The ComputerWeekly.com's Developer Network blog recently reported on the results from a Rosslyn Analytics survey that shows, among other things, that only 23% of businesses closely align their business strategy to the data that's available to them. What I found particularly interesting was not that statistic but the way that the CW blog, in both the headline and lede of the post, managed to shoehorn the Rosslyn findings into a Big Data story: "Gut instinct drives more firms than big data analytics."
But the Rosslyn report is not about Big Data specifically but rather just data in general. When asked to identify the top issues keeping their organizations from getting value out of their data, those surveyed cited not the lack of advanced analytic skills or adequate technology but rather the scattering of data among disparate systems and the poor quality of that data. For all the powerful things that Big Data technologies can do, they don't provide a silver bullet for these fundamental data issues.
Most tellingly, fewer than half the business leaders Rosslyn surveyed considered their data to be a strategic asset, indicating a lack of executive sponsorship for data initiatives.
Which, as always, reminds me of a story. Several years ago I was helping an organization implement and drive adoption of a sales automation solution. The CEO was convinced (by instinct) that the reason the sales teams weren't hitting their numbers was that the reps weren't out in the field making enough face-to face-calls. He kept hammering on the sales managers to increase the field activity, but without a way to measure how much was going on the discussions quickly devolved into finger pointing and rear-end covering.
The sales managers maintained that their teams were already knocking on every door they could. The data needed to support or disprove the activity thesis was scattered in Outlook inboxes and calendars and, more than anything, verbal reports—and the sales reps had gotten quite proficient at rattling off a stream of "updates" during weekly team calls that made themselves sound very busy.
The solution was simple. We created a basic weekly activity report out of the new sales automation system—just a simple grid that broke out, week by week, the number of prospecting emails, phone calls, and in-person visits that each sales rep on each team made. The data was only as good as what the sales reps had entered into the CRM system—and, despite training and weeks of evangelism to try to get the sales reps to record their activity, we knew that it was incomplete at best.
But that didn't really matter. Once the numbers were side by side, the CEO could immediately see which teams and which reps had a lot of activity and which ones didn't (or, at least, weren't recording their activity regularly.) It didn't take machine learning or complicated correlation analysis to see the conclusion. The leadership team knew who the top performing sales people were (they were the ones taking home big commission checks each month), and they could see right off that those same reps were the ones with the highest activity logged. The CEO turned to his VP of Sales and laughed. "How did we ever run the business without this report?" he said.
The business decision was clear: get the teams out in the field more. With the CEO inspecting the report each week, the sales managers were on the hot seat and pushed their teams to record their activity faithfully. As the numbers got better, the organization was able to identify and settle upon the right level of field activity that drove sales—and field activity rose and so did overall sales.
That's one reason why all the hype around Big Data can actually impede an organization as it tries to transform itself from instinct-driven to data-driven. If you can't get from your house to work because you live on a rutted, weed-choked dirt road, the answer isn't to sell your old sedan and buy a faster car. You probably should to start by fixing the road. And that doesn't mean transforming a dirt road into a multilane highway in one big burst but rather by fixing enough of the ditches and potholes that your car can pass over it.
When it comes to business data, fixing the road means taking the data you already have and using it in a way that has tangible business value—enough tangible value that decision-makers will want to see more of it.
Taking Hadoop and pumping a bunch of data into a massive repository and creating a lot of colorful charts and graphs won't do much if the decision-makers don't trust the number underneath the analysis, so data quality is important. Even worse, though, is if the organization's leaders aren't able to immediately see how the "insights" contained in the analyses offer concrete guidance on what the organization should do. In other words, if the data doesn't help them make decisions, they won't adopt it.
And that's where focusing on the technology first is a mistake, and focusing on the data first isn't much better. Rosslyn's survey respondents believe that "product data" and "customer data" are the types of data that are most valuable to organizations. But what's not discussed is why that data is (or, at least, could be) valuable to decision-makers.
By definition, decision-makers make decisions, and if they don't have the data they need to make those decisions they are going to do it the way they always have: by their gut.
What are the key business decisions that your decision makers are trying to make? What is that one report that you could hand to the CEO and give him or her a view that makes them say, "How did we ever run the business without this?"
