Estimating Lifetime Value

A colleague recently asked me to help determine the “lifetime value” of a customer, which is a bit finicky and tedious but not terribly difficult unless you obsess over minor details.   I’m consolidating some of the correspondence and substituting a different examples, as it seems useful for making a case for the value of solutions to acquire customers, increase share of wallet, or influence consumption habits – which are all germane to customer experience design.

In its most basic sense, lifetime value represents the amount of money a customer will spend on a product during the course of their lifetime.  This is significant, and often overlooked in transactional approaches to marketing because they focus only on a single sale, which in turn leads them to employ tactics that are geared to getting a single sale – and discourage repurchasing.   I could elaborate on this quite a bit, but it would be a diversion from the present topic.

The relationship company does not seek to make a quick buck through a single sale, but instead seeks to establish a lifetime relationship with a customer and capture all sales over the course of their lifetime.  As such its basis for determining the value of the customer should not be the revenue of a single sale, but the revenue of all sales that can be made to the customer over the course of their lifetime.

Example: Spectacles

Let’s take spectacles (glasses) as an example: it’s plausible to assume that a customer will purchase a new pair of spectacles every two years from age 18 to 80.  And yes, this is a “plausible assumption” as will be all figures in these examples – research could derive a more precise number, but for now I’m merely ballparking, so a plausible assumption should suffice.

And so, a pair of spectacles every two years for a period of 62 years (ages 18 to 80) results in the purchase of 31 pairs – which I’ll round down to 30.   If the average price of a pair of spectacles is currently $400, the customer represents $12,000 in lifetime value to the company that succeeds in forming a lifelong relationship with him, such that he returns to them for every pair of spectacles he will ever need.

That is a basis for his lifetime value to the vendor – and it should already become clear that a firm that considers the customer to represent a $12,000 stream of revenue should value him more and be willing to invest more in cultivating a relationship than does a firm that sees this customer’s value as a single $400 transaction and ignores the lifetime value of the customer.

The Basic Elements

I’ll pull the basic elements of the equation together in a general sense, which is a bit tedious but necessary to applying it to various products:

1. Duration of use – How many years will the customer purchase the product?
2. Frequency of use – How often will the customer purchase the product?
3. Unit price – The current price of the product

The equation, then, is to multiply duration of use by frequency of use (which derives the number of units they will purchase over that duration) and multiply that by the unit price to derive the total lifetime value of the revenue expected from a customer if you can manage to capture it.

That’s all you really need to know (or estimate) to calculate lifetime value of a customer for a given product.   A few examples show how it can be applied to various products:

• Automobiles – If the average customer purchases a new car every four years from ages 18 to 80, this means they will buy 15.5 cars in their lifetime.  And if the average price of a car is $30,000 then customers have a lifetime value of $465,000.  The equation is: ((80-18)/4) * 30,000
• Coffee – If the average office worker purchases two cups of coffee from the cafeteria in their workplace each working day (200 per year) from ages 22 to 65 at a price of $1.89, then the lifetime value of their coffee purchases to the cafeteria is $32,508.  The equation is: (65-22) * 2 * 200 * 1.89)
• Diapers – If the average household has two children, for whom they purchase 24 diapers per week for the first two years of their lives, and a 24-count package of diapers costs $12.50, then the lifetime value of that household to a diaper brand is $2,600.  The equation is 2 * 2 * 52 * 12.5

These figures are, again, approximations – and would need to be fine-tuned by research, but the basic calculations should be good enough for a ballpark estimate of customer lifetime value.

Adjustments to Duration

From these examples, it’s already plain that certain products require adjustments to the basic equation.  For example, the office cafeteria captures only 43 years worth of business (because people don’t begin work until age 22 and retire at age 65) and the diaper brand sells to a household for only four years (because the product is only purchased during the first two years of life for each of the two children).

The duration of use is somewhat hazy and is the most influential factor in the outcome of the equation, so it’s worth a bit of nitpicking to get it right because small differences in duration can result in significant differences in the estimated lifetime value.

Primarily, when a firm seeks to acquire new customers, their full lifetime value is less germane than their remaining lifetime value – as any purchased made in the past cannot be won.   So if a premium automobile brand accepts that most people can’t afford its products until age 40, it must start its equation at that age rather than at age 18.

I would be more cautious about adjusting the age at which a customer ceases using a product, as many firms look to current customer behavior to conclude that they only keep a customer for a limited time – if your current customers stay with you for eight years, then switch to another brand, that should not be accepted as inevitable in all cases.  Unless there is a plausible reason that a customer will stop using the product altogether, you should not passively accept that you will eventually disappoint them into switching to another brand, but instead work very hard to prevent that from happening.

