š When to hold 'em and when to fold 'em.
A year ago, I stared at a choice: pack it in or persist. Why I choose the latter, what I learned, and why it was the right thing, this time.
Even by startup standards, my past year has been improbable. 12 months ago, I had begun the process to shut down the company, had brought the āteamā just down to me, and I had a final decision to make: shut it down or give it one, last shot.
All the data said it was time to call it. Many who cared for me thought it was time to move on. Iād done it before - shut down a deeply meaningful venture that couldnāt scale in a space that desperately needed a solution to exist.
But this time something felt different. Like I was still just one little twist away from the pieces falling into place.
Dogged by a feeling that I was just missing some one little thing, I decided to give myself a last Hail Mary. With nothing to lose, I focused as simply as I could on the thing I wanted to exist in the world, for myself. And I gave myself 6 weeks - if, by Sept 30 I didnāt have a clear āsomethingā, I would shut down the company. It was a promise I made to myself, my husband, my investors.
So with 6 weeks secured, I got to work on product. No thinking about business models or scaling or technical details. Just make the simplest version of something I would find valuable for myself.
And that has always been: make it feel like I can just send everything family related to someone I trust, and theyāll take care if it all. Read through all the emails, get things to calendars, remind the right person on a Tuesday morning about the library books or the violin.
Take it all out of my mind and my inbox and help me manage it all.
So thatās what I did.
I posted on YCās Bookface, the best community I know to find curious/supportive yet highly skeptical/discerning users. In a few days I found myself with 5 or 6 onboarding calls. 3 families ended up giving it a shot - paying me $40/mo to be this person for them.
They would text/email me all the things, I would get it all back to them, using the surfaces they already used - SMS, email, Gcal/work cals. All via Front and Airtable. And because it was just me, I focused on the more central value with the largest impact. The stuff families would keep paying $40/month for.
Along the way I learned intimately the shape of this problem in a way that I hadnāt yet - what needs to happen with this email? How does one write the most succinct/info-packed title that returns as an SMS reminder? How does collaboration work between partners? Why is that feature a āgamechangerā?
Soon enough, through referrals and itty bitty word of mouth I was up to 10 families and though so small, it was clear I had more here than I ever had before. Still. It was clear this was far from scaleable.
And with a product with baby PMF, I turned to the next thing that could kill this: how to scale me?
From Sept to Jan - 4 short months, I asked this question over and over again (whatās the thing that will kill it next), until I:
found an incredible fellow mama to join the team to support me as the āhuman in the loopā and proved that I could hand it off
started playing around with OAIās playground and got glimmers of how this tech could have been the missing piece all along
with really nothing to lose, I shot a note to Sam (Altman) asking if I was understanding the potential correctly
in a true act of generosity, he not only responded, but connected me to his team to explore it further
found a temporary team to help me build with it further
was funded by OAI with a small group of other incredible founders and spent Dec and Jan having my mind regularly blown about what was going to be possible in mere months
found my dream co-founder
launched one of the first true āassistant/agent/copilotsā aimed at an everyday consumer use case
kept building and iterating until we got here - a product that, a small group of families find to be a āgame changer/lifesaverā, has incredible retention and has the economics to scale incredibly well
Like the protagonist in The Martian said:
You just begin. You do the math. You solve one problem... and you solve the next one... and then the next. And If you solve enough problems, you get to come home.
Of course this story is only just beginning and weāre no where near ācoming homeā. The technology is still so new that everyday we discover a new battle at the frontier of where probabilistic problem solving hits deterministic reliability. We have to scale this to the hundreds and thousands and millions.
But sitting here vs 12 short months ago? Iāll take it.
š¤ 12 lessons from 12 months.
At the risk of boiling a complicated year down to trite bullet points, I do think there are lessons Iāve learned that have enabled this highly improbable year to unfold as it has.
Build genuine relationships and be worthy of taking chances on. One of my favorite quotes is Jobsā: āYou can only connect the dots looking backwards.ā Itās led me to live a life more concerned about pursuing the right dots, than worrying about the lines or ladders to climb. Itās as true of people as it is jobs. When I was building Poppy, I was lucky enough to have partners like Sam Altman helping along the way. He saw first hand the problems Iāve been trying to tackle. I would like to think that being genuine and consistent in my endeavors helped make my long shot email find a recipient willing to help.
Thereās persistence and then thereās stubbornness, ego and fear. Itās a fine line. Itās incredibly hard to call it on a start-up. To āquitā something where persistence is a prerequisite. Funny enough, Annie Dukeās book Quit came out right as I was in the middle of all this. Turns out I was in complete agreement. It is my job to not only make the Koolaid but to drink it and hand it out to others. But it is also my job to know when that has to end. I set Sept 30 as the date and gave myself clear metrics. I put the odds at less than 2% but there it all was. I could take all of the emotions out of the everyday and just execute.
