Lifetime Value: Optimal Resource Allocation

Anonymous Duck
Member

Posted 05-Feb-2006 10:50 AM
Dear all,

in CRM circles there has been much talk about a metric called "Life-Time Value" and how it could be used to make better decisions. (for those who are not into it: LTV equals the net present value of every cash-flow expected from a customer over his life-span)

In your experience, can you use it to decide how marketing resources should be distributed among your customer base?
That is, if you realize (as it is often the case) that some customers account for a disproportionate % of your total profits (both current and future, we're reasoning in terms of LTV!),does it make sense, FINANCIALLY, to invest more on them? Let's say you have 100 customers, 2 of whom account for 40% of your total profits. If you preach by the "all customers are equal" rule, you'll retention budget (which might tale the form of a loyalty programs, a members-only magazine, etc.) will be split evenly among customers (thus, your 2 customers will get 2% of it). But if you want allocation to reflect differences in terms of LTV, those same 2 customers should get more: 40%, perhaps?

Which one makes sense and... why? (again, financially).

P.S. This same question has been posted on the MarketingProfs webboard.

Graham Hill
Guru
Member

Posted 07-Feb-2006 01:49 PM
Customer value in terms of customer lifetime value (CLTV) or customer equity (CE) is a difficult concept to implement. CLTV is hard to measure, is hard to interpret and is not always as straight forward to implement with front-line staff as it would appear.

However, it is a very valuable GUIDE to allocating marketing and other resources. It can be crudely used to ensure that high-CLTV customers get the lion's share of marketing resources spend on them. But that shouldn't be done in a way that the majority of lower-CLTV customers get no resources at all. Or that high-value but very short-lived customers are overlooked.

At the next level, it can also be used in conjunction with return & responsiveness measures to target resources where they will generate the highest return. It may be that there is a higher return to be had from moving mid-CLTV customers up the value-scale than from high-CLTV customers who are already spending everything in your catagory on you already.

At the next level still, CLTV can be used to construct portfolios of customers with different CLTVs that collectively optimise their total return - high-CLTV customer with high-risk may not be morth much more than a low-CLTV customer with low-risk. This is the CLTV "Holy Grail" of treating your customers as (financial) assets. This generally requires the calculation of segment betas which bring risk into the CLTV equation. This is an area still in development in even the most advanced users of CLTV.

On the negative side, it can also be used to "fire" customers whose CLTV is negative. This is a contentious issue and one that companies should only take after a great deal of thought. There are usually many other ways to make unprofitable customers profitable without having to resort to such draconian measures.

In a nutshell, CLTV is a useful tool to help you use your marketing resources. Just be sure to use it sparingly and wisely.

Graham Hill
Independent CRM Consultant

Luis Yndigoyen
Member

Posted 09-Feb-2006 07:25 AM
I use the concept of LTV, I think that it is a metric that is not only useful to distribute marketing budget.

The LTV can be used to stablish price strategies, service level and to design retention programs.

The fact that it is hard to calculate the LTV is not mean that it is not important to know this metric. It is not an exactly figure but you can obtain good aproach with some accurate information and a little "out of the box" thinking process.

Max
Member
Picture of Max

Posted 09-Feb-2006 07:38 AM
This is a good fun topic.Bravo! LTV is a must.Period. Would you go sailing without a compass or a GPS or a radar?
LTV tells you who your customers are and where they are in their buying cycle. Simple. LTV is NOT a means to segregate them in a negative way. It is a measurement allowing you to know the level of interest with what they buy, if they can and want to buy more. You can stimulate them to buy, but, it has been seen that the best custumers if «boosted» too much will simply churn without warning. So , for those who use CLTV to make the high CLTV reach new summits of income, be careful, CLTV could be like a balloon that, if you blow too much, will burst in your face. On the same logic, low LTV customers can become average LTV and then high LTV customers, if you know why they joined.Some low LTV will go nowhere. CRM models will warn you and help you know «who's who». The trouble is that most companies are ready to get just ANY customer that they lure with tricky offers and then they expect CRM to turn them into good buyers with high LTV.It does not work that way! Just remember that. «Front end and back end» results are linked directly to offer and conditions of the offer.
CLTV should be the result the sum of ALTV (A for actual)associated to PLTV (P for projected) then you will know when to stop stepping on the pedal. This is simple and require good common sence only. I'll be glad to exchange more. Thank you.

