Slack was originally a gaming app. Twitter started out as a podcast platform. Brex first pitched Y Combinator a VR headset.
These pivots, now, almost feel like startup campfire stories — we see the results, but what’s often obscured in the successes are the specific, painstaking moves founders made to find product-market fit.
We’ve had the opportunity to share dozens of lesser-known pivot stories here on The Review in our ongoing Paths to Product-Market Fit series, and in the process, we’ve developed a more granular definition of pivoting than simply “going in a different direction.” In our detailed essay breaking down the four levels of PMF, we offered our own take on the four Ps of marketing framework — four levers you can pull to get unstuck and jump to the next level:
- Problem: The fix your customers badly need (or don’t realize they need yet).
- Persona: Who’s buying your product, which can either be an individual decision maker (like a CTO) or a company profile (like a thousand-person manufacturing org).
- Promise: How you articulate your unique value proposition.
- Product: The solution that’ll deliver on your promise.
Many of the founders we’ve spoken to made pivots that might not sound terribly dramatic on paper, but they tuned the dials on one (or more) of the four Ps and soon after saw the first sparks of traction. Clay’s Kareem Amin originally went wide and juggled multiple personas before committing to a specific one. Jack Altman realized Lattice’s original persona had a much more pressing problem than what the company was solving. Jason Boehmig hit a wall trying to convince Ironclad’s persona that this was the right product for their problem — and found it was just a matter of tweaking the promise.
Here, we’ve curated our favorite pivot stories from The Review, applying our four Ps framework to spotlight how these founders climbed out of the pre-PMF trenches — lasering in on the exact aha moments and all the decisions and discoveries that led them there.
PROBLEM
Vanta hunted for solutions before asking what customers needed
Founders tend to fall into one of two camps: those who set out to solve a problem they’ve felt firsthand, and those who’ve always known they want to start something — long before they know what to solve.
Christina Cacioppo was the latter, and she ran into a classic entrepreneurial pitfall before stumbling on the idea for Vanta: building something no one wants.
After stints in VC and product management at Dropbox, Cacioppo knew she wanted to start her own company — so she quit her job and learned to code so she wouldn’t have to rely on a technical co-founder to build. She then scrapped together a series of products that failed to excite anyone.
This was back in 2016, so her first instinct was to ride the audio wave kicked off by Amazon Alexa. First, she built a meeting recorder that automatically transcribed all meetings across a company, and then a microphone that transcribed notes into Slack. But neither sparked much interest from startup buyers.
So Cacioppo turned her focus to a more niche use case: a voice assistant for biologists working in a lab. They do stuff with their hands, they wear gloves, they work with chemicals — how great would a notetaking tool be? She found a lab, shipped them a microphone, and even built them an iPad app. “They were thrilled because no one makes software for biologists in labs,” she says. “But the market for this was the size of my thumb and I didn’t even know anything about biologists. It’s funny to talk about now, but it was a true low point at the time.”
It was a bit like the children’s book, “Are You My Mother?” We had built this tool and then walked around to people asking, “Do you want our tool? Do you want our tool? What about you?” And the answer was no — no one wanted our tool.
After a few unsuccessful paths, Cacioppo resolved to stop building — and start talking to people. She turned to the startup folks she knew to pick their brains. “We decided we weren’t allowed to build anything at all. We had to just talk to people, and talk to them until we had a lot of confidence and a mental model of customers, their jobs, the problems they might have and how we might solve them.”
To find the most pressing problems, she turned to everyone’s most precious resource: time. So she asked people about their daily routines.“To find a good problem to solve, we really focused on what the customer’s day-to-day was like,” she says. “For discovery, the best thing we did was ask people to pull up their calendar. What were the best parts of those past weeks? What were the worst parts? That finally got us to a problem worth solving.”
One thing kept coming up over and over again in those calendar conversations. Startup leaders wished they could focus on security compliance and nab a SOC 2 certification — but could never find the time. That’s when the idea for Vanta crystallized: automate this process that people want to prioritize.
The moral of Cacioppo’s story: Talk to plenty of people to find a problem worth solving before writing the first lines of code.
Shippo realized how broken shipping was while exploring e-commerce marketplace ideas
Shippo co-founders Laura Behrens Wu and Simon Kreuz started a business so they could stay in San Francisco. The friends met in college back in Germany, both found their way to the Bay Area, and figured a startup was a good enough excuse to extend their tenancy. So they began ideating in an area they thought would be easiest: e-commerce (because Stripe and Shopify created a lower barrier to entry for small businesses selling online).
The first thing the co-founding duo built was a Shopify store that was supposed to be a marketplace showcasing independent designers. But kickstarting a brand new marketplace required too much initial inertia, so they decided to create demand by buying a few items from designers and selling them themselves.
After a few early orders, the founders became immersed in the ultra-complex world of shipping. They had to get the items they bought delivered to them from emerging designers, then ship those items outbound to their customers, all while managing inventory, shipping timelines, last mile problems — and more.
“Shipping was the least fun part of that experience,” says Behrens Wu. “We were learning as we were building. So we agreed to just build, get some customers and if we developed a passion for shipping along the way, we’d go all in.”
Once the duo decided to focus on shipping as a broad problem area, they toyed with a series of iterations before landing on the right solution. First was a search engine for buying shipping labels, but Behrens Wu says they realized that SMB customers were shipping too high of a volume per month to need that product. “If you ship many orders (like on a daily basis), clicking through this interface one by one was too tedious and time-consuming,” she says.
