Starting a company is hard. But “failure” definitely isn’t one of the words I’d use to describe my experience.
Since letting friends and family know my co-founder and I were shutting down our startup, a lot of people have asked me what I’ve learned and what I would’ve done differently. So I thought it’d be helpful to write down some of my thoughts.
Dylan Sather and I spent the last year and a half building a suite of tools aimed at helping people get more stuff done at work. We built 4 apps that were used by thousands of people at more than 100 different companies. Though we didn’t set out with a goal of building 4 distinct applications, we iterated our way through a number of ideas until we landed on Rate That Meeting.
Rate That Meeting
Our vision for Rate That Meeting was clear — we wanted to help teams spend less time in unproductive meetings at work. To do this, we built a platform to facilitate anonymous feedback between meeting participants and organizers, along with a dashboard for team leaders to analyze the data necessary to assess the impact meetings have on their team’s productivity.
Deciding to shut down Rate That Meeting was a difficult decision for us both, but there were clear signals that we couldn’t ignore and problems we couldn’t resolve.
We offered vitamins instead of painkillers. When software is presented as a solution to a problem, users want a one-click fix. But improving meetings at work takes more than a couple clicks. The long-term solution requires buy-in from upper management and continuous self-reflection from the entire organization. Technology can certainly help here, but there’s no silver bullet — this took us longer to realize than it should have.
As a result, we had a lot of customers who liked Rate That Meeting but very few who loved it.
The core problem we never solved was securing conviction. Dylan and I decided not to seek capital from outside investors until we both had the confidence that Rate That Meeting solved the problem of unproductive meetings. Spoiler alert: we didn’t get there.
There were glimmers of things that worked well, but they weren’t strong enough and weren’t growing quickly enough. We were trying to fundamentally change meeting culture at companies, and that’s really hard to do.
Things that worked well
#1 — Ruthless prioritization
Since we hadn’t raised money (see above re: conviction as a requirement for raising money), we forced ourselves to set aggressive timelines and milestones for the company.
With every new feature we discussed, we repeatedly asked ourselves, “will this get us closer to our goal of conviction and money?” (“money” = funding or significant revenue) — and yes we actually asked each other that question out loud. If the answer was no we punted on it.
The process was a pain in the butt, but it was also an incredibly useful exercise. It forced us to really think hard about everything we were doing and whether it would directly contribute to our goals.
#2 — Intra-company virality
We built powerful hooks into our app that (1) increased the utility for existing users and (2) helped get more people to sign up. These hooks drove 56% of our total customer signups.
- Meeting feedback (35% of viral-based signups): RTM users could submit anonymous feedback for meetings they attended and we’d invite the meeting organizer to sign up in order to review that feedback.
- Meeting referrals (25% of viral-based signups): as a meeting participant, you could easily invite your peers to rate the same meetings you recently had with them, even if they weren’t signed up for RTM.
- Teams (25% of viral-based signups): we encouraged managers to create teams and invite their direct reports, in order to gain insights into productivity across their org.
- Standard CTAs (15% of viral-based signups): when RTM users organized their own meetings, we’d ask all the internal attendees to provide feedback after the meeting ended and would invite them to sign up after submitting feedback.
#3 — Product Hunt
We built Should It Be a Meeting as a weekend project to help accomplish a few goals — (1) market Rate That Meeting, (2) disseminate tips for having fewer meetings, and (3) have fun.
In order to get eyeballs on it we decided to launch it on Product Hunt in early May 2018. Even though we ignored most of what we read for ensuring a successful launch (we rushed it and didn’t build a big audience beforehand), we still ended up in the top 5 of all products for the day. We credit most of the success to the ease of use and the general nature of the app being free, not requiring users to create an account, and offering immediate value (mostly in the form of, “no you shouldn’t schedule a meeting”).
We saw nearly 10k visitors to the site in 2 days, which was way more than we had anticipated.
This had a material impact in driving additional traffic and signups for Rate That Meeting, even though users had to jump through a few hoops to get from Should It Be a Meeting to Rate That Meeting.
Important note — even though site visits and signups both increased after our Product Hunt launch, customer retention did not, which is why we still felt justified in shutting the app down even amid an uptick in new signups.
#1 — Prioritize finding conviction above everything else
Conviction should act as the guiding star during the early stages of product development. There were times when we lost sight of securing conviction in our product as our top priority, and that was a mistake.
This lapse in proper prioritization drove us to build features we were excited about but that didn’t move us closer to proving our hypotheses or finding product market fit.
We got caught in the trap of thinking that by signing up new users we’d be able to learn more about how they’re using the product, instead of staying focused on delighting our existing customers and turning them into passionate users — focusing on both at the same time didn’t work.
#2 — Either raise $$ or don’t (but make sure to choose one)
We wavered whether or not to raise money at all, especially in the first few months. We spent too much time debating if we could generate enough revenue to sustain the business or whether we should seek outside capital. As a result, our priorities shifted too many times, which distracted us from focusing on conviction and product market fit.
We finally decided we’d likely need to raise money in order to be successful. But we also decided we’d wait until we found conviction, since we didn’t want to ask investors to believe in us and our product if we weren’t yet convinced ourselves. This forced us to be scrappy. That scrappiness proved helpful and forced the ruthless prioritization I described above, but it sometimes forced us to be too scrappy.
Because we had set such aggressive timelines for ourselves we weren’t able to play the long game with most of the features we released — we’d push them out and if they didn’t have an immediate and positive impact we moved on to the next one.
And we drastically underestimated how slowly big companies move and how long sales cycles are for enterprise software. We just needed more time to move the needle with bigger clients, and our runway ran out.
#3 — Ignore bigger clients, latch onto small groups of customers
Maybe this is obvious to some people, but it wasn’t immediately obvious to us. Because we had built such viral hooks into our app we thought big companies presented the perfect environment for our app to grow and thrive.
Our viral hooks worked really well at some larger companies, but new users aren’t valuable if they don’t use and engage with the software. And even when growth isn’t slow, quantitative data alone rarely tells the full story.
Meaningful progress was slow with bigger companies — we attribute a large part of this to misaligned incentives. Every new company or user who signed up was always exciting for us, and we were motivated to move quickly. But giving us feedback on our product wasn’t the top priority for any of our customers (obviously), which meant they weren’t motivated to move as quickly as we were. This lack of urgency often times ended up grinding progress to a halt, which was frustrating.
Getting a feedback loop working with your customers is really hard and is incredibly important, and we found that this was a lot more difficult at bigger companies.
It’s also important to remember that big deals are probably over-hyped.
We learned a lot since we started working on MailCoach in 2016, we encouraged thousands of people to think more proactively about their workflows, and we had a great time.
We’ll undoubtedly take on our next venture with our eyes a bit wider and our skin a bit thicker.
Huge thanks to everyone in our respective networks that helped us along the way — especially Tod Sacerdoti, Pravin Savkar, Tom Schmidt, Mike McCullough, Luke Knepper, Monique Ho, Drake Ballew, Chris C Williams, wini tran, David Rabie, bruce falck, and Justin Smith.
— Danny Roosevelt (say hi at firstname.lastname@example.org)