Adjustments to Frequency

Frequency of purchasing is another factor in the calculation of lifetime value that is based on the assumption of constant and consistent use.   For a ballpark estimate, it is likely sufficient to look at customer behavior in aggregate – e.g., a person purchases a new car every four years is an estimate based on all car buyers.

In some instances, this can be adjusted by market segment: by definition, the average customer exercises average behavior – but if the customers that a given brand attracts are skewed to purchase more or less frequently than the customers of other brands, then the equation must be adjusted.

There are also overall trends in society that can be observed.  For example, the recent economic downturn has people holding onto their cars for longer, and it is now more typical to purchase a new car every six years rather than every four.   There is some debate over whether behavior will return to normal after the economic crisis has passed, or if six will be the new standard – but at least for the present, the frequency of purchasing should be adjusted.

Share of wallet is another factor that affects frequency, and it is adjustable.  For example, a customer may dine in restaurants 150 times a year, but only visit your restaurant 25 times a year.  A marketing initiative to get them to come to your business more often can have a dramatic effect on their lifetime value to your brand.

There are even initiatives that can increase frequency of use for a product category – convincing customers to change their purchasing behavior in general, such as to dine in restaurants more frequently in total and not just at a specific location.  But both this and share-of-wallet are factors that can be changed, so the difference in lifetime value is only germane to a project if the goal of the project is to increase frequency of consumption.

And finally, frequency may not be perfectly linear.   A customer may purchase more frequently at different times in their life.   Sticking to restaurants as an example, people in their early twenties may not have the financial resources to dine out as often, when they get married and settle down they tend to dine out less often, when their children leave the nest they may dine out more often, and once they are retired and living on a fixed budget they may dine out less often.   If these variances are significant, they should be accounted for.

Adjustments to Unit Price

For the sake of comprehensiveness, I’ll mull over the adjustment to the last factor: price.   This does not seem to be quite as significant, because prices do not tend to fluctuate much from a point of equilibrium with the market – and though poor pricing strategy can be harmful to the revenue of a given brand, it generally does not affect consumption.

While it is true that nominal prices tend to increase gradually over time, this is generally in pace with the decrease in the value of money over time.  Accountants use discounting rates to reflect that a dollar that will be earned in five years is worth less than a dollar that will be earned today.   But if you expect that increases in price and debasement of money are more or less in balance, the difference ought to be negligible.

It is likely far more important to consider not merely the lifetime revenue to be received from a customer, but the lifetime profit that will be derived from this revenue.   For example, it’s all good and well to predict that customers will spend $400,000 to purchase all of a given product they will ever consume, but the suppliers will generate different levels of profit from that amount based on their cost of providing the product to the customer.   One firm might make a net profit of 8% ($32,000) from that revenue whereas another might make only 6% ($24,000).

Ironically, cost accounting is an idea with which most firms are already familiar, and they project their costs and revenues into the future when determining ROI and making efficiency improvements … to cut costs by 2% means generating a specific amount of revenue per year.   For projects whose primary benefit is decreasing costs rather than increasing revenue, this becomes very important.

And That’s About Enough

I’m going to stop grinding on this for now – and possibly for ever – because there is a great deal of elaborate and intricate detail that can be applied to perfecting the calculation of lifetime value of a customer, and it’s really something that accountants should be enlisted to help with if you need your figures to be precise.

But insofar as deriving a figure that’s reasonably accurate for the purpose of considering whether the investment in customer experience is worth the cost, this will likely give CX professionals a good way to derive a ballpark estimate as a quick test as to whether a proposal is worth undertaking.

Prestige Brands

I noticed someone using the word “prestige” in the context of brands.  It was a memory slip, I’m certain, as she was referring to the level of quality between standard and luxury brands, which is more commonly referred to as “premium,” but the use of the word “prestige” was interesting in its implications.

Prestige is usually in reference to a person or an institution, not a product or a brand, but as brands are often spoken of as if they had a personality or were synonymous to a company, it’s not so farfetched to consider a brand to be prestigious.   Prestige implies respect and admiration felt for someone on the basis of our perception of their characteristics or accomplishments.

We feel respect for brands for the very same reasons.   But it seems to me that respect and admiration are independent of their categorization: premium and luxury products have significant prestige, but a standard or economy product may be prestigious when we feel some degree of respect for the accomplishments of firms that produce mass-marketed goods.  Ford, Microsoft, Walmart, and Coca-Cola are all brands that have prestige, but do not fall into the premium or luxury categories.

In practice, prestige confers a kind of domination: we defer to prestigious people and institutions without question because we presume that they have merit, simply because they are prestigious.   A brand’s prestige may have the same effect: once it has become “the leading brand” in its category or gains a significant reputation, consumers no longer consider its merit through any deliberate process, but simply assume that it has prestigious qualities.