An aha moment arrived when drawing inspiration from Stripe’s API model. “The Stripe model is API first, and we thought that model could work for the shipping industry as well. So we pivoted to that approach — which we’re still doing today.”
The founders decided to build a dashboard on top of the API and connect that to the Shopify App Store, which was fairly new at the time, to drum up exposure with small businesses whose owners likely didn’t know how to integrate an API.
Shippo rode the rising wave of the Shopify App Store to gain its first few customers. “The numbers were small in comparison to where we are right now, but seeing that chart was incredible. It showed us that we were onto something, and this was working,” she says.
Lattice realized people leaders had a bigger problem than OKR planning
Lattice co-founder Jack Altman initially tried to tackle a problem he experienced firsthand at his previous company, Teespring: quarterly OKR planning was a painful process. “The execs get mad at each other. The employees are like, ‘Why are we doing this?’ Nobody adheres. It was in fact a real problem,” says Altman.
But it took pitching a solution for that problem to prospective customers to realize no one wanted to pay for it.
“The first thing we built was an OKR software tool. We had seen at our last company that quarterly OKR planning was a painful process for companies,” says Altman. “We validated that there was a real problem by talking to a bunch of other companies about how hard it is.”
He soon learned that OKR software was more of a nice-to-have for customers. “While it was in fact a real problem, we never built software that truly solved that problem,” he says. “There were two things that made that clear. One, it was really hard to get people to actually pull out their credit cards and pay us,” he says.
“The second issue was that once we did get people into the product — even entire teams where the whole company would do an OKR planning cycle in Lattice — the next quarter didn’t come easily. They were like, ‘Ugh, we’ve got to do this.’ And then by the third quarter they were like, ‘This is not happening naturally.’ For the employees, the retention was just not there. So both sides were not strong enough,” he says.
He took this feedback seriously and pivoted to a new problem before wasting more of the early Lattice team’s time working on the original problem. But the work wasn’t all for nothing. They’d already built strong relationships with HR leaders while building the OKR software, so they were able to pick their brains about other more pressing problems. Performance management rose to the top of the list.
“After nine or 10 months of working on OKRs, we got to a place where there wasn’t any line we could draw between what was happening and a business that was going to make any sense for the next round of funding by the time we ran out of money. So we ended up needing to pivot a couple of times before we landed on the thing that really worked, which was performance management.”
More often than not, if you’re working on something that is not getting great traction, you’re probably not a 10% adjustment away — you’re probably a 200% adjustment away.
After the pivot to performance management, the difference in traction was immediately clear. “It was very obvious what market pull looked like then. We had people paying us annual upfront contracts without ever touching a real product, just on our design mocks — which couldn’t be more different than the experience we’d been having with OKRs,” says Altman.
GOAT found a much better problem to solve from a serendipitous personal snafu
Perhaps no path to product-market fit is as full of winding turns as GOAT’s.
Co-founder Eddy Lu and his former roommate and co-founder Daishin Sugano tried all kinds of entrepreneurial endeavors to varying degrees of success before eventually landing on the idea for a sneaker marketplace. Their first few ideas? A cream puffs franchise and some of the first iPhone apps to hit the App Store (this was back in 2008). Nothing took off.
When they moved to Chicago from their home state of California with no friends in the city, they had their first compelling idea — why not build something to help them meet new people? GrubWithUs was born: a social dining app for sharing meals with strangers.
Lu and Sugano got into Y Combinator with GrubWithUs and successfully raised Seed and Series A funding. But they eventually ran into some insurmountable marketplace friction points. “People are thinking: What if I’m vegan, and there are no vegan options at the restaurant? What if the restaurant’s too far and people don’t want to travel out of their way? And then there’s the social aspect of having to socialize with strangers. So that social friction, plus all the other friction points, made it impossible for that marketplace to function,” says Lu.
After a few years, Lu and Sugano knew GrubWithUs wasn’t working, but they felt trapped on their current path by the sunk cost fallacy. “We tried to create a new startup using the existing restaurant data we had. But honestly, we were just trying to save face with our investors. If you had asked us if we were passionate about building this next idea, we would have said no. We were just trying to salvage the scraps that we had,” says Lu.
So the co-founders forged ahead anyway, and ended up wasting months building something they didn’t care deeply about. And sure enough, when the new company finally launched, it failed.
Amid these panicked attempts to pivot, there was a bit of bad luck that turned into an unexpected lightbulb moment: Sugano, a lifelong sneakerhead, received a fake pair of Air Jordans from eBay.
To Sugano and Lu, it was obvious that there was a need for a sneaker marketplace where everything was authenticated. “It raised a question in our minds: how come, in this day and age, we have hundreds of options for purchasing sneakers online but still have to worry whether something is real or fake?” says Lu. They knew how massive the resale market was and figured this had to be happening to other people.
Even though they were excited about the kernel of an idea (much more so than their previous venture), Lu and Sugano were nervous about presenting such a dramatic pivot to their investors. They first approached Greg Bettinelli, a partner at Upfront Ventures with a lot of marketplace experience.
After hearing their pitch, Bettinelli was convinced of the new idea and gave Lu and Sugano the green light to drastically change course. So the co-founders got the emotional buoy they needed to email the rest of their investors, who encouraged them to keep their remaining money and go for the pivot.
