How AI connects sales and product management

From feature factory to strategic partnership

The software industry has changed completely in how products are built. I started my career at a thirty-person software company in Hamburg after studying sociology and philosophy, selling knowledge management systems to large German organizations. Back then, the process was simple but flawed.

Sales would return from meetings where customers had seen something elsewhere they absolutely needed. Whether it made sense for the product or the market didn’t matter. Every big deal triggered new feature demands. Product teams scrambled, engineering dropped planned work, and technical debt piled up. The loudest customer dictated the roadmap — a pattern that still exists in many places today.

The shift toward a real partnership between sales and product took years. In smaller companies, product teams focused on core functionality and reacted to customer RFPs. In larger ones like GitLab, product and sales now shape the market together.

At GitLab, I manage enterprise accounts in financial services and insurance, where deal cycles are long and complex. In large software companies, collaboration between sales and product is structured, not reactive. Product shapes direction based on insight, while sales brings context from the field. Both rely on data, not escalation.

Building products in public

At GitLab, I found that the approach to product development is built on transparency, which is also one of GitLab’s core values. Any GitLab customer can write a feature request, describe what they need, and open an GitLab issue. These issues live in the same environment we use internally.

Sales teams can link all relevant customers anonymously to specific issues, adding context about the segment or industry while keeping discussions public. The result is a living record of market needs that both validates and challenges assumptions.

Having a clear view on overlapping requests and problems from customers in banking or insurance, we see clear demand. When an issue sits untouched for years, it might not be as critical as it first appeared.

This public workflow builds trust. Product managers don’t rely on filtered summaries — they can read real customer input, comment, and follow progress from idea to release. Some issues have been open for a couple of years, others move quickly from concept to production. The key is that everything is visible.

That transparency can create the foundation for something powerful: AI can now work on top of structured, open data to connect feedback, context, and action.

Challenging customers is part of the job

Over the years, I’ve learned that saying yes to everything is the fastest way to destroy focus. You have to learn to challenge customers. That feature might seem essential right now, but does it really fit into their process? Have they thought through how it would actually work?

In my early days, the companies I’ve worked at have built many features that customers swore were dealbreakers. We shipped them, and they were never used. They lived in the product, added maintenance costs, but created no value. That taught me that good salespeople don’t just pass requests along — they interpret them.

When a customer asks for something, the real question is: what problem are they trying to solve? Is it unique or something others struggle with too? Could we solve it in a broader way that helps the entire segment?

Every company tends to believe to be special. Every FinServ says it operates differently. But the truth is, most share very similar challenges — compliance, reporting, data security. The job is to find patterns, not exceptions. That’s how sales insights become product direction.

Product transparency meets sales pressure

In Sales, revenue is foremost how we’re measured. It also reflects customer satisfaction, market fit, and business health. That pressure is constant, and it’s easy to fall into short-term thinking. But over time, I’ve developed deep empathy for the product side.

In many companies, engineering is “rebuilding the helicopter while flying.” If we keep stacking on new features, we’ll never fix the fundamentals. Product teams need protected time to improve reliability, scalability, and user experience.

GitLab’s transparency helps manage this balance. The open issue tracker, public roadmap, and regular releases give everyone visibility. Product leaders meet customers directly, while solution architects bridge technical and commercial conversations. Service Pings then close the loop — we can see which features deliver value and which should be retired.

Ideally, when a salesperson pushes for a feature and product shows data proving it doesn’t work, that’s no longer a debate. Evidence should replace opinion. It’s collaboration, not confrontation.

AI as the connective tissue between sales and product

At GitLab, AI isn’t a separate initiative — it’s also an integral part of the product and a part of how we work. From structured notes to shared issue tracking, AI bridges the communication gaps that can slow us down.

Enterprise sales is still a people business and I don’t think AI will replace that. Deals can take twelve to thirty-six months, with many stakeholders and decisions. Empathy and judgment matter more than automation. But AI can help to handle everything that gets in the way of those human moments.

How we at GitLab use AI tools like Claude to manage documentation and context is kept very transparently in our handbook. When I talk to a customer, I always summarize our discussion so my colleague knows exactly what happened — the customers never has to repeat themselves. That simple step keeps continuity across accounts and ensures product has the right context when new issues or opportunities come up.

AI can help turning fragmented human communication into structured knowledge. It connects conversations, documents, and product data — the glue between what’s said in meetings and what gets built later. AI can ensure continuity without replacing relationships.

The biggest limitation today is integration. In enterprise sales, you typically use twenty or thirty tools from different vendors. They need to connect so information can flow freely. That’s when AI becomes truly powerful.

The future isn’t AI closing deals — it’s AI coordinating the details. Systems that surface relevant documents, next steps, and feature discussions automatically. The bridge between sales and product will be as much technological as human.

Size and stage define collaboration

The relationship between sales and product depends entirely on company maturity. The basics are the same everywhere, but how they play out changes with scale.

In startups, when you’re chasing your first customers, product must listen closely to sales. Every deal provides vital market intelligence.

In mature companies like GitLab, product has a broader perspective — usage data across thousands of customers. Product knows what drives adoption and what causes friction. At that stage, sales needs to trust product’s judgment while continuing to surface fresh signals from the field.

Whenever I join a new company, I spend the first few weeks simply understanding where it sits on that spectrum. I have coffee chats with both sides. You can’t fix collaboration before you understand context.

The human element in digital transformation

Despite everything technology can do, enterprise sales will always depend on people. In the not too distant past working for other companies, I’ve seen colleagues walk into meetings without laptops, just to talk. It might sound old-fashioned, but the point stands: relationships drive business.

The best sellers blend tech fluency with empathy. I use digital tools for structure but stay present in the room. I still carry a notebook, but often sometimes it’s better not to stare at a screen while someone’s talking.

The future is hybrid. AI will take notes, summarize meetings, and manage reminders. Humans will focus on creativity, negotiation, and trust. What matters most is technology openness — staying curious and willing to learn new tools, even when they’re imperfect.

That’s how you stay relevant. You adapt, you experiment, and you let technology amplify your judgment, not replace it.

Shared lessons

For product teams:

• Join at least one real sales call each week.
• Use shared issue systems instead of private feedback lists.
• Publish rolling roadmaps with clear disclaimers.
• Protect engineering time for fundamentals.
• Base prioritization on data, not opinions.

For sales teams:

• Document every meeting consistently.
• Link customer requests to context and segment.
• Review patterns with product regularly.
• Use AI for structure and recall, not shortcuts.
• Manage customer expectations transparently.

Both sides want the same outcome: products that solve real problems and relationships that last. The bridge is trust built on shared context — strengthened by AI.

At GitLab I experience what great things can happen, when transparency and technology meet empathy. AI doesn’t replace the connection between sales and product — it deepens it. Open systems, shared knowledge, and structured context make collaboration faster and more grounded.

As AI becomes more integrated, the most valuable skills will still be human: listening, judgment, and curiosity. The future belongs to teams that use technology to connect understanding, not replace it.

Perspectives: Honest conversations on crafting great products over a cup of coffee.

I sit down with friends across design, data, engineering, ops, and more—people who work closely with product leaders, from PMs to CPOs. We talk about how teams really work, where things break down, and how AI and new ways of working are reshaping the future of product.

Coffee of the day

1000. Cups · Muhehewe · Madrid