Crossing the deep tech expectation-reality chasm
What to do when reality isn't lining up with what you promised
Entrepreneurs can find themselves in a reality that has come to diverge greatly from the expectations they set for themselves and publicly. This expectation-reality gap has an even higher likelihood of being large at a deep tech startup compared to a traditional one, given the longer timelines and technological uncertainty these founders are managing. One of the top jobs of a founder is to spot as early as possible when the gap is becoming too big to manage, and to course correct in a way that preserves your relationships and integrity while ensuring your startup has enough cash in the bank to carry on.
Elizabeth Holmes’s botched Walgreens partnership gives us a good example of what not to do when this expectation-reality gap becomes unwieldily. Ms Holmes entered an agreement with Walgreens to open “Wellness Centers” equipped with her company, Theranos’s, Edison machines for rapid blood tests from a finger prick. When it became clear inside Theranos that they were behind schedule and the machine’s efficacy was below that required, they started running the tests on third-party machines, meanwhile leading everyone outside the company, and even many inside the company, to believe the Edison machines were successfully performing the tests. Choosing deception when she noticed the problematic gap set the partnership up to fail, and ultimately this inability to deal with uncomfortable realities among other troublesome behaviors led to Theranos’s collapse.
Of course, you can also try to avoid getting into these situations in the first place. Like Holmes encountered, one source of potential gap comes from the need to engage with customers, validating your business hypotheses, generating real feedback from users, and demonstrating traction. Instead of signing binding contracts with real delivery dates to accomplish these goals, you can leverage softer agreements like letters of intent (LOIs) until your completed product is in clear sight (i.e., the main technical risks have been solved and you can plan the remaining engineering cycles with decent certainty). LOIs can keep you on your customers’ minds, create a sense of psychological buy-in to the idea of working with you, and be used as a proof point for investors and others that there’s real interest.
Speaking of investors, deep tech founders funding their startups with venture capital face a second situation with potential for a large expectation-reality gap. You will likely have to put make-believe dates to your unpredictable R&D phase and may find yourself facing the end of your runway without having hit those technology or product milestones yet. A common misstep that can be fatal to a startup is to manufacture a vanity milestone: a flashy demo or test result that attracts attention but doesn’t burn down key risks or isn’t truly aligned with your path to your first product. This is dangerous: the smart investors you want around your table won’t fall for this, and even fake demonstrations have a sneaky way of sucking up all of your focus and resources in deep tech. Worse yet, engaging in this behavior once may mean you have to keep doing it, since you won’t have made enough progress against the substantial work you should have been doing come the next fundraising time.
Unlike using LOIs, there aren’t great ways to soft pedal your way around dates with investors. Some are sympathetic to deep tech timelines and have raised their funds from LPs who are willing to close their eyes and avoid looking at results for a bit longer. In my experience, investors still operate on the 12-18 month cycle and even with patient capital it's important to be aware of the pressure to externally demonstrate progress on those timelines.
Another strategy is to raise a ton of capital up front and disappear from the public and your investors’ eyes for this phase. Your ability to raise significant sums of money early in the life of your startup when there’s little to no real data on your technology or the market appetite is largely determined by how big of an outcome you can expect and how quickly you and your investors could reap the rewards. This approach is realistically available to very few startups, and requires you to be a disciplined steward of the cash over many years, which is easier said than done.
But there may come a time when you’d exhausted LOIs and just needed to sign a real contract whose date is now impossible to meet, or you’d had to put real dates in your pitch decks and it’s become apparent you’re going to miss them. Spotting these gaps that need to be corrected with enough time to fix the situation and have another go is a key responsibility of a founder. A crisp way to do this is to use “killer experiments” in your R&D, and across the rest of your business.
Killer experiments, borrowed from pharma and biotech, are checkpoints designed to stop efforts before insidious effects steal all your cash: they keep you from falling for the sunk cost fallacy, they prevent you from feeling satisfied with incremental progress when you need exponential, and they force you to examine new approaches even when you thought you’d exhausted them all. They are set in advance when you’ve still got your wits about you. And they make grey areas of progress into black and white decisions by removing the emotional aspect.
A killer experiment at a battery company might look like setting a life cycle target you must hit by a certain date, and agreeing (on your team or even with yourself) to pivot to a different chemistry or electrode if the target’s not met, and being firm about the decision even when progress was trending in the right direction. If you let yourself set or renegotiate pass/fail criteria for an approach in the moment, it’s easy to talk yourself into continuing down the same path (speaking from 8 years of experience down one path). After all, most innovations that don’t violate the laws of physics are possible to realize with enough time and money, but the market or opportunity may pass you by in the meantime.
If the killer experiment fails and it’s clear you’ll have to break a promise, it’s time to take steps to remedy the situation. While falling on your sword with customers and investors is painful, it’s always less catastrophic than the Holmesian alternative and if done early enough can preserve your company, using up just one of your lives. Follow these steps to get back on track, then pick yourself back up and go try again:
Get clear on why the miss happened. Did you just need an extra design-build-test cycle or is there a fundamental problem with your approach and you need to think hard about scrapping it and starting over?
Decide on the best way forward. Do you delay your timeline, perhaps needing to ask for a little more money, or push out a product delivery date so you can meet your promised specs? Bring in a trusted partner at the company or run the scenario by experienced advisors to see if they agree with your thinking and conclusion, and don’t lose sight of your key risks and path to market.
Develop a plan and think through the implications as best you can. For example: renegotiate contracts with customers, share the news with your board, or raise a bridge round from insiders so you can work on the real milestones.
Communicate with your team and implement the change. This calls for a strong leader to reassure the team it’s still heading in the right direction, while also recognizing the effort, roles, or plans that need to be scrapped to get back on track. Hold an honest retrospective and hear from the whole team what went wrong so it doesn’t happen again.