Client lifetime value

The formula has two inputs and they are not equally powerful. Margin enters linearly: improve it by ten per cent and LTV improves by ten per cent. Retention enters the same way arithmetically and behaves very differently in practice, because the range is so much wider.

An agency can realistically move average margin from thirty-five per cent to forty. It can realistically move average retention from eleven months to twenty-two, and that doubles the value of every client it has.

This is the argument for spending on account management rather than new business, and it is rarely made, for a mundane reason: pipeline is easy to measure and retention is not. New business produces weekly numbers. Retention produces a number you can only calculate about clients who have already left.

The practical consequence is that agencies systematically underinvest in the input with the larger effect, and can point at a dashboard to justify it. Calculating LTV honestly, once, is usually enough to change where the next hire goes.

Using revenue instead of margin is the most common error and it is not a small one, because it makes every account look valuable in proportion to its size rather than its profitability.

A large paid media account with pass-through spend and a thin fee can show an enormous revenue-based LTV and contribute almost nothing. A small, high-margin strategy retainer shows a modest figure and funds the business. Ranked by revenue LTV, the agency prioritises exactly the wrong one.

It also distorts acquisition spending, which is the main thing LTV is for. An agency willing to spend three months of revenue to win a client is spending far more than three months of margin, and the arithmetic that justified it was wrong before anybody committed.

The correction is one substitution and it usually reorders the client list substantially. Agencies that do it for the first time are frequently surprised by which accounts sit at the top and which at the bottom.

LTV on its own is a description. It becomes a decision when set against what it costs to win a client: the time spent pitching, the proposals that lost, the referral fees, the marketing that generated the enquiry.

Most agencies do not know that second number because losing pitches is unbilled time that goes unrecorded. The honest calculation includes them, since the cost of acquiring one client includes every pitch that did not land.

With both numbers, ordinary questions get real answers. Is this pitch worth entering. What can we afford to pay for a referral. Should we take the low-margin account with a long expected life or the high-margin one that will be gone in six months.

Without both, those decisions get made on the size of the fee, which is the one variable that correlates least well with whether the relationship was worth having.

A blended LTV across every client an agency has ever had describes a client that does not exist. Retainer and project clients have different retention curves. A paid media account and a brand retainer have different margins. A client won through referral behaves differently from one won through a competitive pitch.

Segmented by service line, the figure starts to be actionable, and the usual finding is uncomfortable: one offering is subsidising another. The service that wins clients is frequently not the service that makes money, and the agency has been pricing it as though both were true.

That is a pricing decision nobody was making deliberately. Sometimes the answer is that the loss-leader is fine, because it reliably leads somewhere. Sometimes the answer is that it leads nowhere and has been absorbing capacity for two years.

Segmenting by acquisition channel is the other cut worth making. If referred clients stay twice as long as pitched ones, which is common, then the value of a referral is not the fee, it is double the lifetime, and that changes what a referral programme is worth building.