Traditional marketing wastes money. Think about it. Every campaign you run creates three unwanted outcomes:
- The Sure Thing Problem: Wasting money targeting people who would buy even if they weren't targeted;
- The Lost Causes Problem: Wasting money on people who are highly unlikely to buy, and;
- The Do Not Disturb Problem: Actually driving away business by targeting people best left alone.
This last response may come as a surprise to many marketers. Research suggests that customers who, over a period of time, receive several irrelevant offers are often prompted to either "turn off" or act upon their dissatisfaction by opting not to respond to marketing promotions. Untimely promotional offers can also turn an impulse purchase into a deliberate and considered purchase, turning the customer away from a simple renewal or purchase to a studied review of competitive offerings.
In a real life scenario this could be a customer who recently got turned down for a line-of-credit loan at his local bank and then shortly thereafter received marketing for high-interest loans. It quickly becomes clear to the customer that the bank is not acting in his/her best interest.
Most marketers today do not take into consideration the possibility that a campaign may prompt customers or prospects to leave or look for alternative solutions. Take for example a customer of a mobile phone service who, towards the end of his service term, receives a message from the retention department encouraging him to renew his contract.
While this marketing intervention may be well timed for some customers and the promotion serves as a great incentive to renew, for others it may in fact prompt them to look for a competitive service contract with more attractive terms.
These types of responses are measurable and virtually every customer experience management campaign suffers from them. However, traditional analytics' inability to identify the people that fall in to these response categories leads to incredible waste in a campaign. Campaigns cost more than they need to, achieve less than they should and simply annoy some customers, driving them away. In an ideal scenario, marketers should be able to identify only those whose behavior they are able to influence positively.
New segmentation modeling techniques, called as uplift modeling, incremental modeling, differential response modeling enable marketers to determine before executing a campaign, between those whose behavior they're most likely to impact and those they won't. Unlike traditional response modeling that looks at the 'probability of response,' uplift modeling looks at the predicted incremental response by modeling the difference in behavior between the treated and control groups to find the people most affected by marketing intervention.
The ability to discern those whose behavior you're most likely to change helps dramatically increase campaign profitability by allowing you to target fewer people, (ignoring those who can't be swayed) and generate greater results by eliminating negative effects (those who are prompted by the marketing outreach to look for competitive offerings). An added bonus is the additional campaign resources it frees up. You can actually spend less and make more by overcoming marketing modeling myopia and integrating uplift modeling throughout your marketing campaigns.
Savvy marketers are quick to catch on...
Many industries are moving on from traditional modeling in favor of uplift modeling. The financial services industry is among the first to test the worthiness of this new approach to segmentation and publicly acknowledge the value derived from switching to uplift modeling.
Increasing the length of customer relationships is an important issue in financial services. This is a competitive marketplace that is flooded with commodity service offers which make a compelling case for high-precision campaign segmentation and targeting strategies that help increase cross-sell and customer wallet-share.
The value of many banking products is high, so even an increase in product sales of a fraction of a percentage can provide a positive return on investment for direct mailings. Leveraging the latest techniques in uplift modeling, one of the largest banks in the US discovered that it could increase total sales by about 10 percent while reducing mailing volumes by 60 percent. These results more than doubled campaign profitability compared with the bank's previous approach to modeling which they regarded as state-of-the-art.
Overcome myopia and improve your ROMI
A myopic view of modeling and of customer response segments has lead to incredible marketing campaign expenditures with low return on marketing investment. Embedding uplift modeling into the campaign planning process will enable even the most successful marketers to see marked improvement in their ability to target the most profitable customers and spend less money in the process.
About the Author
Mark Smith is an executive vice president at Portrait Software. He is a 15-year veteran in the customer analytics and data management industry. He has written numerous papers on the impact of new technology to consumer marketing operations, and speaks regularly at industry and technology conferences.