Ask a digital leader what's really slowing their optimisation programme down, and the answer is rarely 'a shortage of ideas'.
More often, we find backlogs are full, test hypotheses are queued and insight reports are flagging opportunity after opportunity.
The problem sits one step further along: those ideas are waiting on a decision that hasn't been made, and won't be made this week either.
We call this decision debt. Like technical debt, it accumulates behind the scenes, one deferred choice at a time, and each deferral makes the next decision a little slower and a little more expensive to reach.
The decision avoidance pattern
A test result comes in and the finding is clear, but what follows is a request for more data, a slot on next month's meeting, or a note to "align with the wider team first."
None of these responses is unreasonable but they form a pattern: the organisation has a way of avoiding decisions in a way that feels like diligence, and it's a pattern that repeats every time a decision needs to be made.
Why this gets mistaken for other problems
Decision debt hides well because its symptoms look like something else.
A growing backlog looks like a resourcing problem, so the response is to hire. A slow test cadence looks like a tooling problem, so the response is to buy new software. A team that's stopped proposing bold ideas looks like a skills or collaboration problem, so the response is training or team-building.
Every one of these responses can help to an extent but none of them addresses a queue of decisions that nobody has the authority, confidence or mandate to clear. This is why decision debt is so easy to miss inside a digital operating model that looks, on paper, like it's working.
How the debt builds
Every unmade decision does two things: It blocks the specific piece of work waiting on it, and it changes the conditions for whatever decision comes next.
A test that should have happened last month ends up bundled with three others by the time the team finally discusses it, which makes the discussion longer and raises the stakes. A personalisation rule that could have launched quickly is left waiting - so by the time the team reviews it, more people have opinions about it and there's more to undo if it's wrong.
The debt then stops being about any single decision. It's the growing weight sitting on every decision still to come.
Spot it before it spreads
Decision debt has a handful of tells, and once you know them they're hard to miss.
Meetings get scheduled to discuss a result rather than act on one. The same test appears on three consecutive steering agendas with no new information between them. A junior team member has the data to make a call but waits for someone senior to make it instead, not because they're told to, but because that's how it's always worked. Insight teams start pre-empting the question "what should we do about this" by attaching a recommendation nobody asked for, because unprompted recommendations get ignored less often than open findings do.
None of these on their own means much. Together, they're the clearest sign that a digital operating model has a decision problem rather than an idea problem.
Clearing the queue
Clearing decision debt means giving a small number of decisions a clear owner, a threshold for what needs escalating and what doesn't, and a commitment to act on evidence within days rather than within a quarterly cycle. That means someone can approve a test to launch without wider sign-off, a rule for how long a "let's discuss it" can sit before it needs a decision, and treating an inconclusive result as a decision in its own right (stop, adjust, or run again) rather than a reason to wait for more data.
This is core to the operating model work we do inside our Optimise service. A KPI framework tells you what matters. An operating model tells you who decides what to do about it, and how quickly. Without both, insight accumulates but nothing changes.
Reducing the backlog
Fix this, and decision volume doesn't need to change. What changes is the pace: the same decisions, made faster and with less second-guessing, because the rules for who decides and when were set in advance rather than negotiated fresh every time The backlog is reduced because fewer items sit waiting for a decision that was always going to be made eventually.
Where to start
If tests are queued, insight is available, and progress still isn't translating into outcomes, the Digital Optimisation Health Check will show you where decisions are stalling across strategy, data, operating model and performance.
