Many entrepreneurs and tech companies laud the idea of rapidly trying new things and failing fast. But it’s a luxury that most companies can’t afford as well as a waste of time, according to Pandora CTO Tom Conrad.
Since the company pays out a large amount of money to the companies it licenses music from, Pandora had to ruthlessly prioritize. Here’s how Conrad explained it at a talk at venture capital firm First Round Review’s recent CTO Summit:
“There’s this motto in our industry that says fail fast and fail often. But none of us can actually do that right? We have all kinds of constituents — employees, investors, users — they are expecting you to do smart things, not dumb things. So one of the first things I said was let’s not pretend we can just try things and some will work out and some won’t. That’s not winning, that’s losing.”
Instead the company built and serviced a 70 million-person monthly user base, grew to a half a billion in revenue, and went public with only 40 engineers.
Pandora developed an efficient system to prioritize engineering tasks: Projects and planning were limited to 90 days in advance since longer time frames lead to slower reactions. Conrad would start by asking “What would it be stupid not to do in the next 90 days?” and then source ideas from throughout the company. Then, he estimated total engineering capacity, and assigned it a dollar value—$5, for example, would represent a feature that takes an engineer a month, so a 10 engineer team would create a $50 dollar budget.
A team of decision makers then voted with a budget based on engineering capacity. After a period of discussion, consolidation and advocacy, the short list of ideas that got fully “funded” would move forward.
The system, which Pandora has used for 8 years, is an excellent example of the power of constraint. More work, and more important work, was accomplished because projects were precisely mapped to the available budget and supply of engineering talent.
If people aren’t honest about their productivity the system can fail. And adding more hours or people to a task doesn’t always make it better. But assigning value to time, and strictly limiting it, means less gets wasted.