Art Rosenbergb
Member
Picture of Art Rosenbergb

Posted 09-Feb-2006 07:39 AM
Lifetime value can't be the only metric for customer care. The particular service need is a more dynamic metric that must be identified first, then combined, if appropriate, with such value. It all depends upon the business relationship with customers, e.g., ongoing services vs. product sales/support, one-time vs. ongoing relarions, etc.

David King
Member

Posted 09-Feb-2006 08:33 AM
I think that customer value should always be a driver in making decisions on how much to invest in a particular customer. However, I think the concept is often misapplied and that LTV has some particular problems.

One of the problems with differentiated treatment based primarily on value is that, in practice, it often becomes a discount program for valuable customers. The original idea is often that we'll increase our level of service to valuable customers: more frequent and higher quality communications, priority call handling, and so on. However, it all too often devolves into giving discounted promotional offers to valuable customers, which ultimately trains them to be deal hounds, rather than brand loyal.

One poster also mentioned that some companies use value to "fire" customers, which I agree has issues as well. However, I think the far bigger problemis that companies do not pay enough attention to middle-tier customers. Again, a naive program often cannabalizes revenue streams from best customers through discounting, while not investing adequate resources on growing lower-value relationships.

A simple example would be to think about two hypothetical customers. The first spends $1,000 per year with our brand. What will it take to double that revenue, particularly if our stimulus is a 10% discount offer each time? On the other hand, if I have a customer who spends $250 per year, I have a good chance of doubling or tripling my volume, assuming they have the need and the money to buy more.

A particular problem with LTV is that it is often too ambitious. One attempts to build in future revenue streams, event-based costs, probabilities of attrition, and so on. Each factor that is added produces more likelihood of error.

In our marketing practice, we prefer looking forward only about a year or so. Our customer equity models do an excellent job of forecasting future frequency and amount of purchases. Among top-tier customers, we'll observe some who look to increase or maintain their spending, while others are forecast to reduce spending with our brand. This last group is usually the highest priority: these are big dollars that are about to go to the competition. We'll also have forecasts for other value tiers that let us make varying investment decisions.

One useful aspect of value-based decisions is that they can be useful on all sorts of levels for allocating money and resources. (Now that I've complained about the rampant oversimplication elsewhere, I'll present an oversimplified framework for allocating resources.) If we imagine three buckets of marketing dollars, one each for acquisition, growth, and retention, then a value-based decision on allocating these buckets might be:

* Weight retention dollars heavily toward high-value customers. This is where our chief risk is, and we need to retain as much of these revenue streams as possible.
* Weight our growth dollars more toward middle-tier customers, with some investment in some lower-value customers as well.
* Focus our acquisition programs on prospects who can be grown and retained successfully, and make sure that we deliver a great customer experience to each new customer.

In summary, I echo the other posters' sentiments: definitely use LTV, but with care and restraint.

David G. King, CEO
FULCRUM
"Accelerate Your Profits!"

Ed Sander
Member

Posted 10-Feb-2006 12:07 AM
LTV can be used for various purposes but in the past I have found it especially useful in testing scenario's for different strategies and 'selling' them to management. Simply build a LTV table for your current situation and build a second one for the new strategy (e.g. a loyalty program). Play around with the assumptions and you'll find the potential lift of your new program/strategy.

I especially like the way in which LTV tables incorporate all aspects of customer value; retention, annual spend and costs.