You need passion, a solvable problem, and luck: That’s the equation for pivoting.
PERSONA
Clay focused on GTM teams after building for everyone
Clay co-founder Kareem Amin spent six years working on a data enrichment product before figuring out who needed it most — and unlocking up-and-to-the-right growth numbers.
The original idea for Clay was incredibly broad and abstract: to make the power of programming accessible to more people. Amin and his co-founder Nicolae Rusan first built an initial product that scraped dozens of databases to pull information directly into a spreadsheet. But when they started selling it, they were overwhelmed by how vast their ICP could be.
Recruiters told Amin how useful their product was for finding candidates. Salespeople wanted to use the platform for their inbound and outbound efforts. Front-end engineers could use the spreadsheet as a low-code backend. They even had a customer who was using Clay to transform and send data to their accounting software.
So for a while, the Clay team avoided committing to just one ICP and sold wide, tailoring it to different customer profiles. “When you narrow the scope, it feels claustrophobic. Why are we doing something that’s smaller when we could be doing something bigger?” says Amin.
But they eventually realized this approach wasn’t netting customers who actually continued to use the product, and revenue growth stayed low. “People would feel excited by the possibilities the product opened. But they weren’t always going back and using it the next day,” says Amin. “So actually we had a lot of, ‘Wow, this is so powerful!’ and then no usage, or inconsistent usage.”
To gain traction, Clay fully committed to the ICP of outbound salespeople and only sold the same product each time — even if Amin didn’t feel complete confidence that was the right call. “It wasn’t so much that I was 100% sure that outbound was the right call, although it was a faster starting point because all companies need it, versus typically only bigger companies needing inbound support, ” says Amin. “It was more that I realized we need to pick one thing at a time, test it out clearly, and then make sure that it aligned with our larger goals — to get a lot of customers that can come in, self-serve and give us feedback that we can react to quickly. That’s when we’d earn the right to execute on the more expansive parts of our mission.”
This is the line of reasoning that helped us climb our way out of the use case spiral: How do we get customers as quickly as possible, so that we can learn as fast as possible, so that we can improve as quickly as possible?
Plaid took a bet on an enterprise solution after spinning off a consumer budgeting app
Plaid was an early entrant to the fintech space in 2012, the pioneer of the “invisible plumbing” fintech infrastructure that powers everything from paying your friends back on Venmo, to making an investment on Robinhood, to buying Bitcoin on Coinbase. But the bones of the product today were originally built for a consumer budgeting app.
“The response of consumers was such that we pretty quickly realized we weren’t going to be any good at building budgeting apps. We tried for a while. We built six different versions of budgeting management, analyzing your spend types of applications and they didn’t really ever get any traction,” says co-founder Zach Perret.
The idea to switch the target persona from consumers to businesses actually came from one of Perret’s friends. “One day, a friend who was one of the early engineers at Venmo came to us and said, ‘Hey, your consumer products are not so good, but I’d like to license the backend that you’ve built, the way that you’ve integrated with the banks in order to get the data into your app.’ It took awhile, but eventually we decided to make this pivot and shift over to building the platform,” he says.
“We then had five or seven people who said that they would probably use the product if we built it. We even had one who said they would pay us if we built it. And so it was pretty easy to find nascent product-market fit in that very early stage once we landed on the right idea,” says Perret.
But despite feeling some early pull, Perret wasn’t confident that the market was much bigger than those first few customers. “The bigger challenge for Plaid was that the market didn’t really exist. So while we had these five or seven companies that wanted us to build something, the real challenge was market development. That was the reason that almost every VC passed on investing in Plaid in the early stages. They said, ‘You’ve built an interesting product, it could be useful for some customers, but the market’s not very big.’”
With minimal validation in the new direction, Perret and the team had to take a leap of faith with the pivot. “When I think back to the earliest stage of Plaid’s product-market fit journey, the challenge wasn’t finding a product that people would buy. It was creating a market around the product that we believed would be very important for the future, but was yet to be proven.”
PROMISE
Ironclad sharpened its positioning after hanging out with enterprise legal teams
After serving as outside counsel for startups, Ironclad co-founder Jason Boehmig started tinkering with automating parts of his own practice with coding that he had taught himself along the way. After realizing that there was a gap both in new solutions for lawyers and a team that understood how they actually used software products, he decided to leave his law firm life behind.
“It’s wild to think back to this now that we’re in the AI assistant boom because the first version of Ironclad was an automated assistant for legal paperwork — but it was really me on the backend of an email address,” Boehmig says.
“We started by doing everything from corporate filings to NDAs and everything in between. So if you were doing an NDA, you would CC [email protected] and be like, ‘Hey, I’m doing an NDA with Joe Smith. Can you help us get it done?’ And I would send a series of responses as the Ironclad AI such as, ‘Hey, I need the following information,’ or ‘Here’s the template. I filled it in.’ I later met my co-founder, Cai, who was an amazing engineer, and he’d watch me work as the assistant and then would automate what could be automated away.”
But it was spending time in person with their initial customers that helped narrow their product’s focus. “I remember that we flew out the whole company — which was four or five people at the time — to Boulder to see a customer who had a bunch of product feedback for us. We all just sat around and watched them use the product,” Boehmig says.
