Limited Runways and Innovation

Res Extensa #49 :: The benefits of deadlines, budgets, and time-boxing

Jumping off of a previous issue on constraints, let's talk a little about the benefits of a limited runway.

A while back I heard Ryan Petersen, founder of the logistics company Flexport, talking about taking investment in a growing company. Here's his advice to founders who've just raised money, with that sweet cash just wired to their bank accounts:

Immediately do a hiring freeze. Do not hire anybody for 90 days. More people are not going to solve your problems.

Even with the runway extended possibly for years with all that fresh capital, he still says to keep the clamps on. At least for long enough to sort out how to use it wisely. The list of companies that developed terrible habits post-extended-runway is long —  overhiring, extravagant team retreats, unnecessary perks, and general corporate decadence. Tighter belts encourage discipline.

In the world of startups, you'll hear about "runway" constantly. Talk of "cash burn" and "burn multiples" and how much cash runway the company has to reach the next stage.

Startups are capital-intensive bets on a product solving a new problem, in a (hopefully) large and new market, so they need a large payload of new money every 2 to 3 years to keep building the bet, with each new milestone reached. Investors buy equity as a gamble that the company will be able to convert a few million dollars into a trajectory that reaches escape velocity, self-sustainability, and many-X valuations.

The runway defines how long you've got to hit the next milestone. It's a limiter on execution. Like Ryan, I think boundaries like this are utterly essential as forcing functions.

Fixed boundaries or open-ended exploration?

In their book Why Greatness Cannot Be Planned (discussed briefly in RE 41), AI researchers Ken Stanley and Joel Lehman claim that big, ambitious goals require open-ended path exploration, not planned, top-down definition of objectives. The route to any innovative discovery passes through seemingly-unrelated stepping stones (computers required vacuum tubes, et al). So their argument comes across as sort of contra-runways and limits. They propose the search for novelty in a "playground" where we can explore and tinker.

I get the idea. Thousands of innovations in history resulted from accidents, and were side quests from the discover's stated goal. Penicillin, microwave ovens, Post-it notes, Velcro, x-rays — none of these originated with a stated objective: "I want to create _____". Their inventors happened across them. These stories fit with the Stanley and Lehman thesis: "the best way to reach an objective is to never have one." But even though they didn't have these discoveries as defined objectives, they were certainly constrained. What limitations did Fleming have on his work when he noticed mold in his petri dishes? Did Röntgen's superiors have him on a tight budget for research when he ran across x-rays? Even if unconstrained by specific time-bound runways, they had physics, chemistry, and available technology to keep them bound.

I'm all for tinkering as a means of discovery. Innovation is chock full of serendipity. So where does that leave the notion of having a runway? Wandering the earth hoping for lightning strikes clearly isn't how we should be working day to day.

Runways vs. playgrounds

The distinction between runways and playgrounds pivots on the definition of "ambitious goals". There's a difference in how to approach the ambitious unknown versus the everyday (but often still challenging) work requiring deliberate execution. In the second case, we know the objective is reachable, but we've gotta perform to get there. In the first, it's uncertain whether the goal is attainable at all, and the series of steps is opaque.

We can match up the constraint-defining strategies for two types of objectives: runways for known and reachable targets, playgrounds where the future and the path are unknown.

Benefitting from both

This fits together logically for me. There's benefit from each method, if you first understand the class of objective you're pursuing.

I think what causes problems for teams is when they get these matches backward. I've seen it repeatedly in my career.

• Adding too many constraints when chasing far-out, unknown R&D — this could mean tight budgets, delivery dates, being too short-sighted or impatient with the simmering stew of novelty search. Here we need creative freedom but sometimes end up with committees and directives guiding the project. Blind leading blind.

• And on the contrary, projects where the end state is clearly achievable, but have no borders for execution or accountability. Things like this often end up meandering, with no hard deadlines, no tight "definition of done", no constraints where they're essential to good work. Time is of the essence here, and it gets wasted.

I think we can merge these two concepts: a constrained runway with wide enough left/right bounds to explore, but leave room for unplanned discovery.

This parallels the idea of the hard edge and soft middle. It's hard to be productively creative without constraints. Sure, we can just wander around complete aimlessly completely exploratory. The problem is that with no objective — just playful meandering — there's no predictability. There needs to be a forcing function.

The biggest threat to creativity and scientific progress is therefore the opposite: a lack of structure and restrictions. Without structure, we cannot differentiate, compare or experiment with ideas. —Sönke Ahrens, How to Take Smart Notes

Like Ryan's example for hiring freezes, we need to get good at enforcing runways, even if they’re self-imposed. The "arbitrary" date can make people uncomfortable, but nothing crystallizes attention like a ticking clock. The challenge is to design systems for working that respect constraints and value focus while affording the freedom for experiment where necessary. Tightening your range of motion creates healthier habits. If you keep runways tight, you'll develop better hygiene, and be more resilient to future storms.

---