By Stéphane Amarsy,
Chairman of the board of Splio + D-AIM

If we want to boost our knowledge of each individual customer, we must make experimentation a priority. This can only be achieved by collaborating with algorithms and assessing the long-term customer relationship.

Move beyond the obvious

While the quest for individual knowledge of consumers is vital, we must also examine their behaviors from external factors.  This approach provides additional information about what makes consumers tick. For example, an American pizza company noticed that its demand forecasting algorithms were extremely accurate except at a specific point of sale. For two years, things had been going haywire over the same 15-day period.

There was no concrete explanation for this curious phenomenon. An investigation into the matter revealed it was due to works being carried out on the pizzeria parking lot—a detail that hadn’t been incorporated into the algorithms. After some tweaking, the quality of the algorithm increased considerably, followed by the company’s turnover. The raises the question of how knowledge can be exploited for the best results.


Make experimentation a priority

Experimentation is one of humanity’s greatest assets. Algorithms, on the other hand, lack the ability to innovate.  That’s where human-AI collaboration comes in. It lets humans focus on their added value while algorithms do the same. Searching through external data can be time-consuming, especially with open data. It has to be organized— at least until systems are capable of accessing all the knowledge of open data with true interoperability. For that reason, prioritizing must involve experimentation. But working on every customer isn’t necessary. It’s better to carry out tests, analyze the results and generalize if the experiment is successful. You can wait for a great idea or plan the process, but the best way to identify research areas is to monitor technological developments and hold divergent thinking sessions.


Evaluate marketing contributions

Marketing expenses are more often seen as an operational cost and not as an investment that depreciates. The approach we take today creates tomorrow’s value. But how can we judge the value of such an investment and quantify the results? A long-term evaluation of marketing contributions is essential, something that current accounting systems aren’t good at factoring in. We must take a closer look at the customer’s long-term value. Instead of adding up how much the last transaction made, we must apportion value to long-term customer relationships with each individual.

For example, when consumers buy an airline ticket, their value isn’t just the ticket’s cost (with the possibility of added-value from services, of course). It also involves considering the likelihood that this person will continue to buy airline tickets. This is what customer lifetime value is all about—the total worth calculated, according to the estimated lifespan of customer loyalty. Return on investment is something that comes with the benefit of hindsight.


Which businesses will be capable of taking a long-term approach? And what revenue will they get by building solid customer relationships? Having the capacity to do both provides a vital competitive edge and protects consumer insight and its operational use from financial scrutiny.


Want to know more?