“They had two monitors, one showing their email and on the other was Ironclad. They were just doing their day-to-day routine, working on a legal team. And by observing them and seeing how important these routine, repeatable transactions were to their profession, we realized we had to go all in on these corporate legal teams — particularly legal operations, which was a role that was just getting defined back in 2015.”
From there, Ironclad repositioned from an AI legal assistant that promised to automate lawyers’ tasks to a Contract Lifecycle Management platform that promised to help enterprise companies create and manage legal contracts end-to-end. That way, they could play in an existing category instead of trying to create one.
At Ironclad, the early “aha moments” came from watching our customers use the product. It enabled us to pick a niche audience that we could go after — legal operations — and the specific task they were doing — routine transactions.
PRODUCT
Alma moved a brick-and-mortar business online when the pandemic broke out
Harry Ritter had already found product-market fit for his original vision of a co-practicing space for therapists. In late 2018, Alma opened its Midtown Manhattan location, and then a second location in Downtown Brooklyn in 2019, and both were quickly packed to the brim.
Soon after, the pandemic happened. And just like that, 70% of Alma’s revenue disappeared in one weekend. In the year and a half after launching, Alma had found its market, but the market had changed almost overnight. With its main source of revenue gone, Alma had to pivot — and fast.
In those early days of the pandemic, Ritter saw a glimmer of opportunity: “If we’re all about to go into lockdown, and we’re about to experience all of the challenges that are about to come with this emerging pandemic, mental health is only going to matter more. And moreover, these small business owners who struggled to figure out how to run a business during normal times — how much more are they going to need us to show up for them now that things are getting even harder?”
But in the meantime, Ritter had to face his investors. “The first two were the worst board meetings I’ve ever had in my career as a CEO because we were trying to figure out what to do, and I wasn’t so sure,” he says.
So Ritter pored over Alma’s customer data and found a few interesting threads: a spike in providers joining Alma to access its virtual support services and insurance program, with promising early metrics around growth velocity, retention, close rates, time to sale, utilization and customer satisfaction.
From there, Ritter convinced investors at the next board meeting that facilitating providers’ shift to virtual care was the path forward. “That was the change moment where everybody in the room said, ‘All right. If you believe in it, you’re willing to do it, and you’ve got this data that seems to make sense, let’s see what’s going to happen.’”
There were bumps on the road ahead — the unstable economy in 2020 necessitated a reduction in force. But as business started to bounce back, Ritter eventually rebuilt the team by hiring go-to-market talent to grow the provider and client base and product and engineering folks to develop the tech. And his bet on virtual proved to be fertile ground for expansion: Though still New York City-based, Alma was able to open up applications for providers from across the country.
Rupa’s focus on root cause medicine didn’t change — but its product had to
Tara Viswanathan, co-founder and CEO of Rupa Health, had already fallen in love with a problem: giving more people access to root cause medicine to better understand their health concerns, instead of just managing symptoms. She just happened to build the wrong product to solve that problem first.
“When starting Rupa, I had conviction in the problem and where the market was going. I saw that the way we think about health, wellness, and medicine was changing. The first solution I started building was a ‘Zocdoc-style marketplace’ for holistic health providers,” says Viswanthan.
A year went by, and despite Viswanathan’s and co-founder Rosa Hamalainen’s iterations to the marketplace idea, it never netted any customer traction. “We tried everything to get this to work, from concierge, 1:1 text conversations to help people find their doctor, hosting in-person meetups called Rupa Circles, partnering with high-profile doctors at Stanford, Sutter Health, and One Medical, and numerous other product iterations,” she says.
Viswanathan says it was a combination of gut and logic that ultimately told her it was time to pivot the product. “My gut knew we were hitting a wall and this was not the right direction, I just needed my brain to catch up and understand why a marketplace wouldn’t work.”
In startups, there will be times you need to persist, and times you need to pivot — therein lies the founder’s dilemma. To get conviction on the right path, validate your gut feelings with logic and research.
It took a series of product pivots to eventually land on the lab testing platform. “What we discovered is that every product iteration we launched got us closer to the right answer. So the faster we could put products out into the world, the faster we’d get to something that works,” she says. “Insights and feedback from that first product helped us get to where we are today, but if I’d been rigid on sticking to that solution, we would have never gotten Rupa off the ground.”
After the marketplace, Viswanathan and Hamalainen tried their hand at a virtual clinic, doing the heavy lifting to operate it themselves. It was through building the clinic that they became intimately familiar with the hassles doctors face when operating their own clinic.
“After seeing just a handful of patients, it became obvious that labwork was the biggest problem,” says Viswanathan. “We were spending hours and hours figuring out where to order the tests, signing up doctors at different laboratories, determining pricing, and explaining instructions to patients. Without our help, doctors would have to spend hours on this themselves. One day my co-founder and I looked at each other and said, ‘This has to be it.’”
Immediately, Viswanathan fired off a series of texts to a few doctors in the Rupa network, asking if a portal that handled all labwork would be valuable. “At that point, we didn’t even know what a solution could look like, so we didn’t bother pitching a specific product. We focused on pitching the problem. Instantly there was a reaction we had not seen with the virtual clinic platform or the marketplace,” she says. “For all the other products we built, we were pitching it to the doctors. Now suddenly with one text message, it was the doctors convincing us to build and telling us exactly what was needed. Within minutes, we realized we were on to something.”
