4 ways to create a seamless customer experience using data
Saken Kulkarni | May 28, 2014
In my neighborhood in New York, I often see retail employees handing out coupons. Whether it’s raining, snowing, or 90 degrees outside, these guys can’t go home until they’ve handed out every last coupon.
Unfortunately for them, this marketing method is rarely an effective way to reach customers. Let’s take a frozen yogurt chain as an example. How can the employee handing out coupons know which customers live in the neighborhood? What if 80% of the coupons that she is handing out are going to people who are simply passing through the neighborhood? What if 60% of the coupons are going to people who are lactose intolerant?
Unless this retailer uses data to drive its marketing decisions, it’s wasting its marketing budget. Retailers can no longer rely on blanket messages in their marketing campaigns. Tailoring marketing campaigns to the customer, similar to the method of tailoring experiences to the customer, will separate retailers who lead from those who fall behind. In a challenging economy beset with competition from online-only retailers, profit margins can be razor thin, forcing retailers to trim costs wherever possible. Optimizing marketing spend to focus campaigns and promotions on customers most likely to respond can help increase margins.
The popularization of the electronic point-of-sale application in the 1980s opened the door to data-driven marketing by enabling retailers to identify and track customer purchases. The explosion of social and digital data created a prime opportunity for retailers to better understand their customers and optimize their marketing campaigns. There are several important metrics that can create a data-driven marketing culture. Here’s four metrics to consider, which include both established, tried-and-true techniques and more experimental methods.
You may remember that I referenced Professor Mark Jeffery from the Kellogg School of Management in my post on designing your customer analytics methodology . His definition for marketing NPV is very similar to that of Customer Lifetime Value. Think of marketing like a financial security: marketing NPV measures the value of a marketing campaign, adjusting for the time value of money.
C^0 = Startup marketing cost
B^n = Revenue (or cash benefit) from the marketing campaign
C^n = Cost of marketing campaign
r = Discount rate
When NPV > 0, the marketing campaign should be executed, and when NPV < 0, the marketing campaign should not be executed. Treating marketing campaigns like you would a financial investment creates an analytical approach for marketing execution.
Conjoint analysis has been used since the 1970s, but its popularity is still strong today. Conjoint analysis determines how people value different features that make up a product or service. For instance, an electronics company might interview a set of respondents and provide them with a list of different tablets with potential product features (e.g., price, GBs of storage, etc.) and measure which features respondents choose. Several software programs automate and decrease the manual effort involved in this process. This analysis helps marketers understand how they should design products and features.
Diffusion is the rate at which an idea or product is spread from one customer to another. There are several methods for estimating diffusion, but a classic example comes from marketing professor Frank Bass called the Bass Diffusion Model. This model estimates the rate of diffusion by segmenting adopters into “innovators” and “imitators,” and deduces that the diffusion rate is heavily dependent on the interaction between the two segments.
Net uplift modeling
Imagine that a national footwear chain launched a “Friends & Family” discount deal in the month of April, and a monthly report shows that sales increased by 15% in that month. From these results, the marketing team may believe that the marketing campaign was successful simply because sales increased. Unfortunately, this is untrue. We don’t know if sales would have increased by 15% if the marketing campaign was not executed. In other words, we cannot attribute the cause of the sales increase to the marketing campaign because it could have been caused by something totally different (like the Easter holiday or strong consumer confidence).
The best way to isolate the marketing campaign variable is to use the net uplift modeling to measure the incremental impact of a marketing campaign. Think of this type of analysis as a double-blind prescription drug study, where the test group receives a drug, the control group does not receive the drug, and the test group is measured against the control group. In the case of a marketing campaign, the test group would be customers who received the promotion and the control group would be those who did not receive the promotion. The “true response rate” would be the difference in response rates between the two groups.
Use analytics to drive marketing decisions
In times of inconsistent economic growth where downward pressures on margins put marketing budgets at risk, it is critical to use analytics to drive marketing decisions. Rest assured that this type of rigor will not stifle the creative forces at your organization. Instead, it will help tailor and focus creativity to customers who are most likely to respond.
Saken Kulkarni is a business analytics consultant at Slalom Consulting. He focuses on the intersection of customer engagement, analytics, and data visualization.