This article takes a thorough, high-level look at how our iGaming and Forex clients perform customer segmentation to target their marketing efforts and thus increase spend and customer lifetime value. The strategies discussed here are applicable to any company or website with paying customers, with or without our software.
In the old days, companies sent the same marketing campaigns to all of their customers. Later, marketers began to understand that even simple grouping of their customers allowed them to run more relevant, effective and profitable campaigns.
More recently, sophisticated marketers and retention experts have discovered that applying the latest technologies to this challenge deliver far more targeted – and profitable – campaigns. The reason is clear: when each individual customer (or small groups of very similar customers) receives a personalized and highly-relevant offer, there is much more chance that the message will resonate with them. In a world of marketing overload, customers have little patience to hear what you’re telling them. So it’s critically important to tailor your messages and incentives so that they are the most relevant and interesting for each individual customer.
At the heart of relevant, personalized marketing is accurate and fine-grained segmentation of your customers.
Segmenting Customers into Lifecycle Stages
The first step is to divide your customers into distinct lifecycle stages. For a typical iGaming site, the important stages for retention marketers are the New, Active and Churn stages. Customers in each stage behave differently and thus the way you market to them will necessarily be different. The next step is then to dive into each lifecycle stage to identify interesting groups of customers within them.
We define the New stage as customers who recently made their first deposit, typically within the past two weeks. These guys are checking you out, there’s no trust yet and they are relatively sensitive to their experience on your site.
Therefore, the best way to approach them is to give them a good experience, e.g., by giving them extra bonuses or providing them with exclusive game play opportunities.
Within New customers, our clients use Optimove to identify two sub-groups who need to receive extra attention. The first are “quick-churn” – these are customers who lost their initial deposit very quickly and didn’t return to the site. They require relatively aggressive treatment to “repair” the bad experience they had and encourage them to give it another try.
We call the second sub-group “new with high potential” – these are the customers with the highest potential value. The best way to identify this group is by using customer lifetime value forecasting, which uses customer modeling technology to incorporate both behavioral and demographic data to predict how a customer will likely behave in the future. (This is one of the most important abilities of our Optimove software.) If you don’t have access to accurate customer modeling methods, a simpler (and less accurate) method is simply to take the top 5%-10% of New customers in terms of either deposit amount or “net gaming” (net revenue). From our experience, it is most effective to give this group VIP treatment; long-term incentive programs tend to keep them engaged for extended periods of time.
Active customers are those who have played on the site for an extended period of time, e.g., more than two weeks, and continue to do so.
Firstly, it is valuable for iGaming marketers to divide their Active customers into two groups: those who make withdrawals and those who don’t. These are two completely different types of customers. While the withdrawers tends to be more active, making more frequent deposits and playing more frequently, it is the second group that tends to be more profitable for the company. An important practical implication of this difference is that the company’s bonus policy can be much more generous with the non-withdrawal-takers: the bonus monies provided will most likely return to the company.
Next, you want to perform cluster analysis on each of these two groups in order to identify small groups of customers with similar behavior patterns. Commonly available software packages can help you perform cluster analysis using basic models, such as RFM (recency, frequency, monetary). For example, this lets you quickly differentiate between profitable players who play every day and less-profitable players who only play once a week.
Another way to segment your Active customers is to look at their average deposit amount. For example, you might want to target active customers who frequently deposit $20 with a match-up offer if they deposit $30. This will encourage this group to deposit more, but you wouldn’t want to send such an offer to customers who typically deposit $100.
Our Optimove software incorporates a very deep, multi-layered cluster analysis for greater accuracy in generating very small clusters, but the details of that are beyond the scope of this article!
Customers in the Churn lifecycle stage have been inactive for an extended period of time. Churn stage customers also fall into two distinct groups: those who made a single deposit and quickly became inactive, and those who were more active for a longer period of time. The first sub-group is the most difficult group to reengage and thus usually receive the most aggressive offers. The second group can be further sub-divided into those customers who still have a remaining balance at the site and those who don’t. The former group can often be reengaged with a simple reminder about the outstanding balance that they can still use to play, while the latter group will require more aggressive encouragement to return and deposit more money.
Interesting Customer Segments
- About to Churn – our customers have found that it is very valuable to try predict which currently Active customers are about to churn. Retention efforts on this group can be ten times more successful than trying to reengage them after they have churned. The most well-known model used to identify these about-to-churn customers is the “decision tree” model which assigns a probability of churning to each customer based on behavioral data. We work with a more sophisticated and accurate proprietary model which we have baked into our Optimove software.
- Back from Churn – At Optimove, we place significant emphasis on a group we call “back from churn” – these are customers who were inactive for a period of time and then became reengaged with the site. We’ve discovered that, in many ways, these back-from-churn customers behave similarly to customers in the New lifecycle stage. This provides additional opportunities to retain them and increase their spending. (You can read more about this at our back-from-churn blog post.)
- VIPs – Typically, a very small percentage of a company’s active customers are big spenders who make large and frequent deposits and who play often. It is important to treat each of these customers as a “segment of one” and to give them one-on-one, highly personalized attention.
- Bonus abusers – These are customers who find loopholes in a company’s bonus policy and exploit them for their own benefit. It is important to identify these bonus abusers and to prevent them from receiving offers which could work against the company. In extreme cases, these customers can be blocked entirely from using the site. There are different ways to identify these customers, based on the identification of unusual indicators, such as a high bonus ratio or negative revenue. Once such indicators are observed, the site operator will typically need to manually track the activities of individual customers to make a definite determination that they are bonus abusers.
The most effective marketing is personalized marketing, and successful personalized marketing requires fine-grained and accurate customer segmentation.
Achieving this level of segmentation is not simple – it requires day-to-day tracking of a wealth of information about your customers and their activities. Armed with this intelligence, the retention marketer is half way towards the goal of determining the most effective marketing actions to run on each user. The second half is the ability to test, track and measure the effectiveness of the marketing actions run on each user in order to optimize and maximize the impact of every marketing action.