5 tips to get from stuck to unstuck.

Most organizations hire fixers when they're already six months past stuck. 

The analytics implementation stalled. The consent strategy broke. The retail media reporting framework doesn't work. Everyone knows there's a problem. Nobody knows how to fix it. 

It's often because they're looking for certainty that doesn't exist. A comprehensive root cause analysis, stakeholder alignment workshops, and a phased remediation roadmap. 

The solution is simpler: stop having meetings about the problem and start making different decisions. 

Create movement. Create momentum. You'll never get unstuck by doing nothing.

After two decades of unsticking analytics projects, I've learned that decision paralysis is the problem, not a symptom of it. The technical challenges are real, but solvable. What's not solvable is an organization more comfortable analyzing than deciding. 

Here are some of the ways to short-circuit it:

1. Kill the "Let's gather more data" reflex

If you've been stuck for three months, you have enough data. You're avoiding a decision. Name what decision you're avoiding, then make it. 

Monday.

Or better… today. 

I once walked into a retail media project where they'd spent five months "researching attribution models" because nobody wanted to admit the real decision was whether to show advertisers conservative numbers or inflated ones. 

We picked conservative. 

Launched in three weeks.

2. Find the decision that unlocks five others

Stuck projects have dependency chains. "We can't launch reporting until we finalize the attribution model, until we agree on KPI definitions, until..." Find the one decision at the top of the chain and force it. Everything else cascades.

3. Replace "What's the right answer?" with "What can we test in two weeks?"

Perfect solutions take six months to design and fail in production, because the context has already changed. 

Testable hypotheses take two weeks and teach you what actually works. Ship the imperfect version. 

Learn fast. Iterate.

The best consent management fix I ever implemented was technically "wrong" by the textbook. But it worked, it shipped in ten days, and we refined it based on real user behavior instead of theoretical edge cases.

4. Eliminate the veto holders

Every stuck project has someone who can say "no" or "later" but never has to say yes. Legal, compliance, IT security—all valid stakeholders, all capable of blocking indefinitely because of risk, uncertainty, or lack of understanding. 

Give them a deadline: "We launch in two weeks with your input or without it. Your choice."

5. Stop optimizing what you'll never ship

I've seen analytics teams spend four months perfecting dashboard designs that never launched because they were still debating data governance. 

Shipping a useful-but-imperfect solution beats perfecting one that stays in staging.

The pattern

Stuck organizations drown in analysis. Unstuck organizations make smaller, faster decisions and learn from what breaks. 

It's Bezos's two-way doors. Most decisions are reversible. We can change our minds, we can undo most choices. 

But we can't undo inaction. 

You don't need more data. You need someone willing to make a call.

That's what fixers do. We show up, identify the decision everyone's avoiding, and force it. The project usually unsticks within a week.

The hard part isn't the technical problem. It's admitting you've been avoiding the decision.

Sometimes you need an outsider to say it.