Catalyst

Computing the true value of a customer

MAKRAND JADHAV | Updated on January 24, 2018 Published on July 24, 2015

Getting it right There are a number of methods to determine the worth of your customer PATHDOC/SHUTTERSTOCK.COM

Who is your “valuable” customer?

Most organisations will answer this question in revenue terms. A retailer will most likely say it’s the one who spends the most money with him in a calendar year. A telecom service provider would say it’s one with the highest ARPU (average revenue per user), a bank may say it’s one with the highest AQB (average quarterly balance) or two-three active accounts, and an airline may say it’s one with the highest frequent flyer miles in its account. All very valid.

Let’s examine some other techniques used to determine the most valuable customer. One of the largest banks in Canada decided to approach this question rather differently: ‘Can I determine how much profit each customer is contributing’? Needless to say, not a very straightforward or easy task. On the contrary, quite data-intensive and laborious to begin with. After a lot of discussion and deliberation it adopted the following approach to measure customer profitability:

Customer profit = Net Interest Revenue + Other Revenue – (Direct costs + Indirect costs) – (Risk provision)

Obviously it first had to start with unique customer identification and associating multiple accounts with a single customer. It then applied the above to each piece of the customer-level information. What came out was a true measure of a customer’s value to the bank. It then chose to run this across its over-four million customer base. So even if you are a customer with a moderate AQB but transact less and mostly through electronic channels, you may be more valuable compared to a high AQB customer but one who soaks a lot of RM time. This approach formed the basis of its customer communication/promotion programme resulting in $7.5 million in incremental revenue and 15 per cent increase in “high value” customer volume.

Life time’s worth

Another measure deployed is Customer Life Time Value (LTV). Here the organisation tries to estimate the potential earnings that it may accrue from a customer based on the length of his stay with the organisation and his potential to buy more of its products over time. In insurance parlance it will use a formula like:

Customer Life Time Value = Value from current policies (Premium value of current policies + Expected tenure of current policies) + Value from future policies (Probability of repeat purchase + Expected value of future policies + Expected tenure of future policies)

This requires statistical capability to determine the units for the “Expected” parameters, and relevant historical data (payment history, product holdings, channel of acquisition, product category, etc.) to develop it. Such an approach is also common with banks, and telecommunication companies.

Another common measure used, especially in retail and e-comm is RFM - Recency, Frequency, and Monetary value. In this case, the recency of customer visits/transactions is measured/classified over a certain period of time (week/month/quarter), i.e., when was the last transaction done – one week prior, one month prior or two months prior? Next, the Frequency is calculated. Heavy is anyone who does 10+ transactions in the period, Frequent (7-10 transactions), Moderate (3-6 transactions), Limited (1-3 transactions). As for monetary value, if the value per transaction is ₹1,000-1,500 the customer could be classified as Low, someone spending ₹1,500-2,500 could be Medium and ₹2,500+ would be High. This will give 36 classifications where each customer can be plotted. As a final measure each customer could be classified as VIP (High in monetary value, Heavy in frequency and with the most recent transaction not more than the last week) or Dormant (Low in value, Limited in transactions, and a dated last transaction), and treated accordingly.

A holistic measure

A more simplistic approach is the one similar to the one I described in my previous column on Distribution – composite score (see cat.a.lyst edition dated July 10). Identify the top characteristics you want your customer to display – revenue, complaints, payments, tenure, number of visits/transaction, value per transaction, profit, channel, number of products ... the list can be endless. Then weight each of them as per your business objectives. Apply it to your customer base. What you will get will be a composite measure/score for each customer ranked on the parameters identified by you. You may change the parameters or weights on a periodic basis to reflect the business priorities and the score will reflect the same.

As industries move from an acquisition phase to more mature and competitive phases (read telecom, developed economies) organisations will have to decide which customers to focus on. The above measures can assist in delivering custom treatments based on the customer’s overall value to the organisation, via the channel of his preference. Engaging with the “valuable” customers is deeper, relevant, and timely. Needless to say, it leads to better wallet and market share.



MAKRAND JADHAV IS CO-FOUNDER & COO, KLOUTIX SOLUTIONS

Follow us on Telegram, Facebook, Twitter, Instagram, YouTube and Linkedin. You can also download our Android App or IOS App.

Published on July 24, 2015
This article is closed for comments.
Please Email the Editor