The following is a guest contributed post from Christine Oliveri, Vice President of Operations, Client Services for I.Predictus.
Isaac Asimov, the prolific author and biochemistry professor, once remarked that “If knowledge can create problems, it is not through ignorance that we can solve them.”
I would add an addendum: we cannot solve problems through opinions either.
In the business world, there is a tendency to rely on what the folks at Amazon call “HiPPOs,” or the Highest Paid Person’s Opinion. And where there is a dearth of data, there is going to be a surplus of HiPPOs.
When we lack the hard numbers and facts we need to make informed decisions, we substitute fluff since we have to come up with an answer—particular in dips and crises. Or worse, we sometimes have the data, facts and analysis we need to make a great decisions, and it gets ignored.
That is dataphobia. It is cognitive dissonance elevated to a cultural level. It is the furled eyebrows and redness we see in a data scientist’s face when she leaves a meeting where years of collected information and rigorous modeling have been completely ignored in favor of a HiPPO.
Data ignorance is not bliss. As Forbes reported in April, the now former CEO of JC Penny, Ron Johnson, took a classic HiPPO approach to reviving a sleepy retailer. He ignored focus group data, he insisted on making nationwide changes before testing any of his ideas and he never checked to see if his plans were working. JC Penny burned through $1 billion in cash within 17 months, lost an additional $1 billion in revenue and the company’s market capitalization fell by nearly 50 percent.
A successful approach to big data must combine technology and a campaign to transform company culture. In the way that water is a tool of the fire department, data is just the tool of a knowledge-driven culture. The development of data instrumentation, warehousing and reporting goes in step with the cultivation of a culture that values this investment. As I’ll explain below, replacing dataphobia with dataphilia requires attention towards both systems and people.
1) Name Your Data Champion
If big data in action is the holy grail of business expansion, revenue growth and cost-efficiency, then you need a knight to lead the quest. In other words, your big data program is not a project for the summer interns—it needs to be led by someone with deep curiosity, strong soft skills and a critical eye.
Your data champion, whether it is your newly appointed Chief Data Officer (CDO), your Manager of IT Systems or Bob in Marketing, needs to do data inventory. What does the company collect? What should it be collecting? What do we wish we knew about our business and market?
Then, it’s time to do some research and meet the vendors, who are now plentiful. Many will help you realize that there is lots of data collection and modeling possibilities that you were not aware of. They may identify that your data labeling and warehousing processes are different across departments. Your data champion should explore solutions, grill vendors with questions and find an array of options that can be confidently presented at the C-level.
2) Choose Your Weapon(s)
Your data champion hopefully knows their C-level audience. Do they want the best bang for the buck? Do they want the Cadillac of big data? Are they willing to hire people who can administer these systems?
Here’s how the data champion can win before going into the lion’s den: he or she picks systems that are comprehensive, performance-based and do a lot of the legwork.
Whatever vendors may tell you, a lot of data systems require rocket scientists to run them. No seriously—many of the best data scientists have PhDs in astrophysics or equally esoteric disciplines. The data champion will may not get far if they tell the executives that the company just has to hire three software engineers, two statisticians and an astrophysicist to get a decent ROI on the system. Ideally, Bob from Marketing should be able to access and use the system as easily as a data professional, though they may use the system for different purposes.
So look for systems that do a lot of the algorithmic heavy lifting, and look for systems that have you pay according to the value they generate. At some point, you may want data scientists—but try to present and start on a system that does not make them an unconditional part of the package.
3) Test and Tell
The best way to win colleagues over to a data-driven culture is to test and tell. You could have employees read about big data success at other companies, but the “were different here” mentality is resilient. Lay that myth of uniqueness to rest with an in-company test.
Depending on your big data needs and chosen solution, the best test environment will vary. For instance, if you purchase a marketing big data system, invite the media buying team in your marketing department to pilot the system. Explain that they have a serious opportunity boost ROI and distinguish themselves throughout the department and company. If you picked a good solution, they won’t need too much training and direction.
If and when that team generates superior results, the power of data is going to catch on. Internally publicize what the pilot team accomplished. The pilot users themselves will probably become evangelists and the other departments will want in. Reward people who make the transition and produce data-driven results. Suddenly, you have an organization that equates data use with success, attention, promotions and bonuses. The incentives align and the mast of culture swings in the changing winds.
Big data champions fail in their quest when they make the paradoxical mistake of making the case for big data with opinion. In an opinion shootout, we know the highest paid person wins. So as I’ve highlighted above, place the challenge of eradicating dataphobia into the hands of a formidable individual (or team). Remind his champion that numbers speak best when you’re trying to convince an organization to place greater value in numbers.
Do not wait for a competitor to deplete your market share using a data program. Unfortunately in a business blaze, most people forget to stop, drop and roll. In stressful environments, HiPPOs can wrestle control away from knowledge-driven responses.
We know the costs of ignorance, and in business we know that tragedy begins when hubris or panic lead us to reject knowledge. Don’t let dataphobia bring down your organization.
About the Author:
Christine Oliveri is the Vice President of Operations, Client Services for I.Predictus. Christine holds a B.A. in Finance and Marketing from The George Washington University in Washington, DC where she graduated with the highest honors.