But on the other hand, prestige also relies upon distance.   We are more likely to have admiration and respect for legendary people whom we have never met, whether they are remote to us in distance or in time.  A person, after their death, often accrues greater prestige than they ever had in life.  And celebrated persons are often more greatly admired from a distance than by the people with whom they interact regularly.   The more intimately you know someone, the more familiar you are with their flaws, and the less you are likely to idolize them.

That’s not to say that familiarity always breeds contempt, but merely that it dispels the romantic notions we may hold.  It is very often so when the “real” person fails to measure up to the legend that has granted them prestige.  And again, the same may be said of brands: we admire a brand because others seem to admire it – but when we purchase the product we may find that it does not live up to the expectations that were set.

Perhaps a better way to distinguish this is considering the assumed versus earned prestige of a brand.   Before a consumer uses a brand, he assumes it has a level of prestige - this is belief without proof.  After he has used it, the brand may have earned prestige through his experience.   And more to the point, a brand must meet expectations in order to maintain its prestige.   When the brand falls short of expectations, it loses prestige.

There’s also the relationship, in conspicuous consumption, between brand and user: we believe a person who has prestigious brands has earned them, or believe that we can be perceived as better than we are by associating ourselves to prestigious brands.

In all, this has been a meandering post, jotting down some early thoughts and impressions about the notion of prestige as it applies to brand.   I’ll likely write something a bit more focused when I have it better sorted out.

The Design of Everyday Things

Don Norman's Design of Everyday Things has been in my "to read" stack for quite some time, constantly getting bumped back in favor of something more germane to my present interest, but I finally made time to give it attention - and while I've no regrets for putting it off, I don't think the time I spent was a complete waste.

The book is very popular among designers, who seem to pay close attention to everything but the title: The Design of Everyday Things is about designing everyday things.  Light switches, faucets, doorknobs, and other devices that accomplish tasks of very little importance and require very little thought.   And the author provides excellent advice in that context.

The problem is that designers take his advice out of its context - to things that have significant rather than trivial functions, and to things that are not used every day.   It's all good and well to approach the design of a light switch with the idea that it should be easy to use, and that a person should be able to switch a light on or off with little or no conscious effort.

But when it comes to significant tasks that have serious consequences, mindlessness is not an ideal and can be quite dangerous.   You certainly wouldn't want to make turning off the cooling tower at a nuclear power facility to be something that can be done simply and without conscious effort.   Nor should planning a retirement investment portfolio be designed to be done with casual disregard.   Some things are important, can have serious consequences, and merit closer attention.

Norman does acknowledge this in his book, and even suggests methods that can be used to compel a user to slow down and pay attention to what he is doing to avoid unintentional and tragic consequences.  But these parts of the book, like the title, seem to have been ignored by those who advocate that everything should be simple and effortless.

His approach to this is actually quite clever: he first spells out the rules and principles of designing for simplicity, then provides a chapter that tells how each rule can be broken to intentionally add complexity to a device or process in instances in which it is warranted.

And this makes perfect sense: simplicity, like any other principle, is not a panacea but a choice to be applied where it has positive results (and to be avoided when it may have negative ones).   Given the way his work is misinterpreted and misapplied so broadly, perhaps it could use a chapter that explains to readers when to tell the difference.

Fully Loaded Costs

The notion of "cost" is understood to entail more than the money-price of a good or service, but it seems to be only vaguely considered from the perspective of the customer.

When it comes to their own activities, businesses are diligent in determining the fully-loaded cost of any proposed activity.  That is, a business purchasing a machine will include not only the cost of the machine, but the cost of the power to run it, the salaries of the workers who operate and maintain it, and the cost to dispose of it when its useful life has ended.   Cost accountants can be quite clever in identifying all the cash outlays that must be made in order to gain the benefit of a purchase.

I have never seen the same exercise performed with the same punctiliousness for the customer's evaluation of the fully-loaded cost of ownership.   It's likely because the firm feels that it is the customer's problem, and the customer's task to determine how much he ought to pay for something - though for a business that means to sell something to that customer, it's very much their task to predict this with better accuracy.

It's not that non-price costs are entirely ignored, just that they are dealt with very sloppily.   A firm that is seeking a retail location will show great interest in the distance a person will drive to shop a a given store, and will create bulls-eye maps that overlay population grids to determine a profitable location.   But willingness to drive is a rather imprecise reflection of the factors that cause a person to be willing to drive a given distance.  Driving to a store requires money and time - and it's highly likely that these can be calculated with greater precision than they presently are.

I have to concede that time, particularly a consumers time, is far more difficult to quantify.   When it comes to employees, the cost of time is their wage: if you pay them twelve dollars an hour, a task that takes thirty minutes to complete costs six dollars.   But because you do not pay customers for their time, does this mean that their time has no value, that it is worthless or free?  Perhaps to you, but not to them, and if you fail to respect the value of a customer's time, you will certainly fail to obtain or retain their patronage.

