The data factory team is filled with people who create the plan to take ideas to market—and have skills to do so. These are engineers, marketers and sales people. They get the job done, and the product or service out the door on time and within budget. While the data lab is free wheeling and open, the data factory is disciplined and process driven to the point of rigidity.
Step 3: Set up metrics for success
Metrics help you improve results and align your people and processes with organizational objectives.
For a data lab, a metric might be coming up with a certain number of ideas per quarter, or patents per year. If a business objective is to improve the customer experience, one metric might be to derive ideas focused on the customer experience.
Data factory metrics will vary greatly by business and industry. For example, a software company might aim for a certain number of new releases per year. An ice cream store might seek to successfully implement a new supply chain of providers by the time the busy summer season arrives.
Whatever the metrics you set, make them achievable. Nothing kills an innovative process faster than struggling under the weight of unattainable goals. Set one metric for success and then another as a super stretch goal. Also, make sure your employees know what the metrics for success are, and how they, and the metrics, are being measured.
Step 4: Set criteria for idea success
It doesn’t matter if a company successfully implements a new service, process or product if that service, process or product doesn’t add business value. In fact, taking a product or service that far—and only then realizing that it doesn’t add value—may be even worse than having an idea die for lack of testing ability.
Before investing heavily in an idea and its implementation, clarify what pain point the idea is lessening or removing. What will be the business value of implementing the idea? What amount of investment is reasonable and smart to make in the idea, factoring in the potential return on investment? How best should the idea be implemented? This is the step where you build assumptions and then test those assumptions before you break open the bank to fund an idea full force.
Step 5: Create feedback loops
Once an idea gets out of the data lab and the data factory into the real world, continuously track what’s working, what’s not, and what needs to change. If you’re a software company, for instance, your sales team will get feedback from customers about your product. That feedback needs to get to your engineers so they can tweak the product to better meet the customer’s needs. The same feedback loop should occur all the way back to the data scientists. The data might show, for instance, that customers appear to prefer a certain kind of pickle with a certain kind of sandwich and so franchise restaurant owners start serving more of those pickles only to then find that customers are tossing them out in droves. That information, fed back to the data lab, will inform a new round of research and, hopefully, the right solution.
Step 6: Celebrate success
Sales teams do this better than any other piece of a company, and all managers from all kinds of departments can learn from great sales leaders. They celebrate when teams close deals, expand deals or even push deals one step further to closure. This kind of energy builds an “all for one” mentality that always ends up bigger and more powerful than if teams work in isolation. Make sure your data lab gets kudos for the great ideas it comes up with and make sure the data factory gets kudos for getting those great ideas to market.
The Big Payoff
The McKinsey Global Institute estimates that advanced analytics will enable industries to create $9.5 trillion to $15.4 trillion of value. To get there, organizations across industries need to leverage analytics to make more informed strategic decisions. They need to find novel and useful insights in data, and then turn those insights into products and services—and make a profit doing so.
But as companies have already discovered, becoming a data-driven organization requires changes in people and processes. Improve your odds by building a data lab and a data factory, and then hold them both accountable, separately and together, for making a difference to the bottomline.