Know When to Sunset a Project With Confidence

I killed a seven-figure revenue stream in 2019 because I couldn’t answer one question: would I start this today knowing what I know now?

That’s my threshold. Not profitability, not growth rate, not sunk cost. If the answer is no, you’re managing decline instead of building something. When I was running my fulfillment company, we had a white-glove assembly service that generated solid revenue but required constant firefighting. Every month I told myself we’d fix the unit economics. Every month we didn’t. The signal that forced my hand was when I realized I was scheduling around those client calls instead of for them. When you start avoiding something that makes money, that’s your gut telling you what the spreadsheet won’t.

The pivot versus sunset decision comes down to whether the problem is the execution or the premise. I’ve seen founders pivot three times trying to save a fundamentally flawed idea because they fell in love with their solution instead of the problem. With ShipDaddy before I built Fulfill.com, I had to ask whether brands actually wanted another fulfillment provider or if they wanted an easier way to find the right one. Turns out the market was screaming for the latter. That insight came from watching brands spend six months vetting 3PLs only to pick the wrong one anyway.

Here’s what actually works: give yourself a 90-day window with three specific metrics that would make you excited again. Not hopeful. Excited. If you hit month two and you’re already making excuses about why the metrics don’t matter, shut it down. I’ve never regretted killing something too early. I’ve definitely regretted letting things limp along for an extra year.

The confidence comes from accepting that your next thing will probably be better because you’re smarter now. Every project teaches you something even when it fails. My fulfillment company taught me that brands hate the 3PL search process. That lesson was worth millions.

Two thresholds keep me honest. First, a hard margin floor — not revenue, margin. Revenue can look healthy while a service line quietly bleeds senior time. When a line can’t hold its margin floor for two consecutive quarters, that’s not a warning sign, that’s the decision already made for you. Second, the opportunity-cost test: if my best people are spending hours on this instead of work that compounds into multi-year retainers, the project is borrowing from the future to look alive in the present. The early signal I watch isn’t margin itself — it’s rework rate and scope creep climbing while senior utilization on that line stays pinned high. Margin collapse is the lagging report; those two are the leading ones.

I learned this around 2014. We were still doing physical-build ADA consulting alongside digital. The physical jobs felt prestigious and clients were loyal, but margins were a third of our web audits and the sales cycle was double. I sunsetted that line over six months and routed everyone into digital. That call is why ADACP survived the 2010s as a bootstrapped firm competing against much larger consultancies — we stopped letting sentiment subsidize a service the math had already retired.

The signal most founders wait for before making this call is a lagging indicator. Revenue drops, user numbers decline, team morale suffers. By the time those signals are visible the decision has already been delayed past the point where a pivot still has enough runway to succeed. The threshold that has helped me make the call earlier and with more confidence is a leading indicator question: is the evidence that this project can work growing or shrinking with each passing month?

That question reframes the decision from a current performance evaluation to a trajectory evaluation. A project can be performing poorly right now and still be worth continuing if each iteration is producing clearer signal about what works. A project can be performing adequately right now and still be worth sunsetting if each iteration is producing less clarity rather than more.

When I was running MCOOK before founding Tibicle, the honest answer to that question became unavoidable at a specific point. Each iteration of the product was not producing sharper signal about the path to market fit. It was producing more questions without answers. The evidence that it could work was not growing. That was the threshold.

At Tibicle we apply the same question to client products during our consulting engagements. When a client is deciding whether to pivot a struggling product feature we ask them to map the last three iterations against one metric: did each one reduce the number of open questions about whether this works or increase them? If the open question count is rising rather than falling, the pivot conversation needs to happen regardless of how much has already been invested.

Sunk cost is the enemy of this decision. The only question worth asking is forward looking. Is the evidence growing that this works? If the honest answer is no, the call has already been made. You are just waiting for permission to act on it.

The signal I watch is whether the project still teaches me something. Revenue matters, obviously, but early-stage projects at Simply Noted have had stretches where the money was thin and the learning was massive. If a project is losing money and I’m not learning anything new from it, that’s the sunset signal.

