Are your teams measured by how many features they deliver rather than the value those features create? If so, you’ve inherited—or hired—what many organisations call a feature factory. It looks productive on spreadsheets but quietly hollows out customer value and long-term growth.
Why the feature factory is so seductive
There’s an obvious appeal: velocity is measurable, predictable and easily reported to the board. But what gets counted often becomes the target. When leadership rewards shipped tickets, teams optimise for throughput. The result is a stream of releases that may tick technical boxes without moving the needle for customers.
Melissa Perri coined the phrase “feature factory” to describe organisations that prioritise output over outcome — a concept worth reading in full in i.e. Escaping the Build Trap.
Symptoms, and the real cost
- Feature bloat: the product becomes harder to use as new options pile up.
- Slow learning: teams rarely stop to validate hypotheses with real customers.
- Misaligned incentives: delivery metrics trump business or learning metrics.
The hidden cost isn’t just wasted engineering time. It’s lost trust from customers, increasing churn and missed market opportunities. A product that keeps changing superficially will see stagnant engagement; managers will then demand more features, kicking the cycle into another loop.
Three practical moves to turn the ship
Fixing a feature factory is organisational, not tactical. Here are three concrete levers product leaders can pull now.
1. Measure outcomes, not outputs
Replace vanity metrics with clear outcome metrics tied to customer behaviour. Examples:
- Activation rate for a new cohort rather than number of new sign-up features shipped.
- Retention or repeat usage for a cohort instead of count of UI tweaks deployed.
Use a north-star metric for each product area and make it visible. When KPIs tie directly to user value, prioritisation becomes straightforward.
2. Fund outcomes, not projects
Traditional project-based funding encourages lengthy requirements and handoffs. Move to a portfolio model where funds are allocated to teams against hypotheses and outcomes. A small, time-boxed fund for discovery ensures teams can validate ideas before engineering spends escalate.
Practical steps:
- Create a discovery budget separate from delivery budgets.
- Require hypotheses statements with clear success criteria before funding for build.
- Re-evaluate funding at regular milestones, not at single delivery gates.
3. Build empowered, outcome-driven teams
Autonomy without clear goals is chaos. The highest-performing organisations combine autonomy with accountability: small, cross-functional trios (product, design, engineering) owning a measurable outcome. Give them decision rights, but hold them responsible for the metric.
Spotify’s squad model is a useful reference for team autonomy and accountability. The idea is not to copy structures blindly but to adopt the principle: autonomous teams, aligned to outcomes, with rapid feedback loops.
A real-world example that illustrates the shift
When EDF Energy reinvented parts of our digital business, the organisation moved towards customer-centric digital channels as the lead channel for many services. That change wasn’t simply about building new features; it required new metrics, funding approaches and empowered teams to redesign key customer journeys. The result was not just more releases, but measurable uplift in digital engagement and reduced friction for customers.
Practical governance: protect discovery and reward learning
Governance often kills discovery with heavy-stage-gate processes. Instead, protect early-stage work with lighter governance but stricter learning criteria. Make three things clear to every team:
- What hypothesis they are testing.
- How they will measure success or failure.
- What will happen after the experiment—scale, iterate, or stop.
Celebrate validated failures. If a team disproves a costly assumption quickly, that’s saving the company time and money—a win worth rewarding.
Leadership matters: culture and incentives
Leaders set the language and incentives. Replace “What have we shipped?” with “What did we learn and who benefited?” Board reporting must reflect outcomes, not just velocity. That change aligns everyone from the CEO to the engineering manager around value creation, not task completion.
Where to start this quarter
- Pick one product area and swap its primary KPI from an output metric to a user behaviour metric.
- Allocate a modest discovery budget and require a hypothesis and success criteria to access it.
- Form or refine a product trio to own that outcome for one quarter, with full decision rights and visible metrics.
Shift is not one big move but a series of small, aligned experiments. Start with governance and measurement—those two changes alone recalibrate incentives and reveal whether teams will follow through.
Final thought
Feature factories are a symptom of misaligned incentives and brittle governance. The antidote is simple in concept and hard in practice: align teams around outcomes, fund discovery, and change what you measure. If you make a visible, repeatable habit of learning over shipping, you’ll be building products that matter—and that’s the one metric your board should care about.
Leave a Reply