Independence Drives Speed

Published: by

In the last week, I have had several discussions with some really smart technologists, partially focused on what makes technology companies nimble and fast and, therefore, great.

In the last article, we discussed hiring 10x people, and especially the way many great employees compound together to create up to 2 orders of magnitude faster companies.

However, hiring really smart employees is necessary, but it is not sufficient. What these employees need is independence. They need the ability to innovate, create, design, test and, whenever possible, deploy their innovations independently.

What does that mean?

Your employees with great ideas know that the "proof is in the pudding." They may have a great idea for making the user interface better or product faster, but the only real proof that matters is whether or not it works. Great employees actually are hungry to see their ideas in live usage bringing real value. They want, as fast as is possible, to see if their ideas pan out... and then move on to the next great one.

Can your employees test out their concepts independently?

  • John believes that a change in layout and presentation on your Web site will increase conversion rates by 1%. Does he have the ability to deploy and A/B test it, even on a small scale? Or does he need to go through an long and complex complete deployment process that involves many people?
  • Sarah believes that a change in a core software component can make your Web service respond 15% faster. Is she able to run the necessary components of the performance test on her own? Or does she need to await a complete build, release and test cycle from a different team?

More accurately, do your culture and systems enable employees to test their creations independently?

Most companies without such systems will say it doesn't matter. After all, John and his peers rarely ask to do A/B testing, and Sarah's team usually works on larger scale cycles and releases. Why bother investing in giving them faster cycles and tools?

Because it matters not how often they do it; it matters how often they will do it.

In some ways, this is like the debate about dynamic scoring of changes in tax rates. The important question is not if reducing taxes by 1% will reduce governmental revenues by 1%. The important question is what is the impact on the economy as a whole of having both 1% less government spending and 1% more in people's independent pockets.

Similarly, it matters not how often employees A/B test now, or how often they run performance tests now. What matters is that you make it so they can do it on their own.

Even in a small company, let alone a large one, you, as an executive, simply have no idea how often an employee comes up with an idea, wants to try it quickly and test his or her hypothesis, then lets it go upon realizing how hard it is to test and validate.

Thomas Edison is said to have told a French reporter, when asked how it felt to fail nearly 1,000 times in attempting to invent the incandescent lightbulb, that he did not fail 1,000 times, he succeeded 1,000 times in learning how not to make a lightbulb.

The improvements in your product that enable you both to deal with problems but more importantly to grow and accelerate beyond your competitors almost never are due to a few big planned improvements or product releases. They almost always are due to thousands of small improvements that have been tested and discarded, leaving the few successes. The more experiments your people can run, the more successes they will have.


Your chances of success depend on the ability of your employees to try many thousands, if not millions, of small or big things that may or may not matter. You need both to create a culture that encourages independent creativity and time spent that may, yes, "fail", as well as provide the tools that let them run test those hypotheses independently.

What are your product development culture and tools? Are your people trying lots of new small experiments each week, the only way to consistently succeed?

If you are not sure, you do not have them.

Ask us to help you build the culture and tools for independent success.