The perils of a huge pre-PMF Series A: My story at Element AI
Philippe Beaudoin
17 replies
It's the first time I tell this story. ProductHunt felt like the right place. I'll tell you why later.
In 2016, I co-founded Element AI. In 2017 we raised $100M. It went weird.
Peak Zeitgeist
We hit peak zeitgeist. AI was super hot. Many investors believed that early AI would be a talent game. Montreal — where Element AI was headquartered — was seen as one of the largest untapped beds of AI talent.
In the months between December 2016 (our seed) and June 2017 (our Series A), we had executed super well on the talent game. We had signed a lot of big logos who wanted to undergo an AI transformation. We had built a world-class network of academics and managed to actively engage them on consulting engagement with these big customers. More importantly, we did it with enough collegiality that researchers enjoyed the time they spent discussing with us — and amongst themselves — around our customer's problems.
Talent Tsunami
We managed to translate researcher engagement into hiring momentum. Very talented AI researchers and AI engineers heard of the exciting discussions that were going on between top researchers in our meeting rooms. Some of them were stuck in data science jobs, some were at Google or Facebook but dreamt of having more impact.
We brought them in.
At that point, any C-level of any large corporation who walked into Element AI went into jaw-drop mode. They saw equations doodled on the windows in the kitchen, they heard researchers discuss the latest deep learning result. They saw the kind of energy they dreamt of having in-house but that they just couldn't get.
We closed them.
We became the fastest growing startup in Canada.
Customers Everywhere
We had so many marquee customers. I gave talks all around the world. I found myself on a stage next to Andrew Ng, only him and I, addressing a room of hundreds of Hyundai executives. That's where I learned that jokes don't survive live translation.
We had an exceptional team of consultants. Ex-McKinsey. Ex-Deloitte. They looked top-notch in meetings. They could wow our customers by bringing in the brightest AI minds at a moment's notice. We delivered.
We opened offices in Toronto, in London, in Singapore, in Seoul.
Too Much of a Good Thing
Then there were the products. We had a lot of great initiatives. We built infra to help our researchers develop and train their models more efficiently. We built a solid information platform using top-notch NLP, a top-of-the-line visual text processing tool, a product using AI to help manage user permissions in large organizations… Lots of potentially very useful products.
But the problem was that none of these products had the traction our consulting group had. We tried to align our consulting outreach with the products we were building, but there was always this new customer that was too big to pass and that we just had to accept.
We had incredibly talented and driven people. People who wanted to bring value to the organization. Only, that value came mostly from consulting. This led to internal tensions. On one side you had those who believed success would come from products and asked for more freedom to build them. On the other, you had those who wanted to bring in revenue and asked for engineers and researchers to work on consulting engagements.
The Exit
That tension became really hard to manage. After 4 years, it's the kind of thing you could feel in the air, just walking through our gorgeous headquarters in the center of Montreal's Quartier de l'IA.
On March 13th 2020, the day the world shut-down, I caught the last flight out of London Heathrow. This was my last business trip with Element AI.
I had decided to follow my dream and found Waverly. A product-centric startup. Yes, yes, AI. Yes, deep tech. Researchers, all of that. But first: a product. First, value that can scale. Real traction. Something small that could grow, grow, and grow.
A year later Element AI was sold to Service Now.
My Takeaway
I loved my journey at Element AI. I loved my coworkers. This was the best team I ever worked with. They had the brains of my Google coworkers and the grit of entrepreneurs.
As founders, I think we took a reasonable bet. I still think that talent was key. I believe we executed it well. It was bold, and being bold was not our mistake.
Our mistake was not seeing the tensions that could come from having too much money. Money gave us the ability to bring in the best customers, but it did not pressure us into building products. Building products and despairing because they're not finding traction. Building products and testing them, pivoting them, validating them…
That's why I'm posting this on ProductHunt. Because you understand what it means to relentlessly push and pivot a product.
That's what I want to do now, with Waverly, and I know you can help me.
Replies
Jerome Pasquero@iknowjerome
Phil,
thanks for writing this piece. I really appreciated your candor and how you clarify that, even as a co-founder, you only had access to part of the whole picture of what was going on. It’s with that same caveat, limited access, that I will push the product reflection one step further. As you very well know, I was also part of the Element AI adventure, though not as a founder; I joined later than you did. In August 2017, I was hired as a Project Manager because the title Product Manager didn’t even exist yet at Element AI. At first, I was tasked with kickstarting a few of the core applied research teams composed of that unique AI talent that you mention. During my 3 1/2 year stint at Element AI, I went on to play many other roles, including the one of Product Manager.
All of my different functions got me to interact with driven, hard-working and knowledgeable people who were all smarter than I could have ever dreamed off in a professional setting. They also happened to be nice and genuine people. The buzz that you describe was real, and I loved every minute of it. The tension between product and consulting was also very palpable, but I think it’s important to provide a little more detail about the source of that tension. Collectively, we were all riding a hype that made us miss some red flags that, in hindsight, are easy to identify. The consulting and sales functions had never been exposed to AI and deep learning. While deep learning is not everything that we explored, it was an important part of the big bet we were making. Outside the tech giants such as Google, Amazon and Facebook, deep learning was just making its way to the corporate world and that meant that it was very poorly understood by people in industry. For instance, people who were deeply knowledgeable about their own industry could, at times, struggle with understanding what deep learning could and could not do. Expectations were through the roof coming from all sides.
