Velocity With Valour

There is the old adage: “perfect is the enemy of done”.

One might argue that by following the principle of Undeniable Bestness we lose our velocity and become victims of perfection. So how do we square Undeniable Bestness with the pragmatic need to move fast?

The first question is: why move fast at all? Because time is the most precious thing we have. The more of it we have, the more we can learn about what constitutes real value for our market. And the more we learn about that, the more fully we can deliver this value. The more value we deliver, the happier and more loyal our customers are and so finally, the more successful we are as a business.

To learn where real value lies, we need to try things. In startup parlance we need to “run experiments”. Moving fast means we increase the surface area of our experimentation across time. So we learn faster, discover value earlier and deepen that value for our customers. It’s a positive feedback loop.

The second question is: well if that’s all true, why don't we just completely optimise for speed? I'd argue that if we move too fast, if we go too deep on velocity and compromise too heavily on quality, then we destroy the validity of our experiments.

Suppose we’re trying to test a value proposition with this mindset. We build, release, market and sell something fast. Really fast. So fast that its execution is well below our internal benchmarks of quality (i.e our pursuit of Undeniable Bestness) and our market gives a tepid response. Does that mean the proposition doesn’t hold value for our market? I don’t believe it does. Along with the rapid pace that a minimum viable product mentality encourages, there is also I believe, an antagonistic requirement for minimum viable quality. MVP thinking is no excuse for lazy execution.

Balancing these two questions, the tradeoff between quality and velocity, is perhaps the single most defining challenge of building a startup. And while there is no clear cut answer to that tradeoff I do believe there is a middle road, which we can call "Velocity With Valour". Move as fast as honour (aka the quality of execution) will permit.

There are two primary aspects to walking this middle road:

• Quality to match fidelity: We can move faster if we test at low fidelities to start. But low fidelity does not equate to low quality of execution. We should aim to excel at each level of fidelity. Picasso could achieve a heck of a lot with just a single squiggly line - and we too can achieve a lot with just a customer conversation, a sketch, a prototype or a landing page.
• Be judicious about what we test at higher fidelities: as we climb up the levels of fidelity, the time investment at each level becomes greater and greater. This scale is not linear either - it jumps in orders of magnitude. Going from paper models to a basic coded prototype might be ten times the investment. And from basic mockups to beta features, another multiple of ten. Supporting marketing efforts and production readiness, another multiple after that. And the burden of customer success and support of a feature’s lifetime, perhaps another multiple of ten thereafter. So, as the level of test fidelity increases, we need to be increasingly judicious as to whether we should actually do this test. If we take everything in “test everything” too literally, we will produce so many things of such a low quality, that we will end up testing nothing at all.