At the edge of belief, surround yourself with the few who believe but will hold you accountable. I canāt stress enough how the hardest part of this time last year was the fact that I was against the ropes. So few people believed that I could or should go on. I truly felt like the crazy one - seeing something that so few others did. It came down to a handful of people who believed in me. They didnāt have to see what I saw, they trusted that I did and that I had a plan. Just as importantly, they would have been the people to tell me to call it if I hadnāt hit the metrics and I had been unwilling to see it then.
Get going, get good. Something shipped today will always win over something marginally better shipped tomorrow. This little graphic got me through so many days when I wanted to just quit. I put it on my lock screen to remind me my job wasnāt to be great. It was to just get and keep going.
Build momentum by solving the next problem next. Once I had done the hardest part of getting going, I had to keep going. A bit artificial but I started sending a weekly update email to my small group of believers. Both to keep me honest, but to give me blinders on what I said I was going to do that week. If I dwelled on the 2%-ness of my endeavors, I would have never made it. When itās just you against the world, you need resolve in spades. And the way to fake it is to just commit and deliver over and over again.
I only focused on the āthing that would kill me next, if not solvedā. That meant product first. Then money. Then team. Then all over again. With the first glimmers of the product in place, I was running on financial fumes so turned to adding capital. Miraculously, OpenAI funded us and solved that. Then, I turned to: if I didnāt find the right technical partner, this would all be for naught. Having found her, we turned back to product, then $$, then team, as weāll continue doing.Ego kills more startups than anything else. It has been a very humbling couple of years. To the point that I now donāt put much store by the highs or the lows. To pull this off, Iāve had to confront my deepest fears of failure, to get down to manually doing things I never thought Iād have to, to approach everything with humility and gratitude. The ultimate truth is: I wanted this start-up to work more than I cared about my ego or my comfort or my ācareerā. Itās okay if that isnāt the calculus for others. But pulling off these kinds of last-ditch efforts requires a humility that is deeply painful and often not talked about.
The other part is: lots of people are surprised I brought on a cofounder (not just a CTO), with the commensurate equity. Many wondered why, after being the sole founder for so many years, on the precipice of things working and having the ability to hire and pay for a great team. Thatās all ego. Also the past. The only thing that (still) matters is the probability of ultimate success. And there is no question that finding a true partner who could do right by the technical side of this mission was and is my single biggest concern. I only have great gratitude to have found that person and that as tricky as itās been, itās also been incredibly fun to build over the past months.Be open to all the things - weird and wild. For most of 2022 I had been looking for some angle to crack this. The thing of the time was web3 and I loved falling deep into a new rabbithole that forced me to consider new mental models. Ultimately I realized it left more questions than answers for what I wanted to build, but just the act of curiosity, learning and exploration had my mind primed for AI. So when I started playing with LLMs, my mind was already primed to accept improbable things.
Be a cockroach until PMF: give yourself the practical chance to stay alive and pull off a Hail Mary. The only way last ditch efforts are possible are if you have the runway and cap table to actually do it. The most painful part of the past 4 years has been the team. Iāve essentially had to go through 2 whole versions of incredible teams. The very best kind of talented people who believe in the earliest days. Who took a chance on me just for me to have to had cut them loose. But if I hadnāt, when I had, I wouldnāt have had to capital to persist. At the same time, I didnāt want to raise ābridgeā financing because if and when something worked, I wanted to have the cap table flexibility to bring on a cofounder or the right capital which gets harder if you give away too much equity too early.
And letās be honest - I might have been one of the biggest benefactors of cosmic timing. If LLMs (specifically GPT4) hadnāt come along at exactly the moment they did, I also would not be here. One of the biggest points of this story is that I had to stay alive to see the day these capabilities came to market. Because too early is still wrong and I thank my lucky stars that this is very much a story of needing to have bridged mere weeks, not years. But still, that meant I had to make a pre-seed round last for years, not just months. If thereās one lesson people take away from all of this is: make whatever calls you need to conserve cash until you have PMF.Constraint fuels creativity. Contrary to popular belief, I think that constraint frees you to be the most creative in the most important areas. By definition you canāt entertain too many features, so you focus on the one or two that matter. But there, you have lots of leeway to do more interesting and weird and wonderful things. Also: itās quiet at the edges where constraints live. The place where you feel forgotten and afraid and alone. But out there on the edges, there is little distraction either. There is no one to impress, nothing to prove, nothing to lose. So itās actually quite fun again to think up weird little tests and try them out for a week or two. Enjoy that gift.
Surround yourself with kindness and grace no matter what. Because it usually is a 2% chance and chances are it wonāt actually work out like this. And if and when that happens, youāll need all the kindness and grace in the world. I know, because Iāve been there too. There is honor in calling it and calling it well. It wont feel like it at first, but trust me, itāll set up a bigger and better Next.
I suppose the ultimate lesson of these 12 months is: sometimes the universe gives you a one (or two). Then again, Iām a believer of: the harder I work, the luckier I get. So put yourself in the position to be lucky. Do the work. Try to minimize the decisions made by ego. Be curious. Explore weird.
And persist.
Wow - how amazing of you to put this out here. I so resonated with other articles you had on women / career. I am rooting for you!
What an amazing report! So glad to hear youāve found a groove, Avni. And itās great to read your writing again.