Over the past two years I've been wrestling with a puzzle that has a multi-million dollar solution. I believe I've figured it out and want to share the solution with you.
Everyone has heard the stories of success resulting from Big Data and analytics. Business analyzing their data arriving at results that seem almost magical:
- Using buying patterns to predict when women are pregnant, even before their family knows.
- Financial firms using buying profiles and patterns to spot and root out fraud, sometimes before it's even taken place
- Police departments predicting when and where crime is going to take place and acting preemptively to stop it
- Consumer goods manufacturers identifying emerging trends, like the gluten-free craze, and changing their marketing to ride it to profits
- And many, many more.
However, when you look into these cases, what you find is that they're the outliers rather than the norm. For every business that achieves some kind of magical result like this, you'll easily find 100 others that implemented analytics and got no results.
The puzzle is this: why are so few efforts to harness data successful?
The Dirty Secret
What I've found is that people are becoming jaded. They can't find a clear path to take data and turn it into revenue, and the analytics industry is almost universally completely ducking the issue. They're taking the difficulty inherent in achieving these results and trying to pass the buck, under the guise of "do it yourself" and "no data scientist needed" solutions.
I think this needs to be said: business users are NOT the right people to be doing analytics. It is a skillset, just like sales is a skillset and accounting is a skillset. And if you give someone a tool they're not trained to use, you typically don't get a great result. And they may end up looking foolish in the process.
People have been burned by a lack of results, and are becoming skeptical of the opportunity, which IS in fact there, as the success stories point out. The blame for this disillusionment needs to fall squarely on the analytics industry, which is trying to spin this story that analytics is easy, anyone can do it; which in turn implies that if you CAN'T do it you must be stupid. So when people run into these tools and end up confused, they tend to write off the underlying promise rather than admit that maybe they're not able to do something that "anyone can do".
And you can find many examples of people talking about big data being in the "trough of disillusionment" in the hype cycle. Little wonder.
The analytics industry needs to turn back from this disastrous course that it's started to go down, and stop trying to sell solutions that aren't going to work. Let the people trained in data analytics do the analytics work, and let the people trained in turning information into revenue do their job. Turn its focus back to what it was created to do in the first place: deliver relevant information that enables revenue-growing action. We're positioned to achieve tremendous gains in productivity and revenue growth, and this misguided approach is threatening to send us back to the data stone age.
The Secret Sauce
Ironically, the Big Data craze has not resulted in any more meaningful information than we had before. Business "intelligence" solutions only tell you what you already know, or allow you to track situations you've run into before. This is like sitting on a goldmine and trying to mine it with table spoons.
So how can we fulfill the promise of "Big Data"?
I spent of time over the past few years studying this: talking to organizations, looking at their processes, their data, their technology, their teams. And I've found a solution.
Here's the magic formula: if business leaders have a clear picture of the situation--if they have all of the relevant information--they can make the correct choices to grow revenue and drive their business forward. In order to achieve this, I've found one thing that all of the success stories have in common, and the failures do not. Ready?
You Need a Plan
Simple, right? But I'm absolutely floored by how uncommon this is. Would you try to move across the country with a plan? What makes you think you can move a business without one?
So let me share the plan that I've watch be effective over and over with you:
Step One: Ask Questions
The real promise of Big Data (and every data-related movement that came before it, like business intelligence) is to find strategic information that will move the needle on revenue. And more often than not, this information brings visibility to the burning questions in the back of every business leader's mind:
- Why have my sales flat-lined?
- Who should I be targeting with our new product line?
- Why are we struggling with certain customer segments?
- What should I be offering to the segments that aren't responding to our message?
These questions need to be delivered to the data team to find answers. Data teams can live internal to the organization, or outsource them to qualified partners. But that data team's job is to translate the question into requirements that it can use to find the answers in the data, find the answers, and deliver them. That's it.
Step Two: Find Answers
The answers to these questions are the type of information that subject matter experts turn into recommendations for actions:
- Customers in urban areas are responding well to our luxury-oriented product line.
- Our new food line is selling best to our high-income Asian customers.
- Complaints about quality are stemming from a specific packaging plant.
- People who walk to work prefer our premium brand to our economy brand.
But as you probably noticed, answers without any follow-through are interesting, but ultimately useless. They're only meaningful when you can use them to decide what to do about them.
Step Three: Make Recommendations
Subject matter experts are the people who can take the answers and use their knowledge to create specific recommendations about what can be done to address the answers found by the data team. Subject matter experts can be product managers, customer service managers, technical workers, sales consultants, or advertising agencies. It's going to be different depending on the question you're asking and your organization. But ultimately what you need them to do is take the information from the data team and decide the best way(s) to act on it:
- We need to re-allocate our advertising dollars from rural and suburban areas to urban areas
- We should create more products in this line that appeal to the Asian flavor profile.
- We need to investigate QA processes and checks at our West Coast packaging plant.
- We should advertise our premium brand on buses and bus stops.
The data can help with formulating recommendations sometimes, but only in a supporting role. The only people really qualified to make recommendations based on insights are the people whose job it is to know the areas of the business that the insights relate to. The data team doesn't know their business to that degree, and often neither do executives. Good executives are good at delegating, so the people who have been delegated responsibility in the organization need to figure out what the answers mean and translate that into recommendations for action.
Step Four: Make Decisions and Act
This is why the executives make the big bucks: they're the ones who have to pull the trigger on recommendations and make them happen. Fund the new marketing effort, authorize new products, give budget to the division that's entering a new market.
But here's the difference: now they're acting with the information they need in hand--the recommendations, supported by answers to strategic questions, supported by data. Data all the way down at the end of the day, but the result is action that will increase revenue.
How to Make it Work
That's it. You need a strategy, a plan to drive revenue by using data to answer the strategic business questions. And I've given you the outline for it.
Now, there are some tactical issues with putting this into practice. This clearly involves a lot of collaboration between groups, how do you make that work well? And how do you deliver answers from the data in a way that subject matter experts can understand?
Email is one option, it's done the job for a long time. Enterprise social networks are another, they can be hacked to make this type of workflow happen. But it'll be bumpy, because this type of workflow doesn't typically exist yet in organizations, so it will require coordination.
Warning: Shameless plug ahead.
Ideally what you want is a tool that handles this from start to finish and takes all of the friction and communication issues out of the middle. One that can be used by data teams, subject matter experts, and executives, where they can all do what they do best in the same place. But I only know of one such tool that exists, and we made it to help our clients. It's not widely available yet, but if you're interested in learning more about it just click here.