Here’s the thing. Knocking off incumbents has always been tough. Switching costs have always been a tradeoff most organizations just can’t justify.
But, even the incumbents are shitting their P&Ls right now. The low/near-zero marginal cost per-seat license they’ve built their castles on doesn’t work in the per-token world of LLMs.
Their operating leverage of growing revenue faster than OpEx goes out the proverbial window.
Economies of scale aren’t the same either, it doesn’t plateau with growth. It keeps going up. Jobs to be done, as they get done, need LLM tokens unlike the marginalized CPU compute.
Incumbents sitting on healthy profits have to embrace going in red (at least pink) and justify reinvesting in their tech, archaic schemas unsuited for LLMs. Whereas, new Challengers, can come in with VC subsidized plans, newer, much cleaner infra and hit them where it hurts – right in their tokens.
This David vs Goliath game has a new twist.
Goliaths, the incumbents, are merely enjoying the unearned benefits of switching costs: be it data migrations, integration rebuilds, workflow and training costs, implementation risks, relationship or ecosystem loss. Either or all of these. Healthy enterprises seldom switch a critical piece of their ops workflow because of features or cost.
Even hyperscalers haven’t figured this out, they just have deeper pockets to spend a billion (or ten) today and hope to make it back in thousands later, indefinitely. Only model providers have done this successfully i.e. knocked off incumbents. They’re also highly subsidized, that’s more like Goliath v New Goliath, but worth a look on how they’ve taken down industries and not only managed to kill startups but established enterprises that never saw it coming. Look at Chegg, StackOverflow, Quota, WebMD…all were thriving, untouchable at one point. There’s plenty more to come as the model providers become application providers themselves.
Anyway, I digress. Back to the fumes of switching costs that incumbents are still running on. Besides being (arguably) better or not, how have model providers done this [knocked off incumbents]?
- Low Cost, which is highly subsidized currently
- More Appeal, of a magical, addictive product
But…they also went after people first, not their jobs. May not have been an intentional move, but it was brilliant.
Build tools for the day-to-day use, and people will innately trust you with work too.
This personal and work dependency osmosis, while common in physical world and hardware, rarely happens with software. Gmail and Google Workspace is a good example of being done before, but that’s not David, that’s just Goliath playing David.
Decision-makers are people too. Get into their personal lives, you’ll get them at work too.
If Instagram launched a tax filing tool, I bet half of us dump our human CPAs.
For the Davids, now, the key is figuring out how not to face Goliath head-to-head. Not to compete on features, market share, SLA response times, or even cost. It’s guerrilla warfare. Jump out from behind the trees, through the trenches; get to the decision makers beyond their work stations. From outcomes perspective, build and launch addictive daily use tools, tangential to the core business need. Things the incumbents could never justify or prioritize in their quarterly plannings.
Hubspot, I’m not a paid user, does this really well. From their sponsored podcasts, signature generator, Dharmesh’s Agent.ai play…maybe unintentional but brave and brilliant regardless. Way to be in the arena. They’ve managed to go from David to Goliath while keeping all of David’s qualities of being nimble, fast, unafraid to experiment and deliver.
Now, how’s any of this practical, tactical? Great question, Daniel-san (or more like David-san in this metaphor).
Goal: Get them in personal lives, you’ll get them at work too.
Building the next payroll/HR platform? Build the free “what’s my real take-home” and RSU/equity tracker that employees check obsessively. Get the *worker* hooked on understanding their own money, then sell their employer the backend that feeds it.
Starting the next dev/observability platform? Give individual devs a gorgeous free status page and side-project monitoring they’ll use on weekend projects. They drag it into work on Monday.
Creating the next AP / Vendor Management tool for Industry X? Launch a killer app for Lemonade Stand finances, a financial learning aid for kids teaching them how to manage money-in/money-out.
Pick the tangential daily-use tool that reaches your decision-maker *as a person*, at 11pm, on a night out or at home — not as a work-time buyer. That’s the trench Goliath won’t climb into.
Read the official AKOS Blog here: https://akos.ai/blog