The Two Phases Every Product Lives In

Every product, at every point in its lifecycle, is operating in one of two modes. The first is a creative mode, where the work is about vision, judgment, and craft. The second is an analytical mode, where the work is about measurement, experimentation, and iteration. The mistake most product teams make is not understanding which mode they are in — and applying the tools of one to the problems of the other.

I call these Phase 1 and Phase 2. The distinction is simple. Phase 1 is before you have meaningful data. Phase 2 is after. But the implications of that distinction run deep into how a product team should be organised, what skills it needs, and what kind of decisions it should be making.

Phase 1: The Pre-Data Creative Mode

Phase 1 begins when a product idea exists and ends when real users are interacting with a real product. During this period, data is either absent or insufficient to guide decisions. You cannot A/B test a product that does not exist. You cannot analyse a funnel that has not been built. You cannot measure retention in a product with no users.

In Phase 1, the primary inputs are experience, pattern recognition, first principles thinking, and inspiration drawn from adjacent domains. A product manager in Phase 1 is doing something closer to design than analysis. They are making a series of bets about what users need, what the market will accept, and what the technology can deliver — and translating those bets into a product that others can build.

What Phase 1 Requires

Phase 1 demands a particular type of thinking that is difficult to systematise. The product manager must be able to hold a user's perspective without losing sight of technical constraints. They must be willing to make decisions with incomplete information, knowing that some of those decisions will turn out to be wrong. They must resist the temptation to over-engineer the solution before the problem is fully understood.

The most valuable inputs in Phase 1 are qualitative: conversations with potential users, competitive analysis, market research, and the accumulated pattern recognition of people who have solved similar problems before. Numbers that do exist — market size estimates, competitor metrics, early prototype feedback — should be treated as directional, not definitive.

Phase 1 is also where the foundational product decisions are made. The core user journey, the primary value proposition, the key interactions — these are set in Phase 1, and changing them later is expensive. This is why Phase 1 deserves more senior attention than it typically receives. The temptation is to move fast and figure it out later. The cost of that approach shows up in Phase 2, when you discover that the foundations were wrong.

Phase 2: The Data-Driven Optimisation Mode

Phase 2 begins when you have enough real user behaviour to make statistically meaningful observations. The threshold varies by product — a consumer application needs many thousands of users before individual metrics are reliable; a B2B tool might reveal important patterns with a few dozen active accounts. But the transition is real, and it changes what good product work looks like.

In Phase 2, the primary inputs shift from judgment to measurement. Where are users dropping off? Which features are driving retention? What is the conversion rate at each step of the onboarding flow? Which user segments are most engaged? These questions have answers — real answers, grounded in observed behaviour rather than assumption — and the product manager's job is to find them and act on them.

What Phase 2 Requires

Phase 2 demands analytical rigour that Phase 1 does not. The product manager needs to be comfortable with data — not necessarily as a statistician, but enough to distinguish signal from noise, to recognise when a dataset is too small to be meaningful, and to design experiments that will produce actionable conclusions.

The classic Phase 2 tools are A/B testing, funnel analysis, cohort analysis, and session recording. Used well, these tools reveal things about user behaviour that no amount of intuition or qualitative research would have surfaced. Users consistently do things that surprise product teams — click where designers did not expect, abandon flows at unexpected points, use features in ways they were not designed for. Phase 2 makes these behaviours visible.

Phase 2 also introduces a specific risk: local optimisation. A team that is purely focused on improving existing metrics can find itself making the product incrementally better at something that is no longer the most important thing. Conversion rates improve while the core value proposition quietly becomes less relevant. Retention improves for users who were already retained while the product fails to attract new users.

The Loop Between Phase 1 and Phase 2

The relationship between Phase 1 and Phase 2 is not linear. It is a loop. You build something in Phase 1, gather data in Phase 2, and most of what you do in Phase 2 is iterative improvement. But periodically, the data reveals something that cannot be solved by iteration — a fundamental assumption that was wrong, a user need that the current product cannot address, a market shift that the existing design cannot accommodate.

When that happens, you need to re-enter Phase 1. Not for the entire product — the parts that are working should continue to be optimised in Phase 2 — but for the specific problem that iteration cannot solve. This means stepping back from the metrics, thinking from first principles again, and making new bets that will then be tested in the next Phase 2 cycle.

Deliberately Triggering Phase 1 Thinking in a Mature Product

One of the habits I have developed over years of product leadership is deliberately scheduling Phase 1 thinking into the roadmap of mature products. Left to their own devices, product teams in Phase 2 tend to stay in Phase 2. The work is tractable, the feedback loops are fast, and there is always another metric to improve.

But incremental optimisation has diminishing returns. A team that has been optimising the same funnel for two years will struggle to find meaningful improvement within the existing product structure. At some point, the only way to achieve a step change in results is to reconsider the structure itself. That is a Phase 1 exercise, and it requires a different mindset than Phase 2 optimisation.

I schedule this deliberately. Quarterly brainstorming sessions where the team is explicitly asked to think as if the product does not exist yet. Competitive reviews where we ask what we would build from scratch if we were starting today. User research deep-dives designed not to validate existing features but to surface unmet needs. These exercises are Phase 1 thinking applied to a Phase 2 product, and they are where the most significant product improvements originate.

The Skill Set Implications

Understanding the two phases has practical implications for hiring and team composition. Phase 1 work rewards creative, generalist thinkers who can work comfortably with ambiguity. Phase 2 work rewards analytical, detail-oriented thinkers who are comfortable with data and experimentation. Both skill sets exist in the same person only rarely.

The best product teams I have built have had a mix — people who are naturally strong in Phase 1 paired with people who are naturally strong in Phase 2, with enough overlap to communicate across the divide. What does not work is a team that is homogeneous in one direction: all intuition and no analysis, or all analysis and no vision.

Knowing which phase your product is in — and staffing, prioritising, and measuring accordingly — is the product management skill that most job descriptions do not mention and most hiring processes do not test for.