Slack was originally a gaming app. Twitter started out as a podcast platform. Brex first pitched Y Combinator a VR headset.
These pivots, now, almost feel like startup campfire stories — we see the results, but what’s often obscured in the successes are the specific, painstaking moves founders made to find product-market fit.
We’ve had the opportunity to share dozens of lesser-known pivot stories here on The Review in our ongoing Paths to Product-Market Fit series, and in the process, we’ve developed a more granular definition of pivoting than simply “going in a different direction.” In our detailed essay breaking down the four levels of PMF, we offered our own take on the four Ps of marketing framework — four levers you can pull to get unstuck and jump to the next level:
- Problem: The fix your customers badly need (or don’t realize they need yet).
- Persona: Who’s buying your product, which can either be an individual decision maker (like a CTO) or a company profile (like a thousand-person manufacturing org).
- Promise: How you articulate your unique value proposition.
- Product: The solution that’ll deliver on your promise.
Many of the founders we’ve spoken to made pivots that might not sound terribly dramatic on paper, but they tuned the dials on one (or more) of the four Ps and soon after saw the first sparks of traction. Clay’s Kareem Amin originally went wide and juggled multiple personas before committing to a specific one. Jack Altman realized Lattice’s original persona had a much more pressing problem than what the company was solving. Jason Boehmig hit a wall trying to convince Ironclad’s persona that this was the right product for their problem — and found it was just a matter of tweaking the promise.
Here, we’ve curated our favorite pivot stories from The Review, applying our four Ps framework to spotlight how these founders climbed out of the pre-PMF trenches — lasering in on the exact aha moments and all the decisions and discoveries that led them there.
PROBLEM
Vanta hunted for solutions before asking what customers needed
Founders tend to fall into one of two camps: those who set out to solve a problem they’ve felt firsthand, and those who’ve always known they want to start something — long before they know what to solve.
Christina Cacioppo was the latter, and she ran into a classic entrepreneurial pitfall before stumbling on the idea for Vanta: building something no one wants.
After stints in VC and product management at Dropbox, Cacioppo knew she wanted to start her own company — so she quit her job and learned to code so she wouldn’t have to rely on a technical co-founder to build. She then scrapped together a series of products that failed to excite anyone.
This was back in 2016, so her first instinct was to ride the audio wave kicked off by Amazon Alexa. First, she built a meeting recorder that automatically transcribed all meetings across a company, and then a microphone that transcribed notes into Slack. But neither sparked much interest from startup buyers.
So Cacioppo turned her focus to a more niche use case: a voice assistant for biologists working in a lab. They do stuff with their hands, they wear gloves, they work with chemicals — how great would a notetaking tool be? She found a lab, shipped them a microphone, and even built them an iPad app. “They were thrilled because no one makes software for biologists in labs,” she says. “But the market for this was the size of my thumb and I didn’t even know anything about biologists. It’s funny to talk about now, but it was a true low point at the time.”
It was a bit like the children’s book, “Are You My Mother?” We had built this tool and then walked around to people asking, “Do you want our tool? Do you want our tool? What about you?” And the answer was no — no one wanted our tool.
After a few unsuccessful paths, Cacioppo resolved to stop building — and start talking to people. She turned to the startup folks she knew to pick their brains. “We decided we weren’t allowed to build anything at all. We had to just talk to people, and talk to them until we had a lot of confidence and a mental model of customers, their jobs, the problems they might have and how we might solve them.”
To find the most pressing problems, she turned to everyone’s most precious resource: time. So she asked people about their daily routines.“To find a good problem to solve, we really focused on what the customer’s day-to-day was like,” she says. “For discovery, the best thing we did was ask people to pull up their calendar. What were the best parts of those past weeks? What were the worst parts? That finally got us to a problem worth solving.”
One thing kept coming up over and over again in those calendar conversations. Startup leaders wished they could focus on security compliance and nab a SOC 2 certification — but could never find the time. That’s when the idea for Vanta crystallized: automate this process that people want to prioritize.
The moral of Cacioppo’s story: Talk to plenty of people to find a problem worth solving before writing the first lines of code.
Shippo realized how broken shipping was while exploring e-commerce marketplace ideas
Shippo co-founders Laura Behrens Wu and Simon Kreuz started a business so they could stay in San Francisco. The friends met in college back in Germany, both found their way to the Bay Area, and figured a startup was a good enough excuse to extend their tenancy. So they began ideating in an area they thought would be easiest: e-commerce (because Stripe and Shopify created a lower barrier to entry for small businesses selling online).
The first thing the co-founding duo built was a Shopify store that was supposed to be a marketplace showcasing independent designers. But kickstarting a brand new marketplace required too much initial inertia, so they decided to create demand by buying a few items from designers and selling them themselves.
After a few early orders, the founders became immersed in the ultra-complex world of shipping. They had to get the items they bought delivered to them from emerging designers, then ship those items outbound to their customers, all while managing inventory, shipping timelines, last mile problems — and more.
“Shipping was the least fun part of that experience,” says Behrens Wu. “We were learning as we were building. So we agreed to just build, get some customers and if we developed a passion for shipping along the way, we’d go all in.”
Once the duo decided to focus on shipping as a broad problem area, they toyed with a series of iterations before landing on the right solution. First was a search engine for buying shipping labels, but Behrens Wu says they realized that SMB customers were shipping too high of a volume per month to need that product. “If you ship many orders (like on a daily basis), clicking through this interface one by one was too tedious and time-consuming,” she says.
