AI has rewritten the speed limits of software. Teams can ship prototypes in days, not months. Solo founders can launch startups and exit in under a year.
But with speed comes a new kind of risk — building the wrong thing faster.
As Joris Dieben, SVP of Product at Gainsight, shared at SaaSiest Amsterdam 2025, AI doesn’t make product management obsolete. It makes good product management more essential than ever.
“Engineering can go fast. The question is whether product has given them the right thing to build.”
To adapt, Joris developed a framework for how SaaS companies can reach product-market fit in the AI era — by pairing the fundamentals of strategy with the acceleration of AI.
The myth of “build it and they will come”
The temptation today is to skip the hard parts. With AI tools like Cursor, Builder.io, or Galileo, anyone can prototype a new app overnight.
But history is full of examples where great technology failed because it solved the wrong problem.
Take the Concorde — a triumph of engineering that cut flight time from London to New York in half.
The problem?
People didn’t want faster flights; they wanted cheaper, more comfortable ones.
Or New Coke, the 1980s reformulation meant to beat Pepsi’s taste tests. Sales dropped because customers didn’t want a sweeter Coke — they wanted their Coke.
The lesson is timeless: technology can make you faster, but only clarity makes you right.
Step 1: Align on a living strategy
Joris’s first rule: AI can’t fix a fuzzy strategy.
Before running a single prompt, make sure everyone agrees on the basics:
- Who is your product for?
- What problem are you solving?
- How is your solution different?
- Who are your competitors?
- Why will you win?
He challenges teams to fill out this strategy canvas individually and compare answers. If they diverge, you don’t have alignment — you have noise.
“If you and your colleagues can’t describe your strategy the same way, AI won’t help you. It needs clear context to be useful.”
From there, choose 3–4 differentiators and double down.
Are you for mid-market or enterprise? Are you premium or value? Simple or configurable? Without distinct choices, you’ll drown in the noise of a 15,000-vendor MarTech landscape.
Step 2: Combine internal and external data
The next step is building what Joris calls a data spine — connecting the dots between internal feedback and external signals.
Internal data sources:
- Support tickets and chat transcripts
- Sales and success call notes
- NPS, survey responses, and churn reasons
- Product usage and feature adoption
External data sources:
- Competitor G2/Capterra reviews
- LinkedIn posts and ad campaigns
- Pricing updates and release notes
- Buyer discussions in communities and forums
You don’t need a data warehouse to start. Use Zapier, n8n, or even Google Sheets to stitch together basic automations. Then add scrapers or AI browsing agents for public data.
“The problem isn’t lack of data — it’s lack of access and action.”
Once you’ve built your data spine, you can finally feed AI something meaningful to work with.
Step 3: Use AI to spot patterns and opportunities
Now comes the fun part. Once you’ve aligned on strategy and aggregated data, AI can help uncover patterns faster than any analyst could.
Ask AI structured questions like:
- What are the top recurring feature requests by segment?
- Which workflows generate the most support friction?
- What competitor features are trending in user reviews?
- Which product themes drive the most positive sentiment?
Use these insights to generate opportunity briefs — short, structured summaries that capture the problem, audience, data, and potential impact of a new idea.
AI can create the first draft; product teams validate and refine. Score each brief by impact vs effort, and prioritise what matters most.
“We don’t start building anything unless the homework is done.”
Step 4: Prototype faster, validate earlier
AI can cut the time between idea and customer feedback to hours instead of weeks.
Use tools like Figma, Galileo AI, Replit Agents, or Builder.io to create clickable prototypes or light functional apps.
Feed in your opportunity brief and let AI generate wireframes or mockups that match your strategy and brand.
Then — and this is crucial — show them to customers.
Joris recommends validating every prototype with 10–20 customers before any engineering work starts.
“If you build faster, you need to talk to customers more, not less.”
That’s the new product rhythm:
- Align on strategy
- Feed your data spine
- Generate opportunities
- Prototype and validate
- Only then, build
Step 5: Stay grounded in evidence
Even as AI speeds up ideation, it can’t replace customer validation or disciplined measurement.
Product teams still need to define what success looks like — not in vanity metrics, but in real outcomes.
Examples of telemetry that matters:
- Task completion rates in new features
- Time to value for new customers
- Ticket volume per 1,000 users
- Expansion or retention deltas post-launch
Every launch should have a feedback loop built in. Learning is the product.
Step 6: Redefine the role of the product manager
AI doesn’t replace PMs. It changes where they create value.
- Less grunt work. Use AI to summarise data, cluster feedback, and create opportunity briefs.
- More time upstream. Spend that time talking to customers and validating problems.
- Less documentation theatre. Replace PRDs with prototypes.
- More evidence-driven conviction. Every bet starts and ends with data.
The result is a faster, smarter product loop — one where strategy sets the direction, data fuels discovery, and AI accelerates everything in between.
Final takeaway
Product-market fit still starts with people, not prompts.
AI can help you see faster, build faster, and test faster — but it can’t tell you why something matters. That’s still the PM’s job.
“The lean startup approach isn’t dead. It just runs at AI speed now.”
So the new rulebook is simple:
- Get aligned on what you’re solving.
- Build your data spine.
- Let AI surface opportunities.
- Prototype fast, validate faster.
- Ship what matters.
That’s how SaaS companies will win product-market fit in the AI era.
Watch Joris Dieben’s full session from SaaSiest Amsterdam 2025 here: https://saasiest.com/build-product-market-fit-in-the-ai-era/
