Working together with our partner, we ran a complex analysis for a client in the retail industry that focused on the effectiveness of these fields:
- Marketing channels
- Customer loyalty programs
- Price promotions
To start with, we looked carefully at the customer loyalty program in every channel, both online and off. Altogether, customers had made about 250 million purchases. We also conducted a survey of users of the loyalty program.
During the marketing channel analysis, we focused on KPI’s for, among other metrics, email campaigns. The last element requiring examination was the results of price promotions and reductions based on data from the last five years.
Print media is going, going….gone?
The results of our work paint a very clear picture of the effectiveness of traditional marketing channels:
- The popularity of familiar forms of marketing, like newspaper ads, has fallen by 40% in the last 5 years.
- Email marketing with promotional prices and product newsletters resulted in a low CTR.
- The client did not use any tools for campaign sending-time optimization.
The customer loyalty program also showed some deficiencies:
- Two-thirds of the members of the program were not aware of its basic benefits. Customers were not encouraged to take part in the program, which resulted in the benefits not being available earlier.
- The client didn’t use any information gathered in the loyalty program to personalize customer communication of any kind.
- Most price promotions were characterized by a negative ROI, in particular products from popular brands. Also, each customer received an identical offer, regardless of purchase history.
Personalization to the rescue
The Synerise team immediately saw a solution for the problem of a lack of personalized communication. Knowing that such personalization can boost conversions, we started with the newsletter and using information about customers gathered in the loyalty program. Going forward, content for individual customers would depend on their purchase histories.
The results of the campaign exceeded our expectations. The increase in CTR for the newsletter compared to the control group was 82%. Thanks to the fact that customers received personalized products in their email content based on their shopping preferences, the conversion rate increased by 30% and the average value of their carts by 22%. This was a huge step in the right direction for our client but we weren’t done yet. We had another idea that we wanted to check out.
We suspected that a personalized offer had a much greater impact on the customer’s conversion than the price incentive itself. The next step was to try a personalized newsletter that contained products not included in the discount, that is, items from outside the promotional leaflet or new products in the assortment.
This experiment yielded similar results to the first. What conclusion can we draw from this? In the context of customer conversion, personalization of the offer has more impact than a price incentive. This, in turn, creates an opportunity to reduce the costs of promotions, which results in an increase in other financial ratios.
Will it sell?
Taking a step further, we proposed a solution based on the personalization of discounts for loyalty program participants. Through the use of anonymous transactional data, information from the loyalty program and the use of the discount system, each customer received price promotions tailored to their shopping preferences. They could be collected in print or via a mobile application.
In the selection of discounts, we used an algorithm that not only chose the best product for each user but a product that at the right level of discount would increase the value of the basket. For example:
If a customer bought one product – say, Coca-Cola – during any visit, offering a discount would not increase the Net Revenue value. The amount of discounts in the newsletter is the same for everyone. In our case, the amount of the discount was dependent on the probability of purchase and the expected value of Net Revenue. Customers received discounts on other products that were potentially interesting to them, as indicated by our algorithms.
It seems that, thanks to artificial intelligence algorithms, the well-known concept of promotional newspapers has gained a new dimension. While the standard newsletter contains more than 100 products every week, we delivered seven or eight individually-selected products, only those most relevant to a particular customer.
How far can you go with personalization?
In this case study, we presented the benefits of personalization and using AI algorithms in the retail branch, but they apply to practically any business area. Collecting customer data is a must in today’s business world and any database can be used to provide intelligent, personalized communication.
Artificial intelligence and real profits
We’re now able to collect more data than ever before but the human mind still has its limitations. Our ability to process the incredible amount of information simply can’t catch up. This is where Artificial Intelligence comes in. Algorithms can execute operations and identify patterns on a scale that is almost unimaginable without the help of technology.
We can now achieve a level or personalization that only algorithms can deliver. Even though we call it “artificial”, AI helps us to better understand how we think – and buy. Put the power of AI to work for you and see the real profits it can bring.