An aha moment arrived when drawing inspiration from Stripe’s API model. “The Stripe model is API first, and we thought that model could work for the shipping industry as well. So we pivoted to that approach — which we’re still doing today.”
The founders decided to build a dashboard on top of the API and connect that to the Shopify App Store, which was fairly new at the time, to drum up exposure with small businesses whose owners likely didn’t know how to integrate an API.
Shippo rode the rising wave of the Shopify App Store to gain its first few customers. “The numbers were small in comparison to where we are right now, but seeing that chart was incredible. It showed us that we were onto something, and this was working,” she says.
Lattice realized people leaders had a bigger problem than OKR planning
Lattice co-founder Jack Altman initially tried to tackle a problem he experienced firsthand at his previous company, Teespring: quarterly OKR planning was a painful process. “The execs get mad at each other. The employees are like, ‘Why are we doing this?’ Nobody adheres. It was in fact a real problem,” says Altman.
But it took pitching a solution for that problem to prospective customers to realize no one wanted to pay for it.
“The first thing we built was an OKR software tool. We had seen at our last company that quarterly OKR planning was a painful process for companies,” says Altman. “We validated that there was a real problem by talking to a bunch of other companies about how hard it is.”
He soon learned that OKR software was more of a nice-to-have for customers. “While it was in fact a real problem, we never built software that truly solved that problem,” he says. “There were two things that made that clear. One, it was really hard to get people to actually pull out their credit cards and pay us,” he says.
“The second issue was that once we did get people into the product — even entire teams where the whole company would do an OKR planning cycle in Lattice — the next quarter didn’t come easily. They were like, ‘Ugh, we’ve got to do this.’ And then by the third quarter they were like, ‘This is not happening naturally.’ For the employees, the retention was just not there. So both sides were not strong enough,” he says.
He took this feedback seriously and pivoted to a new problem before wasting more of the early Lattice team’s time working on the original problem. But the work wasn’t all for nothing. They’d already built strong relationships with HR leaders while building the OKR software, so they were able to pick their brains about other more pressing problems. Performance management rose to the top of the list.
“After nine or 10 months of working on OKRs, we got to a place where there wasn’t any line we could draw between what was happening and a business that was going to make any sense for the next round of funding by the time we ran out of money. So we ended up needing to pivot a couple of times before we landed on the thing that really worked, which was performance management.”
More often than not, if you’re working on something that is not getting great traction, you’re probably not a 10% adjustment away — you’re probably a 200% adjustment away.
After the pivot to performance management, the difference in traction was immediately clear. “It was very obvious what market pull looked like then. We had people paying us annual upfront contracts without ever touching a real product, just on our design mocks — which couldn’t be more different than the experience we’d been having with OKRs,” says Altman.
GOAT found a much better problem to solve from a serendipitous personal snafu
Perhaps no path to product-market fit is as full of winding turns as GOAT’s.
Co-founder Eddy Lu and his former roommate and co-founder Daishin Sugano tried all kinds of entrepreneurial endeavors to varying degrees of success before eventually landing on the idea for a sneaker marketplace. Their first few ideas? A cream puffs franchise and some of the first iPhone apps to hit the App Store (this was back in 2008). Nothing took off.
When they moved to Chicago from their home state of California with no friends in the city, they had their first compelling idea — why not build something to help them meet new people? GrubWithUs was born: a social dining app for sharing meals with strangers.
Lu and Sugano got into Y Combinator with GrubWithUs and successfully raised Seed and Series A funding. But they eventually ran into some insurmountable marketplace friction points. “People are thinking: What if I’m vegan, and there are no vegan options at the restaurant? What if the restaurant’s too far and people don’t want to travel out of their way? And then there’s the social aspect of having to socialize with strangers. So that social friction, plus all the other friction points, made it impossible for that marketplace to function,” says Lu.
After a few years, Lu and Sugano knew GrubWithUs wasn’t working, but they felt trapped on their current path by the sunk cost fallacy. “We tried to create a new startup using the existing restaurant data we had. But honestly, we were just trying to save face with our investors. If you had asked us if we were passionate about building this next idea, we would have said no. We were just trying to salvage the scraps that we had,” says Lu.
So the co-founders forged ahead anyway, and ended up wasting months building something they didn’t care deeply about. And sure enough, when the new company finally launched, it failed.
Amid these panicked attempts to pivot, there was a bit of bad luck that turned into an unexpected lightbulb moment: Sugano, a lifelong sneakerhead, received a fake pair of Air Jordans from eBay.
To Sugano and Lu, it was obvious that there was a need for a sneaker marketplace where everything was authenticated. “It raised a question in our minds: how come, in this day and age, we have hundreds of options for purchasing sneakers online but still have to worry whether something is real or fake?” says Lu. They knew how massive the resale market was and figured this had to be happening to other people.
Even though they were excited about the kernel of an idea (much more so than their previous venture), Lu and Sugano were nervous about presenting such a dramatic pivot to their investors. They first approached Greg Bettinelli, a partner at Upfront Ventures with a lot of marketplace experience.
After hearing their pitch, Bettinelli was convinced of the new idea and gave Lu and Sugano the green light to drastically change course. So the co-founders got the emotional buoy they needed to email the rest of their investors, who encouraged them to keep their remaining money and go for the pivot.
