How to Know It’s Time to Pivot Your Startup

Stella Garber knows when it’s time to pivot—she’s done it. Recognizing that her company’s decision-making software was stuck with sluggish growth, she and her team made the difficult decision to halt its development and turn their attention to an AI-based task management app called Hoop. In this essay, the third in her series on business frameworks, Stella reflects on this decision and how she and her team figured out what’s next, and offers guidance to founders who sense they’re at a crossroads with their companies.—Kate Lee

Was this newsletter forwarded to you? Sign up to get it in your inbox.

Last fall, my team and I had a difficult decision to make. 

We had spent the nine months prior building an app that helped teams of people make better decisions faster at work. Employees could pose questions to one another and independently chime in to reach a decision without biasing each other with responses. The problem was that, while many individuals liked our app, they struggled to get their coworkers on board. As a result, we weren’t growing at the rate we needed, and as the CEO, I was worried that my own team was starting to lose morale.

In the spirit of practicing what we preached—making better decisions faster—we had to be honest with ourselves: It was time to pivot.

Fast forward a year. Our new product is called Hoop, an AI-first task manager. Many of the people we surveyed when we were building our initial product were overwhelmed by everything going on at work, between endless meetings and active Slack workspaces. They worried about missing important tasks and didn’t have consistent ways of staying on top of them. Hoop solves this by automatically capturing tasks across meetings, Slack, and email. You connect Hoop to your tools—your work email, calendar, and Slack—and tasks automatically populate in a to-do list where you can manage everything in one spot. No more missing tasks, writing things down frantically during meetings, or marking messages as “Unread” in Slack for later.

On September 10, Hoop had a successful ProductHunt launch, scoring as the top AI product of the day. Our customers say they can’t imagine going back to life before Hoop because they’re more focused, have less anxiety about missing important tasks, and are generally able to be more productive at work.

So how did we go from one extreme to the other—from stagnation to user love? How did we pivot with conviction, taking what we’d learned and accelerating our path to product-market fit in the next iteration of our product? And how were we able to leverage new technology, in this case AI, to solve age-old problems?

Here’s the story, along with a framework for other founders in the same situation as myself, who might be feeling similarly frustrated and unsure of how to proceed.

A framework for pivoting

It may seem like we knew exactly what we were doing, but we were figuring it out on the fly, doing the best we could given our collective experience. Luckily, that experience was relevant: Most of our distributed team of six had worked together previously across the product, design, engineering, and marketing teams at Trello, where we built and scaled one of the most popular task management apps.

We knew that in moments of uncertainty, figuring it out on the fly never involves guesswork, but begs for structure and organization. To guide our thinking, we developed a framework based on the following questions:

1. Is it time to pivot?
2. What have you learned so far that you don’t want to repeat?
3. What are the most compelling opportunities?
4. Which is the strongest option that everyone is excited to pursue?

Following this framework, along with the templates below, helped us stay calm and rational during what could have otherwise been a confusing and emotional moment for our startup. In the spirit of our longstanding commitment to thinking independently before diving into a big decision, we decided to do a lot of solo work ahead of an off-site gathering. We used that in-person meeting as an opportunity to make big decisions.

Let’s go step-by-step.

Step 1: Is it time to pivot?

Pivoting a startup is excruciating. It’s never clear exactly when the time is right to make such a monumental change. Many advice-givers say that you have to continue working on something for a long time before you know whether it’s any good. That’s true. But, at the same time, if you don’t feel a hint of product-market fit, your best move is to pull the plug sooner than later and move on.

Startups often fail not because they run out of money, but because the founding team runs out of motivation. The adage that startups are marathons, not sprints, is 100 percent true. As a result, you need to think about your momentum and pacing constantly, cultivating both even when they’re scarce through customer conversations, product launches, and fast feature development. We knew that once we made the decision to pivot, we couldn’t look back. We needed to be totally focused and confident in our new direction.

According to growth expert Sean Ellis, product-market fit is achieved when 40 percent of active users or customers say they’d be “very disappointed” if your product no longer existed. If you’re not close to that and your product isn’t growing, it might be time to consider a pivot. It was for us. For our product, our users were “slightly disappointed” most of the time.

In this moment, it’s easy to fall for the sunk-cost fallacy, believing that the killer feature is right around the corner. If your metrics are telling a different story and your users are telling you they can’t figure out how to fit your tool into their workflow, you should recognize what’s going on and act quickly. 

With our decision-making product, we were seeing a lack of adoption across a team. One person would be excited to try it, but that didn’t translate into enough energy to spread the behavior change to drive adoption in a group. We saw this play out dozens of times. After launching new features to try and mitigate this situation, we had to be honest about what was going on. Within our own team, we didn’t all agree that it was time to pivot at the same time, but over a couple of months, it became painfully clear that something had to change for the business to be successful.

Step 2: What have you learned so far that you don’t want to repeat?

After we made the decision to pivot, we felt it was important to take a moment to reflect on all that we had learned so far so we would not repeat the same mistakes. We wanted to move forward in a direction that benefited from all of our learning.

We asked ourselves:

• What should we prioritize in our next product?
• What should we leave behind?
• What have we learned from the previous business model?
• What are some must-haves—be it product, business, or otherwise—for the next iteration?

Everyone on the team completed a “Start, Stop, Continue” retrospective to answer these questions—something we regularly did when we worked at Trello and later Atlassian (which bought Trello in 2017 for $425 million). The exercise helped make sure teams were learning after each product launch. We each named items we wanted to “start” during the pivot—developing new processes and features, solving new problems, and reaching new target audiences. “Stop” was about things we wanted to leave behind, and “continue” was there to help us recognize what we were doing well and didn’t want to abandon completely.