Start with that and you'll be well on your way to creating a data-driven organization, which is the necessary prerequisite and motivating force—not the result—of unlocking the potential value of Big Data.
But the Rosslyn report is not about Big Data specifically but rather just data in general. When asked to identify the top issues keeping their organizations from getting value out of their data, those surveyed cited not the lack of advanced analytic skills or adequate technology but rather the scattering of data among disparate systems and the poor quality of that data. For all the powerful things that Big Data technologies can do, they don't provide a silver bullet for these fundamental data issues.
Most tellingly, fewer than half the business leaders Rosslyn surveyed considered their data to be a strategic asset, indicating a lack of executive sponsorship for data initiatives.
Which, as always, reminds me of a story. Several years ago I was helping an organization implement and drive adoption of a sales automation solution. The CEO was convinced (by instinct) that the reason the sales teams weren't hitting their numbers was that the reps weren't out in the field making enough face-to face-calls. He kept hammering on the sales managers to increase the field activity, but without a way to measure how much was going on the discussions quickly devolved into finger pointing and rear-end covering.
The sales managers maintained that their teams were already knocking on every door they could. The data needed to support or disprove the activity thesis was scattered in Outlook inboxes and calendars and, more than anything, verbal reports—and the sales reps had gotten quite proficient at rattling off a stream of "updates" during weekly team calls that made themselves sound very busy.
The solution was simple. We created a basic weekly activity report out of the new sales automation system—just a simple grid that broke out, week by week, the number of prospecting emails, phone calls, and in-person visits that each sales rep on each team made. The data was only as good as what the sales reps had entered into the CRM system—and, despite training and weeks of evangelism to try to get the sales reps to record their activity, we knew that it was incomplete at best.
But that didn't really matter. Once the numbers were side by side, the CEO could immediately see which teams and which reps had a lot of activity and which ones didn't (or, at least, weren't recording their activity regularly.) It didn't take machine learning or complicated correlation analysis to see the conclusion. The leadership team knew who the top performing sales people were (they were the ones taking home big commission checks each month), and they could see right off that those same reps were the ones with the highest activity logged. The CEO turned to his VP of Sales and laughed. "How did we ever run the business without this report?" he said.
The business decision was clear: get the teams out in the field more. With the CEO inspecting the report each week, the sales managers were on the hot seat and pushed their teams to record their activity faithfully. As the numbers got better, the organization was able to identify and settle upon the right level of field activity that drove sales—and field activity rose and so did overall sales.
That's one reason why all the hype around Big Data can actually impede an organization as it tries to transform itself from instinct-driven to data-driven. If you can't get from your house to work because you live on a rutted, weed-choked dirt road, the answer isn't to sell your old sedan and buy a faster car. You probably should to start by fixing the road. And that doesn't mean transforming a dirt road into a multilane highway in one big burst but rather by fixing enough of the ditches and potholes that your car can pass over it.
When it comes to business data, fixing the road means taking the data you already have and using it in a way that has tangible business value—enough tangible value that decision-makers will want to see more of it.
Taking Hadoop and pumping a bunch of data into a massive repository and creating a lot of colorful charts and graphs won't do much if the decision-makers don't trust the number underneath the analysis, so data quality is important. Even worse, though, is if the organization's leaders aren't able to immediately see how the "insights" contained in the analyses offer concrete guidance on what the organization should do. In other words, if the data doesn't help them make decisions, they won't adopt it.
And that's where focusing on the technology first is a mistake, and focusing on the data first isn't much better. Rosslyn's survey respondents believe that "product data" and "customer data" are the types of data that are most valuable to organizations. But what's not discussed is why that data is (or, at least, could be) valuable to decision-makers.
By definition, decision-makers make decisions, and if they don't have the data they need to make those decisions they are going to do it the way they always have: by their gut.
What are the key business decisions that your decision makers are trying to make? What is that one report that you could hand to the CEO and give him or her a view that makes them say, "How did we ever run the business without this?"
Start with that and you'll be well on your way to creating a data-driven organization, which is the necessary prerequisite and motivating force—not the result—of unlocking the potential value of Big Data.
No comments:
Post a Comment