You need passion, a solvable problem, and luck: That’s the equation for pivoting.
PERSONA
Clay focused on GTM teams after building for everyone
Clay co-founder Kareem Amin spent six years working on a data enrichment product before figuring out who needed it most — and unlocking up-and-to-the-right growth numbers.
The original idea for Clay was incredibly broad and abstract: to make the power of programming accessible to more people. Amin and his co-founder Nicolae Rusan first built an initial product that scraped dozens of databases to pull information directly into a spreadsheet. But when they started selling it, they were overwhelmed by how vast their ICP could be.
Recruiters told Amin how useful their product was for finding candidates. Salespeople wanted to use the platform for their inbound and outbound efforts. Front-end engineers could use the spreadsheet as a low-code backend. They even had a customer who was using Clay to transform and send data to their accounting software.
So for a while, the Clay team avoided committing to just one ICP and sold wide, tailoring it to different customer profiles. “When you narrow the scope, it feels claustrophobic. Why are we doing something that’s smaller when we could be doing something bigger?” says Amin.
But they eventually realized this approach wasn’t netting customers who actually continued to use the product, and revenue growth stayed low. “People would feel excited by the possibilities the product opened. But they weren’t always going back and using it the next day,” says Amin. “So actually we had a lot of, ‘Wow, this is so powerful!’ and then no usage, or inconsistent usage.”
To gain traction, Clay fully committed to the ICP of outbound salespeople and only sold the same product each time — even if Amin didn’t feel complete confidence that was the right call. “It wasn’t so much that I was 100% sure that outbound was the right call, although it was a faster starting point because all companies need it, versus typically only bigger companies needing inbound support, ” says Amin. “It was more that I realized we need to pick one thing at a time, test it out clearly, and then make sure that it aligned with our larger goals — to get a lot of customers that can come in, self-serve and give us feedback that we can react to quickly. That’s when we’d earn the right to execute on the more expansive parts of our mission.”
This is the line of reasoning that helped us climb our way out of the use case spiral: How do we get customers as quickly as possible, so that we can learn as fast as possible, so that we can improve as quickly as possible?
Plaid took a bet on an enterprise solution after spinning off a consumer budgeting app
Plaid was an early entrant to the fintech space in 2012, the pioneer of the “invisible plumbing” fintech infrastructure that powers everything from paying your friends back on Venmo, to making an investment on Robinhood, to buying Bitcoin on Coinbase. But the bones of the product today were originally built for a consumer budgeting app.
“The response of consumers was such that we pretty quickly realized we weren’t going to be any good at building budgeting apps. We tried for a while. We built six different versions of budgeting management, analyzing your spend types of applications and they didn’t really ever get any traction,” says co-founder Zach Perret.
The idea to switch the target persona from consumers to businesses actually came from one of Perret’s friends. “One day, a friend who was one of the early engineers at Venmo came to us and said, ‘Hey, your consumer products are not so good, but I’d like to license the backend that you’ve built, the way that you’ve integrated with the banks in order to get the data into your app.’ It took awhile, but eventually we decided to make this pivot and shift over to building the platform,” he says.
“We then had five or seven people who said that they would probably use the product if we built it. We even had one who said they would pay us if we built it. And so it was pretty easy to find nascent product-market fit in that very early stage once we landed on the right idea,” says Perret.
But despite feeling some early pull, Perret wasn’t confident that the market was much bigger than those first few customers. “The bigger challenge for Plaid was that the market didn’t really exist. So while we had these five or seven companies that wanted us to build something, the real challenge was market development. That was the reason that almost every VC passed on investing in Plaid in the early stages. They said, ‘You’ve built an interesting product, it could be useful for some customers, but the market’s not very big.’”
With minimal validation in the new direction, Perret and the team had to take a leap of faith with the pivot. “When I think back to the earliest stage of Plaid’s product-market fit journey, the challenge wasn’t finding a product that people would buy. It was creating a market around the product that we believed would be very important for the future, but was yet to be proven.”
PROMISE
Ironclad sharpened its positioning after hanging out with enterprise legal teams
After serving as outside counsel for startups, Ironclad co-founder Jason Boehmig started tinkering with automating parts of his own practice with coding that he had taught himself along the way. After realizing that there was a gap both in new solutions for lawyers and a team that understood how they actually used software products, he decided to leave his law firm life behind.
“It’s wild to think back to this now that we’re in the AI assistant boom because the first version of Ironclad was an automated assistant for legal paperwork — but it was really me on the backend of an email address,” Boehmig says.
“We started by doing everything from corporate filings to NDAs and everything in between. So if you were doing an NDA, you would CC [email protected] and be like, ‘Hey, I’m doing an NDA with Joe Smith. Can you help us get it done?’ And I would send a series of responses as the Ironclad AI such as, ‘Hey, I need the following information,’ or ‘Here’s the template. I filled it in.’ I later met my co-founder, Cai, who was an amazing engineer, and he’d watch me work as the assistant and then would automate what could be automated away.”
But it was spending time in person with their initial customers that helped narrow their product’s focus. “I remember that we flew out the whole company — which was four or five people at the time — to Boulder to see a customer who had a bunch of product feedback for us. We all just sat around and watched them use the product,” Boehmig says.