In comparing prices in areas that use different locations, economists pay greater attention to the notion of the value of time.   It is not merely a matter of converting the price in foreign currency to a domestic one, though that little trick often gets a great deal of attention from the ignorant:  things seem to be much cheaper overseas if you pay attention to the price and ignore the amount of time it takes to earn the money to purchase it.

Considered in that manner, it might be better to price a product in terms of the time involved - that is, in minutes instead of dollars.  A product that costs six dollars has a time-price of 30 minutes to a worker who earns twelve dollars an hour.   If it requires 10 minutes of driving (each way) and 20 minutes in the store, the product now costs him 70 minutes.   Add in the cost of gasoline and the taxes on the purchase, and it is now a cost of about 80 minutes.   Remonetize that, and 80 minutes at $12 an hour means that the "six dollar" product costs him $15.60 to obtain.   So the same customer would be better off paying nine dollars to purchase the product online and have it delivered to his home than he would to buy it for six at a store ten minutes from his home.

And this pertains only to the cost of acquisition.  The cost of use is another matter, which is the reason that products that are more convenient to use are of greater value to consumers.   A tool that will last five years and reduce the time required to perform a task that is done once a day by one minute represents a savings of 1,826 minutes of about $365 to the person who values their time at twelve dollars an hour.

I have the sense that I have become tedious - but these examples should underscore the point with which I opened: that the cost of a product is more than the money price, and significantly more in many instances.   Were this accounted for with greater precision, producers would understand the true cost of their product to the consumer, and have a more reliable measurement of the value proposition their product presents.

Importance Versus Urgency

Most commercial communication is predicated on the notion that a person who recognizes that their needs are important will purchase a product that is relevant to those needs.   It’s sound logic and entirely plausible, but messages that persuade a customer that a product is important do not have an impact on immediate sales.  This is because importance and urgency are two different things.

Human beings are notorious procrastinators, and in an environment in which they are constantly bombarded with things that demand their attention, procrastination (along with ignorance) is entirely necessary to avoid spending every moment in frenzy to satisfy all the demands on our attention.   So even if you are capable of convincing someone that your product is important, they are likely to disregard your overtures unless they also feel that it is urgent.

Urgency is the perception that something needs to be addressed right away – and it is a perception, in that logic doesn’t enter into the assessment of urgency.  The urgency of eating and drinking, particularly in developed countries, defies logical criteria: people feel the need to take a meal at an appointed time even if they are not hungry, and to have a beverage constantly within reach even if they are not thirsty.   Survival needs have nothing to do with the degree of urgency they feel.

Meanwhile, things that are very important are not regarded as urgent.  It’s particularly evident when it is necessary to act in the present for a need that will not occur until a future time.  Retirement savings are an excellent example: people recognize that they will need to save for retirement, and that the earlier they start in life the less difficult it will be to accumulate sufficient funds.  Yet the vast majority do not even think of it, particularly when it is twenty or more years in the future.

With this in mind, urgency is a more reliable predictor of human action than is importance.   This is likely a reason that sales promotion often ahs to be split off from advertising in marketing departments: advertising convinces its audience that something is important, promotion that it is urgent.

While the tactics that succeed in getting a person to regard a product as being important are straightforward (cause them to recognize their need and the product to be relevant), the tactics of urgency are not very well developed, and are so clumsy that they are ineffective.

Attempting to cause a person to feel fear and panic of an impending problem, or suggesting a problem will arise if they fail to act immediately, have been attempted so often that it has lost all credibility.  A melodramatic sales pitch is automatically regarded as false and is more often ignored simply for its tone.

Attempting to give a person the sense that an opportunity will be lost is also less effective that it once was, particularly in a competitive market.  The artificial “deadline” of a promotional event has less impact to a customer who has multiple options (what is “on sale” at one store this week will be “on sale” at another next week, and is probably available from a third vendor at a regular price that’s lower than either of the promotional ones).

In all, promoters are relegated to waiting for urgency to naturally occur, and left to deal with importance – in hopes that convincing someone that something is important will implant the brand in their mind for a time when they feel a natural sense of urgency to have it.  That tends to be effective, but the effects are not immediate.

It’s also worth noting that advertisers very often focus overmuch on importance, even for trivial products.  It is certain that every chewing gum brand has a staff of marketers who have devoted an inordinate amount of time and energy to mapping out the mental model of various market segments in various gum-buying scenarios to determine how best to associate their brand to the need.  It seems highly unlikely that the gum-buyer puts much deliberation into the decision, and merely chooses among options available at the time of need based on their superficial preferences.

But at that, it seems this line of thought has petered out and is moving in an altogether different direction.