We went through this with an early direct-to-consumer play. We launched a personal handwritten note service aimed at individuals, thinking there was a big market for people wanting to send birthday cards and thank you notes without the hassle. The concept made sense on paper. In practice, our customer acquisition cost was brutal because individuals don’t reorder the way businesses do. B2B clients send hundreds of notes a month on autopilot. Consumers send five cards at Christmas and disappear.

I gave it six months and tracked two things: repeat purchase rate and customer acquisition cost trend. When the repeat rate stayed flat and the acquisition cost kept climbing, I knew we weren’t approaching product-market fit. We weren’t getting closer. We sunset it and doubled down on B2B, which is now our entire business.

The threshold that works for me is straightforward: if three consecutive months show no improvement in the one metric that matters most for that project, it’s time to call it. You don’t need a complex framework. You need one honest number and the discipline to read it without making excuses.

Rick Elmore, Founder/CEO, Simply Noted (simplynoted.com)

When deciding whether to sunset or pivot an underperforming project, I usually look directly at user urgency versus our maintenance effort. At distribute, I spent my early days building out formal presentation decks and polite, automated re-engagement email sequences to handle user drop-off. It took constant tweaking to keep those systems running, and they rarely actually earned their keep in driving real renewals.

The clear signal that helped me confidently kill that initiative rather than just pivot it came during a specific account breakdown. A user’s outbound flow completely halted right before their renewal date because of a misconfigured plain-text trigger. Instead of letting our standard, polished sequence run, I just recorded a messy, unedited screen-share pointing directly at their broken n8n backend and dropped the video into a direct message. They got their distribution flowing again that afternoon and renewed two days later.

That gave me a hard threshold: issue-to-resolution time. When I saw that dropping a raw video cut our resolution time from days to a few hours, I realized tweaking or pivoting our formal corporate materials was a waste of time. We didn’t try to improve the polished emails. We sunsetted them entirely and shifted to live screen-shares.

I ask one question before anything else: if this ran perfectly—full capacity, zero waste, best-case staffing—would the unit economics actually work? If the answer is still no, that’s not a turnaround problem, that’s a math problem, and no amount of founder grit fixes it.

Temporarily struggling means the ceiling is fine but execution is off. Structurally broken means the ceiling itself is too low to cover fixed overhead. We’ve sunset offerings at Oakwell that guests genuinely loved because the contribution margin at full capacity still couldn’t carry the labor and space they consumed.

The bias I had to beat was sunk cost dressed up as loyalty to the original vision. My ISTEC finance training helps me separate the two: sentiment belongs in the guest experience, not the P&L. The internal debate with Jessica was real—she anchors brand, I anchor the numbers—but once we modeled the best case instead of defending the current case, the call got quieter.

At EV Cable Hub I had a product range that limped along for too long, an accessories line that sold the odd unit but tied up cash, shelf space and attention out of all proportion to what it returned. I run a small online shop selling EV charging cables, so a slow line is not a rounding error, it is money and focus I do not have to spare. The question I kept dodging was whether to kill it or change it.

The signal that finally made the call clear was demand versus effort, not sales alone. A thing can sell a little and still be worth dropping if it eats hours every week. So I started asking a sharper question: is the problem the product, or how we present it. If the searches and the interest are there but the page is doing a bad job, that is a pivot, you fix the listing and the photos and try again. If nobody is even looking, no search demand, no enquiries, that is a sunset, because better presentation cannot create want that does not exist.

That range was the second kind. The interest simply was not there, so I cleared it rather than pouring more work into rescuing it. The threshold I now use is that a line gets one honest pivot if the demand exists, and if it still does not earn its keep after that, it goes. Roughly 1 in 5 of the products I have tried over the years has been quietly retired this way. Cutting the dead ones is what gives the survivors the attention to grow, and a small business cannot afford sentiment about a product that customers were never asking for.

A project has to earn its keep in one of three ways: revenue, retention, or strategic learning. If it is not improving one of those, it becomes noise. The mistake is keeping a project alive because the team already spent time on it.

At Slickplan, the clearest signal is whether the project is creating repeatable value or just occasional interest. If customers use it once, ask for constant exceptions, or require too much support to see value, that is a warning sign. A pivot makes sense when the core problem is still real but the execution is wrong. Sunsetting makes sense when the problem is not important enough for customers to change behavior.

