holding the guardrails while the process carried the load.”
The conventional enterprise answer to a window like this is a big one, and I get why. A new category feels like it deserves a department. But a department has to coordinate with itself before it does anything, and that overhead’s massive and never shows up on the plan. Spend a quarter planning and somebody else defines the category first. Then you’re screwed, selling against their definition.
PerformLine went the other way, and that decision is most of this case study. The team was already strong. What the engagement added was a way of working that lets a handful of people move at a speed the org chart says should take a department.
- A team of four, on purpose: two developers (one part-time), a part-time designer, one product lead. Nobody to hand anything off to, nobody waiting on anybody.
- Scope questions got answered the same day, because the person who could answer them was already in the conversation.
- The roadmap stayed one product deep. One thing, shipped well.
The build ran like a startup discovery loop, inside an enterprise, which is nearly unheard of. Why so rare? Enterprise process exists to reduce risk, and from outside a discovery loop looks like risk, so it gets planned away before anybody tries it.
- 01
Research first
Frame the problem before writing software. Which AI platforms say what about regulated products, and where does the compliance exposure actually sit?
- 02
Rapid iteration
Working software early and often, with AI holding the guardrails: research compressed, prototypes in hours, a tiny team kept consistent at speed. Take the AI away and the method still works. Take the method away and the AI just builds the wrong thing on steroids.
- 03
Users in the loop from day one
Sales took the concept into client conversations while it was still taking shape, and what clients said went straight back into scope and positioning. By launch, the account teams had helped shape the product they were selling. That’s the closest thing to magic in the method.
| The default | This launch | |
|---|---|---|
| Build team | A staffed-up initiative | 4 people, two part-time |
| Concept to public launch | Quarters | Under 70 days |
| What’s left behind | A one-off product | A platform reusable across the product line |
AI Response Monitor was announced May 4, 2026, first into a category PerformLine’s clients were only beginning to ask about. It moved the company from monitoring the channels everyone knows to defining the one nobody else had touched. The technology underneath was architected so other PerformLine products can build on it, so one small team’s ten weeks turned into infrastructure for the product line.
“I love that PerformLine is ahead of the curve… we’re just starting these conversations internally and PerformLine’s already found the solve.”
Company, product, and launch date are public record (press release, May 4, 2026). The build-team size and timeline are from inside the engagement, published with PerformLine’s knowledge.
Four people launched a category-defining enterprise product in under 70 days, for reasons that aren’t exotic. One person owned it. The team was small enough that decisions got made in the conversation where the question came up. And research, iteration, and user feedback ran as a discipline, where most companies keep them on a slide. The capacity is almost always already in the building, and a fast launch runs on process and ownership, the cheapest things to add.
What could four people in your business ship if every decision got made in hours?
The Next Version finds where the money is actually stuck (starting with whether your numbers are even real) and names each move, what it’s worth, and who runs it. You keep the roadmap either way.