Member-only story
Landings over launches: How Google thinks about successful products
Product Management for Engineers (part five)
On the 19th October 2016, the Schiaparelli Mars lander dropped into the Martian atmosphere. Three minutes later, it deployed it’s parachute before jettisoning the chute at an altitude of about 2.3 miles (3.7 km). It fired its thrusters for 3 seconds rather than the expected 30 seconds. Shortly afterwards it impacted on the surface of Mars, travelling at about 335mph (or 540km/h). Going over data from the crash it became clear that the lander was spinning so fast that it overloaded the guidance computer. The resulting confusion led the lander to think it had reached the planet’s surface when, in fact, it was high in the atmosphere. What does this teach us about building products, you may ask? Read on.
One thing that became more and more apparent at Google during my time was that releasing features was becoming the end in and of itself. This led to people prioritising the release of new features over maintaining the quality of existing products or making sure those features were fit for purpose. People were rewarded for getting their project to the finish line rather than winning the race. We started to refer to wanting to know about how something landed rather than the fact it was launched in the first place. It’s relatively easy to launch a human being into space. At least when compared to getting them there and back alive and undamaged. We realised that you don’t celebrate when your rocket leaves the launchpad. You celebrate when your humans get back in one piece.
Landings are, on the whole, much harder to measure than launches. Yes, in the case of Schiaparelli, you can measure it in terms of the black smear that the lander made on the Martian landscape. But, the reality is a lot of product engineering teams launch things without really knowing how to verify that what they’ve built is making a positive difference or solving the problem they expected it to. This leads to some behaviour which is both value and soul destroying:
- Relying on our instincts rather than data. This can also been seen clearly in product decisions being made based on seniority rather than data or people asking you to accept LOFAs (leap of faith assumptions). If you’re launching a spaceship to MARS then faith is a…