Everyone added these items to a Notion page asynchronously over a period of a week before our in-person meeting. 

Source: All images courtesy of the author.

We ended up with a list of about 20 items in each column and looked for themes. 

A common one was that we needed to get our customers to derive real value from our app much more quickly. 

Step 3: What are some opportunities?

Rather than throw out random ideas for products we could build, we standardized the way we’d qualify ideas. We needed these ideas to start from an acute customer pain point and build into opportunities for a venture-scale business. In order to suggest an idea, the team had to answer a series of questions and create an “Opportunity Page” in Notion.

The questions we had to answer to submit an opportunity were as follows:

• What is the specific customer pain we are addressing?
• Who is the target audience?
• Why is this worthy of our time?
• How does the last three months inform this proposal?
• Why are you excited about this?

Then we all submitted at least three opportunities that we were excited about. Each opportunity had its own Notion page. In the end we had 21 potential opportunities to explore.

Notice that we were focusing not on solutions, but on problems—for example, notification overload and task management. This was our way of making sure we weren’t too fixated on fleeting ideas—“What if we built a human resources tool or collaboration platform?”—which could lead us down a rabbit hole without a strategic lens.

Step 4: Which is the strongest option that everyone is excited to pursue?

The next step was to independently score the 21 opportunities and discuss them as a group. We read through the Notion database of new ideas and completed an evaluation form for every opportunity, rating it from 1–10 for each of the following questions:

• How well supported is this pain point? How clear and specific is the target audience for this pain point?
• How reachable is the target audience? If you could improve this opportunity framing by changing, adding, or removing one thing about it, what would it be?
• Squint your eyes and imagine a well-executed solution to this market opportunity. How likely is that solution to lead to a successful business?

Here’s what our Google form looked like:

Bringing it all together

Research shows that independently gathering ideas is the best way to get divergent options ahead of a group discussion. So we did all of this prep work of writing Notion pages remotely and asynchronously, ahead of an off-site where we could sit together, analyze our independent thinking, and figure out the best path forward.

Meeting up in Scottsdale, Arizona, our team took stock of all of the potential opportunities and discussed our ratings. Before looking at the quantitative data, we talked about which opportunities we were most excited about. One option stood out from the beginning, but we pressed forward with our review anyway. We looked at the data and, as it turned out, we had all rated that same opportunity the highest.

While we were building our initial product, we heard a lot of the same themes from enthusiastic users who failed to get their teams to adopt our product. There were already too many tools to adopt another one, and too many tasks flying in from different directions—code for, “Your product isn’t useful enough."

Then, in November 2022, one moment changed everything: ChatGPT was released to the public. We had started using its API to build features, such as AI-generated summaries and suggestions, into our decision-making product. Everyone was excited about the possibility of building an AI-first product from scratch, and starting over was a chance to change the way we thought about building a product. I knew that generative AI was about to change the expectations customers had about software, and now we had an opportunity—and an excuse—to start with a blank slate.

How to do a ‘warm pivot’

Not surprisingly, it's common for startups to comprise teams that had previously worked together. They might start out building something completely different from their experience, but often will pivot into something that’s familiar for them. I’ve heard this referred to as a “warm pivot,” and that’s exactly what we found ourselves doing.

Four team members, including myself, had deep experience with project management from our time at Trello. The standout idea we were interested in examined how we might use generative AI to help people better manage all the tasks coming their way. Generative AI has the potential to change rote workflows in ways we couldn’t have dreamed of previously, and my head was spinning with possibilities.

Sitting around a couch in our Scottsdale Airbnb, we threw around the language of a “to-do list that writes itself.” We were tantalized by this idea. What if you didn’t have to write Trello cards or tickets, but instead, they simply appeared based on what you talked about in a meeting or in Slack messages? What if you never had to write down another task? That was the new idea, and we were ready to run with it.

On the second day of our off-site, we got more tactical about how to test our hypothesis. We wrote specifications for a minimum viable product and discussed how we could validate it as quickly as possible.

We also needed to prove demand before we built too much. While the engineers talked through the technical details of the product, I planned some ways to test demand and figure out how to position our product with only a landing page and some copy. 

We made a plan to build an MVP, test demand, and release something—anything—within about a month. To test the viability of our idea, we needed to see if we could quickly and reliably capture tasks from multiple sources. For meetings, this meant generating transcripts and building a system to recognize what a “task” is. Then this had to be repeated and customized for other sources of tasks, like Slack. And all of this had to be presented to users in a way that was delightful and easy to comprehend.

From start to finish, we had something out in about two months, with messaging that had been tested by buying traffic from Meta and presenting different ads.

There’s no shame in pivoting

Within a few months, our product was taking off. We were able to attract more users, and the pain we were addressing was resonant to a target audience we had identified: leaders, founders, and executives at growing companies.

More importantly, the validation we got at every step in our pivot helped us quickly see that not only did we make the right choice, but we made it in a way that gave us confidence and saved us time in the long term. What felt difficult previously—getting people interested, trying out the product, sharing it with others—became palpably easier. 

If you’re a founder who has been working on a product for a long time and something’s just not clicking, there’s no shame in pivoting. In fact, as I like to remind my team each day, the beauty of a startup is that you can change anything at any time. And with AI, there are seemingly daily updates and new capabilities to draw from. 

However, being systematic, thoughtful, and reflective—following tried-and-true processes and minimizing careless risks—will help you move with more conviction and confidence. 

To read more essays like this, subscribe to Every, and follow us on X at @every and on LinkedIn.

We also build AI tools for readers like you. Automate repeat writing with Spiral. Organize files automatically with Sparkle. Write something great with Lex.