Congratulations – your new customers have successfully completed their first purchase. Now that your customer has had her first positive experience with your brand, you should ensure that she becomes a second buyer so that initial customer acquisition costs pay off and regular repeat purchases become a habit. The following article on “Second-Order Push” explains how this can best be achieved and the role physical mailings can play in this.

When is the right time to initiate a second purchase?

There is no general answer to this question. At first glance, a static welcome route in which all new customers are equally addressed with pushes over a certain period of time seems to be the simplest solution. Often companies proceed even more rudimentarily and approach new customers who have not made a second purchase simply, for example, four times a year, within the framework of “second purchase incentive initiatives”. However, this results in unnecessary costs and the ROI is reduced for no reason, since not all customers have the same potential for a second purchase after the initial purchase and the time period until the meaningful stimulus can vary greatly. The following chart shows the total number of second purchases over a 24-month period at a fashion retailer. You can see here that second purchases can often still take place after 6 or even 12 months. Of course, these considerations always depend on the product.

Classical cohort analyses for new customer segments generally show that only around 25 – 30% of new customers make a second purchase in the second year. In the subsequent years the decline is much lower and may even stagnate, as the following chart shows:

One thing becomes very clear here: If you manage to reduce the loss rate in the first year, you can drastically increase customer value. Therefore the Second-Order-Push is of immense importance for your CRM activities.

How do you know when a customer has the highest potential for a second purchase?

With the help of predictive models, such as those created by the company Gpredictive, you can very well determine which customers have the highest potential (value) for a second purchase at a certain point in time. As a rule, you should check this value again and again at intervals of 30 days. In this way, you create a tightly meshed network of new customers. The system checks again and again whether a customer has a sufficiently high value so that a kick-off is most likely to be worthwhile.

As a rule, only the best 10-20% of a new customer cohort that has a sufficiently high value at the time of an initiation is worthwhile. The following figure shows the distribution of customer values. The values are divided into so-called deciles. All customers in this cohort are divided into 10 equally sized groups. The groups are put together systematically so that only the 10% best customers are found in the best decile, and so on.

The light bar represents the predicted conversion rate. The dark bar represents the actual conversion rate after the launch. The effect that only the best 10% to 20% are really worthwhile is clearly visible. Over time, customers wander through the deciles. For example, a customer who sends out signals early on that he is a seasonal buyer may be found in the weaker deciles in the first months after the first purchase. However, when his time has come, this customer will jump into the top decile. And that is exactly when he needs to be approached, agile and without delay. With a relevant incentive.

How do you set the right incentives?

If you know when you want to address which customers, the impulses are given in the usual CRM channels (e.g. email, social media, apps and physical mailings), depending on the product and business model. The important thing here is that you tailor your messages individually (as far as possible) to the needs of each customer. This is especially true for physical mailings such as letters and postcards. Since no opt-in is required for postal address, you can also reach valuable customers in this channel who are not reached via online channels.

Manually determining the right customers every 30 days and then addressing them manually with personalized mailings is, of course, of little use. In the data-driven CRM world, these processes can be easily automated with solutions such as Gpredictive and optilyz – because the effort to implement this manually would be too great and would make the procedure inefficient. Once your automation environment is set up, you “only” have to monitor and of course constantly think about further optimization.

Practical example: second-order push in the fashion vertical

As the operator of an online shop for fashion, you want to persuade your new customers to make a second purchase at the right time with personalised mailings. For this second-order push, for example, you decide on a maxi-postcard because this format conveys the selected message excellently.

The slogan is: “Thank you for your first order! Here is your exclusive new customer discount code ‘IQAXSKF’ over 15% for your next order”.

Since you offer different collections for women and men, you decide to segment this group by “female” and “male”. The design for female first-time buyers shows impressions from the women’s summer collection, the one for male first-time buyers shows impressions from the men’s summer collection.

To automate this process, you have connected optilyz to your system. In this way, all necessary customer data for the segments is transferred automatically. You can automate this process with a volume trigger campaign. As soon as 4,000 addressees (male/female) meet the criteria “first-time buyer, high CLV and high conversion chance”, the maxi-postcards are sent automatically. This ensures that you benefit from low-cost dialogue postage on the one hand, while at the same time not allowing too much time to pass until customers with a high current purchase probability are addressed, thanks to the small batches. This is because purchase probabilities often decrease over time and thus also conversion rates. The approach “four times a year we activate our first-time buyers” is a relic of the past.

The most important facts at a glance

• Create a cohort analysis of your new customers.
• Use a predictive model to determine the 10-20% best customers in a period. You should repeat this step every 30 days.
• Address these customers with personalized messages.
• Also target physical mailings to these customers in an agile manner and at the right time.
• Automate these processes to minimize the effort.