Killing the Stage-Gate: How One Team Rebuilt the Innovation Pipeline
A snack manufacturer threw out its stage-gate process. Decision velocity tripled. Innovation hit rate didn't fall. Here's what replaced it — and what the team gave up to make it work.
The stage-gate process — five sequential gates, each requiring a cross-functional review committee, financial modeling, and a packaged business case — was the standard innovation governance model for nearly three decades. It produced predictable, defensible decisions and a steady cadence of mediocre line extensions. By 2024, the head of innovation at a $2 billion regional snack manufacturer had concluded that it was also the company's single largest obstacle to growth.
"We were running 18-month cycles on products that competitors were testing and launching in six months," she said. "The process wasn't filtering bad ideas — it was draining the energy out of the good ones."
What Replaced It
The replacement structure has three components. First, a smaller portfolio of explicit "innovation bets," capped at twelve at any time. Each bet has a single executive sponsor, a defined hypothesis, and a hard 90-day milestone. Second, a public scoreboard reviewed monthly by the leadership team, showing every bet's status against its milestone. Third — the most important and the hardest culturally — a kill criterion attached to each bet at inception. If the bet hasn't cleared its 90-day milestone, it ends. No second chances, no extensions, no rescue committees.
"The discipline isn't in starting new things," the innovation lead said. "It's in stopping the old ones. The stage-gate process let bad bets continue forever because nobody had to actively kill them. They just stalled."
The First Six Months
The transition was uncomfortable. In the first quarter under the new structure, four of the inherited innovation projects were terminated within their 90-day windows, including two that had been in the pipeline for more than two years. One of the four had a senior executive sponsor and a working prototype; it died because it couldn't articulate a sufficiently distinctive consumer benefit when measured against the new bet criteria.
"We thought we'd lose people over this. Instead, we kept the people we wanted to keep — and they were the most relieved."
By month six, the team had launched two products that under the old process would still have been in gate-three review. One was a clear commercial success; the other was a clear failure, terminated within a quarter of launch. The innovation lead considered both useful: "Fast wins and fast losses are equally valuable to the business. Slow either-or is the worst outcome."
The Numbers
Eighteen months in, the manufacturer's innovation metrics tell a clear story. Time from concept to retail test has fallen from 14 months to 4.5 months. The number of products in active development is down from 31 to 11. The number of products launched per year has roughly doubled. The hit rate — measured as products meeting their year-one revenue target — has held steady at 38 percent, slightly above the company's prior five-year average. The total cost of the innovation function has fallen, primarily through the elimination of stage-gate review committees that had grown to consume an estimated 1,200 person-hours per quarter.
The Cultural Cost
Two senior managers left the company during the transition; both had built their internal reputations on stage-gate stewardship. The innovation lead does not regret either departure but acknowledges the disruption. "Killing a process that people have built their careers around is a leadership tax. You pay it, and then the company moves faster. There isn't a free version of this."
For executives considering similar moves, her two pieces of advice: do not pilot the new process alongside the old one — that produces the worst of both — and put the kill criterion in writing before the bet starts. "If you negotiate the kill criterion at month three, you've already lost the discipline. The criterion is the whole point."