Innovation Doesn’t Always Follow a Straight Line


Customers love it when products work as designed and are released on time. 

Who doesn’t, right? But for Innovation teams in small companies like dataxu, building reliable, impactful products while balancing temptation to chase net-new projects can be a delicate dance. While I certainly aspire to release products on time, a dogmatic insistence on only working on ‘blessed’ projects cannot be the whole story. Countless inventions have come from ‘skunkworks’ projects or even mistakes.

How to keep customers and your Innovation team happy 

The key to keeping customers happy, while still generating new products, is to maintain focus on committed projects while building in some flexibility to study new ideas. In our research team, we achieve this ‘exploitation vs. exploration’ tradeoff in several ways:  

Tactic 1

First, members of our team spend time helping external clients and internal teams with issues and ad campaign setup. Sometimes, after seeing the same concern come up time and time again, they find a way to resolve the problem using Data Science and a new product is born.   For example, recently, several improvements to our retargeting algorithms came from deep diving into customers’ issues.  

Tactic 2

Second, on a semi-regular basis, we hold a company-wide “Innovation Day” at dataxuInnovation Day is time put aside for everyone to explore ideas that have been brewing.  People work in cross-functional teams with the end goal of pitching their idea to an executive judging panel, often with a cool demo. Countless new products have come from Innovation Day from many parts of the company. In Data Science, Cruise Control and Lookalike Audiences are examples of projects that started as Innovation Day pitches.

Tactic 3

Finally, ‘intern projects’ are a great way to explore ideas we thought were interesting but didn’t want to commit to fully. They can also be a way to bring in an expert with fresh eyes to look at an existing system. This can sometimes be the catalyst for a new product. For example, our Spend Forecaster and recent machine learning system upgrades launched as ‘intern projects’.

Embrace innovative ideas and projects  

The ideas above are examples of ways you can build in extra time for your teams to explore new areas. The exact amount of time will vary according to your situation and may vary week to week. What’s important is to recognize that Innovation doesn’t always follow a straight line. When it comes down to it, the best you can do is plan to deviate from your plan.