By Nick Harrison and Deborah O'Neill
This article first appearded in Harvard Business Review on August 25, 2016.
Customer insight, segmentation, and behavior tracking have proliferated in recent years. But their impact on sales has been underwhelming, primarily because they ignore the needs of one key constituent: the frontline employee working to make the best possible marketing decisions day by day.
Across industries, staff such as retail category managers, sales representatives, financial advisers, and wealth managers are awash with reports and insights that comfort their companies’ top executives and by making them feel that they are leading a “customer-centric” organization. But managers in the trenches often describe the data in these reports as unhelpful, contradictory, and distracting. Worse, they become demoralized when centralized “black-box” solutions and algorithms make strategic decisions for them.
In theory, deep insights from customer big data should enable highly skilled employees to be more creative and free up time to connect with customers in new ways that add value. In practice, however, many lose interest in visiting stores and talking to customers, instead spending most of their time in front of their computers. Some just give up entirely and look for a different role.
The trick is to deliver the end goal of customer-focused decisions by delivering not insights on customers but customer key performance indicators at the product level.
There is a way to use customer insight data to strengthen, rather than weaken, connections to customers. We’ve observed businesses make big improvements when they strike a balance between the creativity of their people and the science of sales. What made the difference?
Five broad rules enable managers to move from working for customer data to having the data work for them.
Determine decisions first, data later. Most managers know that data is not an end in itself: It must serve the business. Yet, curiously, many managers rush to collect all of the data they can or the insight they find most interesting, rather than analyzing the decisions that need to be made and working backward to decide what data, analytics, and insights will help.
This is a potentially fatal mistake. To be sure, most major retailers find a wide range of customer segmentation data helpful in understanding customers’ lifestyle habits and needs. Revelations from this data can shape business priorities. But this data is a far cry from the specific information that retail managers need to figure out which products to put into a particular store — one of the most important frontline decisions in an organization.
Retail leaders such as Kroger and Tesco overcome this challenge by focusing on collecting exactly the right product-focused insights to drive these crucial assortment decisions, such as the item’s selling power, or its “incrementality” (which shows whether customers loyally purchase the product week after week or happily switch between alternatives). The trick is to deliver the end goal of customer-focused decisions by delivering not insights on customers but customer key performance indicators at the product level — because those are the true deciding factors for assortment.
What’s worse, putting a black box in place disempowers category managers and may rob them of any motivation to correct the mistakes of the algorithm.
The same principle applies in other areas of retail. For example, local weather forecasts, event calendars, and delivery schedules assist with carrying the optimal mix of inventory in each store. They enable stores to seize chances to sell more ice cream during a heat wave and higher-margin snacks to students when big sports events happen at nearby stadiums. When a cereal’s sales suddenly slump, store managers can see if it’s because the product is no longer popular or because a delivery truck never arrived. The approach is the same: Don’t just supply insights that may look interesting. Ensure that insights are “shaped” around the decisions that must be made.
Empower, don’t automate. Marketing and commercial decision makers often complain that customer analytic tools developed to improve their productivity actually make them less effective. More often than not, this is because the company has built a centralized, impenetrable black box that makes decisions for people rather than helping them make better decisions themselves.
But data analytics has blind spots. It can’t take into account all considerations, and, by definition, it’s based on history. Making decisions based on customer analytics alone is a surefire way to become a prisoner of your past.
Continuing with the retail management example, trying to make assortment decisions based purely on analytics misses many important insights that only the category managers can provide: discussions with suppliers about next year’s trends and promotional events, predictions about what’s going to be “hot” next season, new product ideas and developments, and broader impacts on the market. What’s worse, putting a black box in place disempowers category managers and may rob them of any motivation to correct the mistakes of the algorithm.
Contrast this with a system that serves up the correct key performance indicators in the right format to help frontline managers make better decisions so that they can make the final call. In this situation, decisions are improved because even if data challenges gut feelings, there is still space for creativity.
Today nearly all of the account managers receive business insights from the tool and are spending about 20% more time with their clients.
Design “with the users, for the users.” Once a company decides to try out an insight tool, it should keep its first data project small and focused on frontline decisions, that way participating managers can be engaged and help to ensure its success. We find that small, test-and-learn data projects on 90-day cycles help to steer clear of committee consensus. Incremental changes also help to surface errors, which can be picked up more quickly and fixed without losing the project momentum or stakeholder buy-in. After achieving sufficient buy-in, the project can be easily scaled.
That’s exactly what one bank is attempting by encouraging its managers to spend more time with clients and less time poring over research. The bank started out working with a small number of account managers and clients to identify developments that would improve the business. It then developed a system that would send these account managers real-time alerts about potential opportunities. When oil prices dropped, the system alerted managers to call their top five clients whose businesses would be impacted by the change. Over multiple 90-day cycles, the company was able to try out new alerts, remove the ones that weren’t working, and continue to refine the tool as it was rolled out to more managers. Today nearly all of the account managers receive business insights from the tool and are spending about 20% more time with their clients.
Continue to remove distractions. Companies must continually review customer data reports with managers on the front lines to be sure that they are restricted to those that provide actionable insights. They must resist the temptation to get more and more reports out of the system just because they can.
An effective report is one that leads to measurable benefits. Using that lens, one major European grocer axed half of its sales reports because they didn’t provide actionable information. Managers loved receiving a weekly list of the 10 best-selling products — but it turned out that no one was making decisions based on it.
So the grocer’s chief information officer replaced the list with a more detailed one of the top 10 items in local store groups. Category managers took actions based on this report because they could see what some stores were selling well and determine whether a product should be stocked on their own shelves.
When the right insight reaches the right hands on the front lines of a business, the results can be truly transformational.
Build analytical capability and culture across the organization. Finally, even successful data projects will fail eventually unless a company builds an organization that can continue to refine and maintain the new approaches, so that demand for them spreads across the organization.
Leading consumer-facing companies are putting huge efforts into building their own customer insight and analytical organizations — particularly those based in big cities such as London, Paris, or New York, which can access top analytical talent. For example, one leading retailer in London recently formed a new analytics and digital division, led by a senior executive. The new group is now attracting top customer analytics talent, in part because it operates and feels like more like a startup than a blue-chip corporation, right down to its office layout. Another London-based retail bank has set up a separate business unit with its own profit-and-loss statement. The unit’s mission is to improve the performance of the bank’s core business and to develop new business streams, all via the use of customer data and analytics.
In addition to establishing these “centers of excellence,” both of these companies are simultaneously using their best people to help bring the new approaches into the front lines. That’s because they realize that their new groups’ customer data and analytics must work well for the frontline staff, or else their companies will revert to their old ways of working.
It’s often said that the devil is in the details. When the right insight reaches the right hands on the front lines of a business, the results can be truly transformational. Carefully identifying and executing your company’s customer insight mission in consultation with frontline managers can make all the difference.
This article is posted with permission of Harvard Business Publishing. Any further copying, distribution, or use is prohibited without written consent from HBP - firstname.lastname@example.org.