All of this so-called "dark data" can be used for new insight, for new actions. You may not know which ones yet -- but you won't know until you have had a chance to look.
Explore new ways to use your data
With reason, give unrestricted access to all data to your analysts. I say "with reason", because some industries have specific regulations that apply, and all industries need to be careful about data privacy. But, assuming you can trust them, and the proper governance exists that would curb any abuse, let your experts explore the data. They will find ideas -- some that will work and some that won't.
It's always difficult to measure the return on investment of innovation. But innovation happens when people are let to pursue ideas. Google got it right, with their "20 percent time" program.
You may not have official data scientists. But a good business analyst, equipped with modern tools for data preparation/exploration, can achieve amazing results.
Release often, test all the time and fail fast
This last point is probably the most critical one. Digital organizations are agile. They always test innovations in real-life conditions. Once a feature is complete, go live with it and measure its effectiveness. If it does not yield the expected results, be ready to pull back -- revert to the previous version, remove the option, try something else. And whenever possible, test different alternatives in parallel to see which one works best.
The key is not to always get it right -- nobody does. The key to success is to fail fast, and change course before it's too late.
Sign up for CIO Asia eNewsletters.