How can A.I. offer the shortest path to value?
A founder's framework for integrating A.I. into an existing product or company
There's a wave of AI companies being created every single day. And I'm going to be honest, I get a little jealous sometimes that they have the luxury of thinking about AI use-cases from scratch because there's so much to be built that has never existed before. In any form. Ever. Blue ocean nerds are drooling at the opportunities right now.
But for the vast majority of people who are leading or working at existing companies, the challenge ahead is very different. Everyone is tripping over themselves to integrate AI into their existing products, some better than others.
Here's my framework for how we're thinking about integrating AI at my company 👇
Let's get into it.
Welcome to Making Product Sense
Join thousands of product leaders and founders from the top companies in the world, who are learning how to hone their product sense.
Framework: AI as a Multiplier
One interesting framework on the psychology of money is the idea of money as a multiplier. It posits that if you have bad money habits, the bad money decisions you make when you're poor are multiplied when you're rich. If you have good money habits, the good money decisions you make when you're poor are multiplied when you're rich. It's used to explain why folks who win the lottery or suddenly make millions when they're drafted into the NFL are likely to blow it on unwise purchases and find themselves in an even worse financial position than before they became rich.
I've started to think of AI in a similar way - as a multiplier. If you're a company that has bad bones - no PMF, no differentiation, no value offering - AI isn't going to solve your woes. In fact, it will often multiply them. For a time, you might get some press coverage, you might raise another round of funding, you might get a surge of new users, but the eventually the press will eat you alive, the investors will demand results you can't deliver and your users will churn. All you will have managed to do is waste time and money.
On the flip side, if you have good bones - PMF, a competitive advantage, a high value-offering - those same benefits like a surge of users or another round of funding could be the tailwind you needed to grow incredibly quickly.
Note: I'll caveat this because there's one case in which this framework doesn't hold up: if the company with bad bones needed a technology like AI for their vision to actually work and they were trying to hack together some sub-par version of it until now. In that case, maybe AI unlocks everything else.
AI short-circuits the “path to value”
During the crypto surge a couple of years ago, many companies attempted to bolt on a token or an NFT to their existing products. The problem was that if the application of blockchain technology didn't further the original value prop of the company, it became either an extra source of friction or a disappointment. In the same vein, if a company tries to bolt on AI to their existing offerings, it likely won't be all that useful. For example, low-hanging fruit like text generation or summarization are being added to products that don't need it, simply because it's easy to throw an "AI" tag on your website.
So here's the question I'm going to be asking to help me and my team be more thoughtful about how we implement AI:
How does AI offer a shorter time-to-value?
AI is fantastic at synthesizing a lot of data within a set of guardrails. And it can do so within a natural language chat interface that speaks every language which is possibly the most intuitive interface aside from voice. How can you use that strength to your advantage so that it effectively short circuits the user's traditional path to value?
Let’s play out a simple example. If your product is a library of every historical legal case and lawyers spend hours in your product searching, filtering, and skimming cases for research, adding a summarization or “related cases” feature is a great application because it's expediting the user’s path to the value they already come to you for. But if you're Twitter, using AI to summarize threads is a bit like using a bazooka to hunt a rabbit - it's overkill. The difference in time-to-value is negligible unless the thread bois get more verbose.
Stories from the Trenches
We're still in the early days of the AI hype cycle, so a lot is going to unfold over the next year. But I think of this newsletter a bit like my own personal diary to clarify and document my thoughts and learnings along the journey as a founder so I wanted to share some of my thoughts from the trenches.
Before I wrap up, what do you think about these shorter articles that just cover a single tactic or idea that I’m thinking about?
In the past, some of my most popular articles have been the deep dives on companies like Waze, Zillow and Fortnite that extract some lessons from their rise to success. Do you like these shorter ones more or less than those deep dives?
That’s all for this one - I’ll catch ya next week.