Must-read material on LTV is Arthur Hughes Strategic Database Marketing:

http://www.dbmarketing.com/Books/DMS.htm

The winner of the Dutch Marketing Literature Price 2004/2005, Event Driven Marketing, also contains information and cases about LTV:

http://www.eventdrivenmarketing.nl/

Ed

Ed Sander

Business Development Manager
Viking Direct/Office Depot

Database Marketing Specialist
Failsafe Database Marketing

Cathy Allington
Member

Posted 10-Feb-2006 05:37 AM
LTV is only one factor in looking at your most valuable customers. We have a CRM software programme which interfaces with POS systems - we bring across data which retailers have already collected - for up to 10 years to date - and the results have been amazing! It is so different talking the "theory" of CRM, as against the real stuff of working with actual data.

LTV's most valuable factor is, that once you can show a business which has repeat sales, how much a customer is potentially worth to them, they are going to focus more on each customer, and they are going to make an effort to collect customer information. When you can show a business that even though their average cash sale is only $80, but they have customers who have spent over $100,000 with them - that is powerful recognition of the potential value of each customer.
I believe this is the true power of the LTV concept - being able to show a business what each customer could potentially be worth - from their own actual records - and helping them to realise the value of focussing on their customers. This is the key to making CRM work within a business.

However, we then look at customers from the RFM point of view - Recency, Frequency, and Monetary. A client gets excited about finding out who are their best spending customers, but then we show them who are their most frequent customers. This starts to get interesting. One client may have spent $100,000 with them, and has made 350 purchases. However, other clients have made more than 350 purchases, but have spent less on each sale. Their LTV to date is only $50,000. But, they are very happy customers and a potential source of referrals - they may also have the ability to buy more - perhaps the business hasn't promoted their other products - the business previoulsy looked at the transactional value of each sale and thought it wasn't worth it.

Then of course, Recency comes into it. If we looked at LTV alone, without considering the last time a customer bought from us, it would give us the wrong picture. If your product has a 3 month lifecycle, and your highest spending customer hasn't bought for 6 months, then you have problems.

LTV is a good metric as it helps a business to understand what the potential value of each customer is. But used on its own, it gives a biased picture of who the best customers are.

www.gyob.net.au

Graham Hill
Guru
Member

Posted 12-Feb-2006 09:40 AM
As David & Cathy point out, CLTV shouldn't be the only thing you look at when deciding how to allocate resources for customer management.

You need at least three other views in addition to the view of value afforded by CLTV: First and foremost, you need a view of what outcomes customers are looking for from doing business with you. I have recently stopped looking at customer "requirements , needs & wants" in favour or outcomes. Outcomes are crisper than the rest, express what customers are really looking for in their language and how they want to receive them. It is very difficult to grow the value of customers without understanding what outcomes they are looking for. That goes doubly for customers who are already giving you most of their business to you.

Second, you need a view of your CRM capabilities which will be used to deliver the outcomes customers are looking for and thus to grow their CLTV. Capabilities composed of combinations of processes, supporting systems, information flows, skilled staff, organisational routines and other assets & resources lie at the heart of today's options-based approach to CRM Strategy. It is very difficult to grow the value of customers without understanding what capabilities you need to deliver the outcomes they are looking for.

Third, you need a view of how and where in the customer experience the outcomes customers are looking for will be delivered. It is during different touchpoints in the end-to-end customer experience that the capabilities will be used to deliver the customer's outcomes that grow their CLTV. It is very difficult to grow the value of customers without understanding how and when the outcomes they are looking for will be delivered.

Taken together, the CLTV, customer outcomes, capabilities and customer experience views should allow you to develop an appropriate portfolio of CRM options that continuously grow customer value.

The CLTV view by itself is NEVER enough. Indeed, I will go further and say that CLTV or numbers in general should NEVER be the starting point for the development of CRM Strategy.

Graham Hill
Independent CRM Consultant

terence
Member

Posted 13-Feb-2006 12:03 AM
In my organisation, we use the value of the customer and customer tenure as a matrix for service and retention. This may not exactly equate the concept of CLTV which uses NPV to compute a LTV based on an assuemd or forecated tenure. So far, it has worked well although it may be based on more historical data.

I'm also in full agreement that this tool must be used together with other customer info such as customer needs, profile, service costs, etc.

At the end, what is perhaps critical is our ability to relate to the customer at that point in time based on his personal "situation".