This year I’ve met over 350 startups building AI-first applications. I’ve been struck by how different company building is for these businesses compared to SaaS companies of yesteryear. The TLDR is that it’s easier than ever to start, but harder than ever to win. Three observations below.
1. A strong wedge product is table stakes
Historically, the biggest challenge for an early-stage founder was finding a high ROI wedge product. If you could find the wedge, you could layer on future product lines and build a layer cake of value.
But AI has democratized building wedge products. Even in obscure verticals (e.g. chemical manufacturing, employment law, home health), I have seen dozens of startups launch within months of each other, with many hitting PMF in record time. The playbook of picking overlooked verticals, identifying labor-heavy workflows and then automating them is no longer enough. There’s simply too much competition for each idea. As a result, founders need to turn their wedge into something of deeper value as quickly as possible.
Legaltech - a category that was uninteresting to investors a mere two years ago - is already a hugely crowded space
2. It’s critical to build defensibility “beyond the AI” from the outset
Defensibility comes from building value “beyond the AI” and is something that founders need to consider from the early days. Below are some strategies I’ve seen work, many of which can be used in conjunction of each other.
- A land grab is optimal when the wedge product has high ROI but is replicable. Executing on a land grab requires fundraising at a rapid clip, building a war chest and signaling your ambition to competitors
- EvenUp, a legaltech startup focused on personal injury law, has pursued this strategy. They were first to market and have grown quickly, but have also seen several fast followers emerge. To stay ahead, they’ve raised very large sums - $180M+ in less than two years - to bolster GTM and capture the market
- Proprietary distribution can help a commoditized AI product win the market
- The AI medical scribe category has become incredibly noisy, with 20+ companies coming to market this year alone, with most being at parity in terms of accuracy. Despite this, companies like Abridge have been able to stand out, securing an integration and partnership with Epic (a dominant EHR with 40% market share) to supercharge distribution
- Quickly becoming multi-product (“Rippling for X”) is another strategy to stand apart in a crowded vertical. This can be particularly attractive in enterprise markets with long procurement cycles, where there is a preference for adopting one vendor instead of several point solutions
- In many healthcare verticals, providers are reimbursed by insurance upon furnishing accurate clinical documentation. Many startups are working on automating documentation, yet, from the perspective of the provider, this is just one of several back-office tasks that can be automated. A startup that quickly builds out an end-to-end offering (e.g. intake, scheduling, scribing and RCM) would be able to command substantially higher ACVs and differentiate from competitors
- Building deep integrations + workflow software within a vertical can create a position of strength for a startup
- Our portfolio company Revv builds software for auto shops. They’ve built a deep integration (and synergistic value prop) with the dominant workflow software in the space, which ensures their product can be seamlessly accessed by shop technicians. From this entrenched position, they are able to layer on additional product lines to the same customer
3. Be more like Travis at Uber
Every venture fund says “it’s all about the founders.” While this cliche can be tiresome, it’s evident that in the current environment - with fervent competition for every idea - founder ambition + strategic thinking matter more than ever. The founder who encapsulates this for me (for better or worse) is Travis Kalanick at Uber.
Charles Moldow from Foundation Capital wrote a great post about the two archetypes of founders he’s seen in his investing career and the environment that each is best suited to. The TLDR is that founders in blue ocean categories have the luxury of being more deliberate and systematic with company building whereas those in crowded categories need to operate with much more urgency. It’s evident that most AI businesses today are in the latter bucket. As a result, speed of execution, vision and fundraising skills are founder traits that will be prized by VCs, and will also be more valuable to founders than ever.
Concluding thoughts
Overall, it’s a wild time to be investing in and building startups. Since becoming a VC six years ago, I’ve never seen more industries being transformed by software. But I’ve also never seen the amount of competition for each idea. To cut through the noise, founders need to think beyond their wedge product and build defensibility from the very beginning. In Darwinian terms, only the fastest moving and most aggressive founders will survive.
If this post resonates, I’d love to hear from you - I’m at samit@1984.vc.