How optimised direct mail saved Mother’s Day at Valentins
In this guest post, we highlight a use case that shows how Valentins used CrossEngage AI to significantly optimise a direct mail campaign. The result: an impressive €34,400 increase in revenue. Learn how this optimization was implemented in a very short time and what key insights it yields for companies looking to improve their marketing strategies.
Flowers per order: Who is Valentins?
Valentins is a pioneer in online delivery of fresh flowers and gifts and has been delighting private customers, companies and cooperation partners alike since it was founded in 1999. With a dedicated team of 55 employees, the company has made a name for itself by offering hand-tied floral arrangements perfectly tailored to occasion-driven purchases. From Valentine’s Day to Mother’s Day, Father’s Day and birthdays, Valentins aims to make it easy to give flowers as a token of appreciation.
Why optimise direct mail?
Direct mail is an effective advertising medium that can achieve high conversion rates. In addition, it remains in the household for a long time and can even generate revenue weeks or months later, thus generating sustainable contacts. To achieve this, direct mail must be managed like a performance channel. What levers did Valentins use with CrossEngage to achieve this?
- Comparatively high costs per mailing
- Estimation of optimal circulation (expected contact value vs. unit costs per contact)
- Customer lifetime value (CLV) model
- Ranking according to expected contact value
- Usage of CrossEngage Scores vs. regular selection
The challenges at Valentins
Estimation of the circulation size
Previously, marketing was based on a rough selection of those customers who would have a higher expected contact value vs. unit cost per contact.
Inaccurate, time-intensive customer selection
On the one hand, this approach was imprecise for customer selection and, on the other, required a great deal of operational effort.
High advertising costs
Accordingly, large print runs were printed and the cost per mailing was relatively high.
1. Make data-based decisions
Using AI to identify patterns in purchasing behavior and thus forecast customer sales trends.
2. Future purchasing behavior as a basis
From now on, the print run can be optimised and controlled on the basis of data. The basis for this is the forecast of future purchasing behavior, which enables a reliable assessment of the profitability of the mailing (for each customer individually).
3. Strategic resource planning
This thus enables more efficient use of resources in the form of lower circulation and costs for direct mail.
A race against the time
To ensure timely delivery of the Mother’s Day direct mail, a race against time began on April 1: Mother’s Day was approaching and the mailings had to reach existing customers on time. Just three weeks passed from the time the contract was signed until the mailings were delivered. Not much time to save Mother’s Day.
Details of this direct mail campaign for Mother’s Day
The goal was to send out a direct mail campaign with a circulation of 40,000 that would allow ranking by expected contact value. For this purpose, a machine learning model for a product-unspecific customer lifetime value was created with just a few clicks. The target group was all existing customers who had made at least one purchase to date. The evaluation period extended from the postal delivery date (21.04.) to Mother’s Day (09.05.). We conducted an A/B test with Valentins.
The test design
For the A/B test, the total potential (all customers with at least one purchase) was randomly divided. In the end, 20,000 recipients each were selected via Valentins’ previous approach and 20,000 via the CrossEngage scores.
The A/B test was evaluated by comparing the selection group from Valentins with the selection group from the CrossEngage scores. This targeted selection from the scores generated an uplift of 13% in sales over the total group.
If you’d like to learn more, feel free to check out our partner CrossEngage.