Founder-Only Approvals Kill Innovation: ButterflyMX's Playbook for Fixing It
ButterflyMX's founder says innovation dies in the gap between spotting an opportunity and acting on it. His fix: delegate real authority, protect workload time, and treat failed tests as evidence.
By Nathan Brooks
4 min read

What's News
- ButterflyMX's founder argues innovation stalls when approvals built for a five-minute decision now take three meetings and a week
- The essay cites a two-week stalled approval as the typical way employee ideas die in growing companies
- Recommended delegation model: define independent decisions, boundaries and escalation triggers, then honor them
- Failed experiments should be distinguished from careless execution before any response
A two-week wait for a founder's approval can kill a good idea faster than any competitor can. That's the scenario ButterflyMX's founder lays out in a new essay on Entrepreneur: an employee proposes a small fix to a customer problem, the manager likes it, the founder's sign-off never arrives — and at the next company meeting, that same founder asks everyone to be more innovative.
The contradiction is common in growing businesses, he argues. Founders want initiative but keep approval processes built for a stage when they knew every customer and reviewed every expense. Employees adjust. They stop bringing ideas that take effort to get approved and stick to work they know they can finish.
"As founder of ButterflyMX, I've learned that a culture of innovation depends on what happens between someone noticing an opportunity and being able to act on it," he writes. Leaders shape that gap through the decisions they delegate, the time they protect and how they respond when an experiment disappoints.
Where do good ideas get stuck?
Before launching another creativity initiative, the author says, look at the last few employee suggestions. Did anyone have authority to move them forward? Was there time to do the work? Did the person who raised the idea ever hear why it was accepted, postponed or rejected?
His example: a support team keeps answering the same onboarding question. An employee proposes changing the customer instructions. The proposal bounces through departments, each with reasonable concerns, and nobody owns the final decision. The instructions stay the same.
Growth makes it worse. An approval that took five minutes when everyone sat around one table may now require three meetings and a week of waiting. His fix: trace a few stalled ideas through the business, look for repeated handoffs, unclear ownership and decisions that come to you out of habit. Removing one obstacle gives employees a concrete reason to bring the next idea.
What does usable autonomy actually look like?
"Use your judgment" sounds empowering. It fails when employees can't tell which decisions will trigger second-guessing, the author writes. Useful autonomy requires specificity.
His method: pick one recurring decision your team brings to you. Define what they can decide independently, which boundaries apply and what requires escalation — and explain the reasoning, so people can apply the rules to situations you haven't anticipated.
A support manager, for instance, could test revised onboarding instructions with a small customer group. The test stays within brand guidelines, preserves required disclosures and includes a rollback path. Within those limits, the manager proceeds without asking.
Oversight should match consequences. A limited, reversible test moves fast; changes involving customer commitments, sensitive data or significant spending get more review. And founders must honor what they delegate. Reopen a decision whenever you'd have chosen differently, and employees learn to check with you first.
How do you find the time?
Innovation time has to come from somewhere. If every hour is committed, improvement work gets pushed to the margins — something people attempt after finishing everything else.
The author's advice: ask a team which repetitive task eats time without judgment. Manually routing service requests, assembling recurring reports, entering the same data into multiple systems. Simplify, automate or stop it. Automating part of a request-routing process could free an employee to investigate why a particular request keeps appearing — which might produce clearer instructions, a product fix or a better onboarding flow.
Keep each experiment small. Define the problem, the change, the responsible person and the evidence of success. Set a review date. Leaders also have to say what can wait; if every existing commitment stays urgent, employees won't believe experimentation is real.
What happens when a test fails?
Employees watch closely what follows a disappointing result. A team tests a shorter onboarding process, expects faster customer ramp-up — and support requests rise instead, because customers miss information they need.
That's evidence about where simplification went too far, the author argues. A useful review asks: did the team stay within agreed boundaries, notice the problem promptly, and can it explain what to retain, revise or stop?
Leaders should distinguish a disappointing result from careless execution. Ignoring agreed limits or continuing past warning signs deserves a different response than a thoughtful test that disproves a hypothesis.
Don't reward raw activity either. A team that launches five experiments hasn't necessarily contributed more than a team that runs one and makes a valuable decision. Recognize clear reasoning, responsible execution and findings others can use — then share what changed, what was learned and what happens next.
That follow-through, the ButterflyMX founder concludes, is what gives employees the confidence to bring forward the next idea, knowing they can act on it.
Original: uk.entrepreneur.com
More from Nathan Brooks
Show full bio
News editor covering regulation, policy and their impact on entrepreneurship, startups and venture funding at Business Bearings.
2,352 articles