Immersive Bold Targets — Why Incrementalism Kills Innovation

Every organization I’ve led or advised eventually hits the same wall. Not a crisis — something quieter and more dangerous. The goals get smaller. The targets get safer. Quarter by quarter, the collective ambition of the company deflates until you’re celebrating a 3% improvement in conversion rate like it’s a moonshot. I’ve seen this pattern at startups scaling past Series B and at enterprises with thousands of employees. The mechanism is always the same, and left unchecked, it will hollow out your competitive position.

The Gravity of Incrementalism

Growth-stage companies are especially vulnerable. In the early days, everything is a bold bet — you have no choice. But as the company matures, something shifts. You hire experienced operators. You build a management layer. You introduce quarterly planning and structured goal-setting. All good things, in theory. But here’s the trap: those structures naturally optimize for predictability, not ambition.

Middle management — and I say this as someone who has built and managed many layers of it — becomes a filter for ambition. A director sets a target they’re confident their team can hit, because their performance review depends on it. Their VP rolls up those targets and presents a plan that looks achievable. By the time it reaches the exec team, you’re looking at a portfolio of modest improvements that nobody will get fired for proposing.

The incentive structure is the root cause. When people are rewarded for hitting targets and penalized for missing them, rational actors will set targets they can hit. This is not a people problem — it’s a system design problem. And if you’re a CTO or CPO who hasn’t addressed it, you’re complicit.

The Real Cost of Playing It Safe

The insidious thing about incrementalism is that it feels like progress. You ship features. Metrics go up slightly. Dashboards look green. But zoom out 18 months and ask: have we meaningfully changed our market position? Have we built something a competitor couldn’t replicate in a quarter? Usually the answer is no.

I learned this the hard way at one of my earlier companies. We spent nearly a year optimizing our identity verification flow — shaving off seconds here, improving pass rates by a percentage point there. All measurable. All defensible in a board meeting. Meanwhile, a competitor launched a fully automated onboarding pipeline that made our incremental improvements irrelevant overnight. We’d been polishing the deck chairs.

The math is stark. Ten initiatives each delivering a 3% improvement do not compound into a 30% transformation. They compound into organizational exhaustion and a product that’s marginally better but fundamentally unchanged. Meanwhile, the team that bet on one 10x initiative — even if they only achieved 5x — has leapfrogged you.

OKRs as the Antidote — But Only If You’re Honest About It

I’m a believer in OKRs, but I’ll be the first to admit that most implementations are garbage. They degenerate into task lists within two quarters. Teams write Objectives that are really just project descriptions and Key Results that are really just deliverables. “Objective: Launch new dashboard. KR1: Ship dashboard by March. KR2: Add three chart types. KR3: Write documentation.” That’s a project plan wearing an OKR costume.

The “O” in OKR needs to describe a state of the world you want to create — not a thing you want to ship. “Our enterprise customers self-configure 80% of their onboarding without support intervention” is an Objective. It doesn’t prescribe a solution. It forces the team to think about what would have to be true for that outcome to exist. Maybe it’s a better dashboard. Maybe it’s an AI assistant. Maybe it’s a fundamentally different product architecture. The Objective doesn’t care — it demands the outcome.

Key Results should be measurable proof points that you’re getting there. They should make the team slightly uncomfortable. If every KR feels achievable with the current approach, you’ve set the bar too low.

The 10x Forcing Function

This is the principle I come back to most often: if you set a target of 10% improvement, your team will optimize the existing system. If you set a target of 10x improvement, they have to throw out the existing system and design something new. Both are valid management tools, but only one produces innovation.

I used this at IDnow when we were looking at our fraud detection pipeline. The team’s initial proposal was to improve detection accuracy by 5-8% through better model tuning and additional training data. Reasonable. Defensible. Incremental. Instead, I challenged them: what would it take to reduce false positives by 10x while maintaining the same catch rate? That question forced a completely different conversation — one that led to a multi-signal architecture combining behavioral analytics, device fingerprinting, and document forensics in a way none of us had originally envisioned. We didn’t hit 10x. We hit roughly 4x. But 4x is a competitive moat. 8% is a footnote.

The key insight: the 10x target wasn’t about literally achieving 10x. It was about making the existing approach obviously insufficient, forcing the team into creative problem-solving territory. You can’t get to 10x by working harder or adding more engineers. You get there by rethinking the problem.

What I’ve experienced consistently is that bold targets make people think about the problem in a completely different way. When the target is 10%, the team asks “how do we improve what we have?” When the target is 10x, they ask “what would we build if we were starting from scratch?” That shift in framing — from optimization to reinvention — leads to fundamentally different approaches. At IDnow, the fraud detection team didn’t just tune their existing models. They stepped back and asked “what if we combined signals that have never been combined before?” That question would never have surfaced under a modest improvement target. The bold target didn’t just change the outcome — it changed how the team thought about the problem entirely.

Creating the Conditions for Bold Bets

None of this works without psychological safety — and I don’t mean that as a platitude. I mean very specific, observable leadership behaviors.

Score OKRs honestly and celebrate ambitious misses. If a team sets a genuinely bold Objective and achieves 60% of it, that needs to be treated as a better outcome than a team that set a safe target and hit 100%. If you say this but then promote the person who hit their safe targets, everyone notices. Your incentive structure speaks louder than your all-hands slides.

Separate OKRs from performance reviews. This is non-negotiable. The moment individual compensation is directly tied to OKR attainment, you’ve killed the incentive to be ambitious. OKRs should inform performance conversations but not mechanically determine bonuses.

Fund exploration explicitly. At every company where I’ve driven real innovation, we carved out explicit capacity for exploratory work — typically 15-20% of engineering bandwidth. Not “20% time” in the Google sense where it’s a perk. Structured exploration with clear problem statements but no predetermined solutions. This is where your 10x ideas come from.

Kill projects publicly and without blame. Bold bets mean some bets won’t pay off. When you shut down a project that didn’t work, how you handle it sets the tone for the next round of goal-setting. If the team that took a big swing gets reassigned to maintenance work as punishment, you’ve just taught the entire organization to play it safe.

The Portfolio Approach

I’m not arguing that every goal should be a moonshot. That’s a recipe for chaos. What I am arguing for is a deliberate portfolio balance. At any given time, roughly 70% of your engineering effort should be on reliable, incremental improvements to your core product. About 20% should be on strategic bets — ambitious but grounded in clear market signals. And 10% should be on genuine high-risk, high-reward exploration.

The problem in most organizations isn’t that they lack this framework — it’s that the 70% gradually consumes the 20% and 10%. Every quarter, some urgent customer request or competitive response eats into the exploration budget. Within a year, you’re running at 95/5/0 and wondering why your product roadmap feels uninspired.

Protecting that 20-30% of non-incremental work is a leadership discipline. It requires saying no to short-term revenue opportunities. It requires defending a team’s time against the gravitational pull of the backlog. It’s uncomfortable, and it’s one of the most important things a technology leader does.

The Uncomfortable Truth

Here’s what I’ve come to believe after 25 years of building products and leading technology organizations: the biggest risk in a scaling company is not that you’ll take a bold bet and fail. It’s that you’ll never take the bold bet at all. That you’ll optimize your way into irrelevance, one safe quarterly target at a time.

The organizations that break through — the ones that build genuine competitive moats and create outsized value — are the ones where leadership has created the conditions for ambition. Where the goal-setting process rewards creative leaps, not just reliable execution. Where missing an audacious target is treated as more valuable data than hitting a mediocre one.

Set targets that make your team uncomfortable. Then give them the safety and space to pursue those targets without fear. That’s the job.