On the other side of the same coin, our AI talent (AI devs, applied researchers) in charge of developing what would give us our industry edge was fairly new to the corporate and business realities, a lot having left academia not much before joining Element AI. Across the company, there was this misguided belief - though not shared by everyone - that “AI” was something that could be sold. In other words, that it held intrinsic value in itself for our clients. But you can’t sell AI-as-a-product. If that’s what you think you are doing, it’s called consulting. In the product realm, AI is a means to an end. It’s a way to make your product better at solving a client’s current or future pain points. We were also operating under the pretense that our consulting branch would talk to customers and identify opportunities through early consulting engagements. In turn, the product teams would turn collected insights into a successful products built behind closed doors. But if there is one golden rule that should never be broken when developing new products, it’s this one: there should be no middle man between customers and the Product Managers building the product.
I believe it would do no one a service to avoid mentioning another factor that contributed to us failing as a product organization. Our executive team was in way over its head with something that seems to have caught fire overnight way above anyone’s expectations. As you point out, there is such a thing as too much of a good thing... or rather a couple of good things: the money we raised and the hype wave we were riding.
Today, I have come to appreciate Element AI for what it was. A fantastic learning experience for the AI/tech Montreal and Toronto communities, as well as all the international talent that joined the adventure. We tried. We gave it all we had. And we sure learned a lot. Not to mention the very rich professional and personal relationships that were forged. By the way, these lasting relationships cross the product/consulting boundary.
Given the opportunity, I would do it all over again. But perhaps a little differently this time… :)
Share
Thank you, @iknowjerome! Thank you @philbeaudoin for writing this.
About my time at EAI, won’t repeat many of the wonderful things you mentioned.
On the product side, two questions for the two of you:
- What are the risks of developing products around one or a few large clients? It seems the more involved they are, the bigger the risk of translating that into internal frictions and misalignments.
- AI is not a product, agreed. What can be said the selection of products or few products for an AI company that started based on talent? Perhaps the talent could hint it’d be easier to develop more technological products.
I think there were also conflicting forces about what products and technology would look like. My view is partial as I was not involved in the consultancy angle.
@iknowjerome @hectorppal On the risk of building a product around one large client: building a product only for that client. This is similar to a consultancy. If only one client needs your thing, it's a solution — not a product.
Selecting a product: It's hard. I think part of the problem at Element AI is that we desperately wanted AI to be a product. We wanted something super-duper-big to justify our existence... I believe the right approach would have been to start with a good collection of small-AI-products-with-real-customers but striving to connect our products more and more into what might have become the AI platform we were dreaming of.
@iknowjerome @philbeaudoin exactly. You cannot describe a straight line with a single point.
I’m very partial to this, but I think there was/is confusion about that our so-called AI can do. Lacking definitions, then scope and evaluation. Demos are a just a good start. Otherwise it’s just another kind of cargo cult.
Ps: I would never imagine I would participate in a conversation about this. EAI happened.
@iknowjerome @philbeaudoin I agree that would have been better. A platform visited and revisited. I think they call this “middle-out” design.
It’s tricky but it won’t happens without explicit meaningful attention.
@hectorppal
It’s often said that product management is about using data to inform what should be built and in which order. It’s also about capturing current and future user pains by getting users to talk about their business and frustrations. Both of these things are true, but I believe the hardest part about product management - especially in the context of launching new products - is about walking that fine line at the intersection of consulting and product. You need to be connected to a few clients to guide your roadmap, but you can’t cater to all their asks, and definitely can’t be at the mercy of how they imagine the solution (product?) to look like. If you stray too far away from that fine line in the early phases of creating a new product, you run the risk of either building something that solves no real problem, or building a custom solution that only applies to one specific customer. As far as I know, there are no hard rules to follow to avoid going too far in one direction or the other. But a good safeguard is to make sure the people building the product are also the ones talking to the users/customers. And as a rule of thumb, I wouldn’t recommend focusing solely on one or two very large customers because it will set expectations that the product be fully tailored to them.
Nice one Phil!
Very interesting. The myth of Element AI explained! A lot of learnings in this. We see a lot of start-ups out there idolize hyper-growth but it comes with a cost. Thanks for sharing
@alexandre_caron Yes. Based on these learnings I purposefully built Waverly with a small initial round, a slow burn, a "minimal tech" approach and an MVP.
This is beautifully written from the product perspective Phil. Bravo!
Diamonds emerge from pressure being applied to the right elements at the right depth for the right duration. Getting all these aspects to align correctly is tricky, an absolute art in both flexibility and discipline, certainly a fine tightrope to walk nevermind to walk well. The people at Element were absolute magic yet the pressure points, as you mention, were not aligned to drive product revenues. I am so proud of the learnings you have translated from this formative experience into the ethos of Waverly and cannot wait to see it grow, grow, grow. Félicitations mon ami!!
Bravo Phil, candor and transparency are so important. It's essential that these stories are shared because we gain considerably more from this information than from some of the borderline glib stories coming out of growing startups, which are inclined to 'hide their shortcomings' all of which gives young entrepreneurs a woefully distorted view of how things really work.
@james_alexander_ i agrée. It’s just my perspective, I was not privy to many of the most important discussions and decisions that were made at the board and CEO level. No doubt other analyses will emerge and will differ from mine. And yeah, its kind of a high risk / low reward situation for a founder to share their story candidly like this, but I just felt I had to…
Great writeup Phil :) Magical years but ones that left many with a tantalising sense of what might have been...
Salut Philippe, j'ai lu ton billet et je suis 100% d'accord. Les grand esprits se rejoignent, car je viens de commenter un billet de Yoshua : « Most important, have a REAL PRODUCT for a REAL MARKET, if you don't want to do just value-added recruitment for GAFAM and sell your company 1 million for each ML engineer or data scientist, you've hired, based on a well known gold rule. But it’s up to you… ». En passant, j'ai une idée de produit qui va « casser la baraque ».... Eurekia.