“They had two monitors, one showing their email and on the other was Ironclad. They were just doing their day-to-day routine, working on a legal team. And by observing them and seeing how important these routine, repeatable transactions were to their profession, we realized we had to go all in on these corporate legal teams — particularly legal operations, which was a role that was just getting defined back in 2015.”
From there, Ironclad repositioned from an AI legal assistant that promised to automate lawyers’ tasks to a Contract Lifecycle Management platform that promised to help enterprise companies create and manage legal contracts end-to-end. That way, they could play in an existing category instead of trying to create one.
At Ironclad, the early “aha moments” came from watching our customers use the product. It enabled us to pick a niche audience that we could go after — legal operations — and the specific task they were doing — routine transactions.
PRODUCT
Alma moved a brick-and-mortar business online when the pandemic broke out
Harry Ritter had already found product-market fit for his original vision of a co-practicing space for therapists. In late 2018, Alma opened its Midtown Manhattan location, and then a second location in Downtown Brooklyn in 2019, and both were quickly packed to the brim.
Soon after, the pandemic happened. And just like that, 70% of Alma’s revenue disappeared in one weekend. In the year and a half after launching, Alma had found its market, but the market had changed almost overnight. With its main source of revenue gone, Alma had to pivot — and fast.
In those early days of the pandemic, Ritter saw a glimmer of opportunity: “If we’re all about to go into lockdown, and we’re about to experience all of the challenges that are about to come with this emerging pandemic, mental health is only going to matter more. And moreover, these small business owners who struggled to figure out how to run a business during normal times — how much more are they going to need us to show up for them now that things are getting even harder?”
But in the meantime, Ritter had to face his investors. “The first two were the worst board meetings I’ve ever had in my career as a CEO because we were trying to figure out what to do, and I wasn’t so sure,” he says.
So Ritter pored over Alma’s customer data and found a few interesting threads: a spike in providers joining Alma to access its virtual support services and insurance program, with promising early metrics around growth velocity, retention, close rates, time to sale, utilization and customer satisfaction.
From there, Ritter convinced investors at the next board meeting that facilitating providers’ shift to virtual care was the path forward. “That was the change moment where everybody in the room said, ‘All right. If you believe in it, you’re willing to do it, and you’ve got this data that seems to make sense, let’s see what’s going to happen.’”
There were bumps on the road ahead — the unstable economy in 2020 necessitated a reduction in force. But as business started to bounce back, Ritter eventually rebuilt the team by hiring go-to-market talent to grow the provider and client base and product and engineering folks to develop the tech. And his bet on virtual proved to be fertile ground for expansion: Though still New York City-based, Alma was able to open up applications for providers from across the country.
Rupa’s focus on root cause medicine didn’t change — but its product had to
Tara Viswanathan, co-founder and CEO of Rupa Health, had already fallen in love with a problem: giving more people access to root cause medicine to better understand their health concerns, instead of just managing symptoms. She just happened to build the wrong product to solve that problem first.
“When starting Rupa, I had conviction in the problem and where the market was going. I saw that the way we think about health, wellness, and medicine was changing. The first solution I started building was a ‘Zocdoc-style marketplace’ for holistic health providers,” says Viswanthan.
A year went by, and despite Viswanathan’s and co-founder Rosa Hamalainen’s iterations to the marketplace idea, it never netted any customer traction. “We tried everything to get this to work, from concierge, 1:1 text conversations to help people find their doctor, hosting in-person meetups called Rupa Circles, partnering with high-profile doctors at Stanford, Sutter Health, and One Medical, and numerous other product iterations,” she says.
Viswanathan says it was a combination of gut and logic that ultimately told her it was time to pivot the product. “My gut knew we were hitting a wall and this was not the right direction, I just needed my brain to catch up and understand why a marketplace wouldn’t work.”
In startups, there will be times you need to persist, and times you need to pivot — therein lies the founder’s dilemma. To get conviction on the right path, validate your gut feelings with logic and research.
It took a series of product pivots to eventually land on the lab testing platform. “What we discovered is that every product iteration we launched got us closer to the right answer. So the faster we could put products out into the world, the faster we’d get to something that works,” she says. “Insights and feedback from that first product helped us get to where we are today, but if I’d been rigid on sticking to that solution, we would have never gotten Rupa off the ground.”
After the marketplace, Viswanathan and Hamalainen tried their hand at a virtual clinic, doing the heavy lifting to operate it themselves. It was through building the clinic that they became intimately familiar with the hassles doctors face when operating their own clinic.
“After seeing just a handful of patients, it became obvious that labwork was the biggest problem,” says Viswanathan. “We were spending hours and hours figuring out where to order the tests, signing up doctors at different laboratories, determining pricing, and explaining instructions to patients. Without our help, doctors would have to spend hours on this themselves. One day my co-founder and I looked at each other and said, ‘This has to be it.’”
Immediately, Viswanathan fired off a series of texts to a few doctors in the Rupa network, asking if a portal that handled all labwork would be valuable. “At that point, we didn’t even know what a solution could look like, so we didn’t bother pitching a specific product. We focused on pitching the problem. Instantly there was a reaction we had not seen with the virtual clinic platform or the marketplace,” she says. “For all the other products we built, we were pitching it to the doctors. Now suddenly with one text message, it was the doctors convincing us to build and telling us exactly what was needed. Within minutes, we realized we were on to something.”