The threshold I use is simple: can we clearly explain who it helps, why they need it, and what behavior proves it matters? If we cannot answer that with real customer usage or demand, we stop investing. Good teams do not just ship. They also know when to clear space.

Client behavior often reveals the answer before internal dashboards do. Projects worth keeping tend to generate clearer communication, faster approvals, and steadier trust over time. Projects losing relevance usually create hesitation, repeated scope questioning, and more defensive conversations around value. In agency settings, that shift matters because once confidence weakens, delivery becomes slower, margins tighten, and even good execution feels harder to prove.

I watch for a consistent drop in decision velocity across two or three review cycles. When clients need more reassurance but show less conviction, the issue is usually deeper than packaging or positioning. A pivot makes sense only if the trust gap comes from something adjustable. If not, sunsetting protects both sides from prolonged inefficiency.

Not every project is meant to last forever and we’ve learned that the hard way. Since we handle different customized packaging projects at once, we regularly review what products, services, marketing efforts or processes are contributing to the business. If something consistently requires a lot of time, effort and resources but produces little to no meaningful return, that’s usually the first sign that we need to take a closer look and take action.

Before completely dropping something, we first ask whether it can be improved or repositioned. Sometimes the issue isn’t the product itself but how it’s being presented, marketed or priced. We’ve had different situations where adjusting the design, target audience, content strategy or offer produced better results than abandoning it. If there’s still demand or potential, we would be more likely to pivot than shut it down.

The clearest signal for us is when the project starts taking resources away from areas that are already performing well. Since we work with a small team in a hybrid setup, every asset, like time, budget and production slot, matters. If a project repeatedly consumes time that could be spent supporting other projects or existing clients and is costing us too much, that’s usually when we decide it’s no longer earning its place. At that point, it’s often better to redirect those resources toward opportunities that better support us.

Conor Keenan, Accredited Wealth Management Advisor® professional (AWMA®), Co-Founder of CompareAccounts. Has been helping Americans make better financial decisions over the past decade.

I tend to call these “zombie projects” – ideally, they can be avoided with a clear roadmap and expectations, but sometimes you make bets that don’t pay off.

We usually classify something as a “zombie project” when the cost of maintaining it outweighs the value we or our clients receive from it.

We then have a meeting to discuss ways to resuscitate the zombie project, and if we don’t think it can be saved, then we decide to scrap it. Afterwards we have a post-mortem to discuss what went right, what went wrong, and how we can avoid these types of projects in the future.

The qualitative metric we use for zombie projects are when people are no longer excited about working on it, and the quantitative measure we use is when the cost of maintaining it outweighs the value we or our clients receive from it.

This is a hard one to answer as it really depends on the specific circumstances of the project. I tend to look at it as if it’s no longer growing (and it’s not making any money) and you can’t think of more things you could do to try and grow it, that would be the time you might sunset it.

The reality is though there’s often an infinite number of things you can do, so you have to focus on things that would actually move the needle.

If there’s still plenty of things you can do that you haven’t yet done, you might think about pivoting at that point in time instead.

Whenever you drop a project, all the compounding that you’ve done up until that point gets wasted. Often it might be better to pivot, either slightly or majorly, rather than having to start the compounding again.

Think about what the project could be turned into. Is there a different goal you are working toward where tweaking that project could help you accomplish it? Most projects, unless entirely unrelated to anything else you are doing, can be pivoted in one way or another to serve a different purpose. Make sure that when you are making this decision, you’re including those directly involved with the project in the conversation.

Even when the numbers don’t look good on a project, I’ll often take the time to dig deeper before deciding to discontinue it. Yes, we want everything we do to ultimately be tied to generating revenue, but that can take the form of more than just sales or utilization numbers. If we have one very loyal client who is continuing to use a dated or inefficient model, we’ll often keep it in operation just for them. That relationship will bring us more value down the road.

We generally make these decisions collaboratively. It’s really important that you get multiple perspectives with these kinds of situations. Just because a project may seem like it’s no longer earning its keep from your specific situation, that doesn’t mean that perhaps another department feels as though it will be helpful or even vital for them in some way in the long run.

It depends on whether the project has any prospects. This is especially important for trending or hyped projects, where timing is everything. For all other businesses, it depends solely on the founder, because any business can be made profitable.