BuzbyTimeRead the requirements guide

Time and money · 5.6

Which work is actually profitable

Which work is actually profitable. What decides it, what it costs, and what usually goes wrong. For another implementation reference, see Monitask's reference.

The client who looks fineThe reason this matters

Every professional business has one: substantial revenue, paid on time, a name worth having, and unprofitable once the hours are counted properly.

It looks fine because revenue is visible monthly and cost is not. The hours are spread across people, some are never recorded, and nobody adds them up. For related guidance and industry context, see Gusto.

What has to be in the calculation

  • All hours, including unbilled and never-recorded.
  • Everybody who touched it, including principals and administrators.
  • Loaded cost rates rather than salary.
  • Direct costs: subcontractors, travel, software bought for the job.
  • The cost of servicing the relationship: meetings, reporting, pitching for the next phase.

The last two are where the unprofitable client hides. A client with monthly reviews, extensive reporting and a partner in every meeting can consume a great deal that never reaches a project code.

The overhead allocation problem

How you distribute costs that are not attributable determines the answer, and any method is arbitrary to some degree.

Allocating by hours is the usual choice and it penalises labour-intensive work. Allocating by revenue penalises high-value work. Neither is right; what matters is picking one, documenting it, and not switching when the answer is unwelcome.

Three views worth having

By client. The one everybody wants and the one most likely to produce an uncomfortable meeting.

By service line. Frequently more actionable, because it says what kind of work to sell rather than which relationship to end.

By project shape. Fixed price against time and materials, small against large, new client against existing. The patterns here are usually stronger than the client-level ones.

What to do with an unprofitable client

Not necessarily leave. The options in order of preference: change how the work is delivered, change what is delivered, raise the rate, reduce the servicing, and only then decline the next engagement.

Most unprofitable relationships are fixable at the first or second step, and the analysis is usually presented as though the fifth were the only option.

The strategic exception

Some unprofitable work is bought deliberately: a reference name, an entry into a sector, a capability to develop. That is a legitimate investment and it should be recorded as one, with an expected return and a review date.

What it must not be is a rationalisation applied afterwards to every loss the analysis finds.

How often

Quarterly at client level, annually at every other level. More frequently than that and you are reading noise; less and an unprofitable arrangement runs for a year before anybody notices.

The first calculation to run

Take last year, one client at a time. All hours by all people at loaded cost, plus direct costs, against revenue recognised. Rank them.

Most businesses running this for the first time find that a small number of clients produce most of the profit and at least one large one produces none. Both facts change decisions.

Project profitability against client profitability

A client can be profitable across a year while individual projects lose, or the reverse. Look at both: the project view informs pricing and scoping, the client view informs whether to continue.

The cost of winning the work

Pitching, proposals, and the hours before a contract exists. Substantial in some businesses and rarely attributed to the client eventually won or to the ones lost.

Recording business development against a named prospect costs nothing extra and it makes the cost of acquisition visible for the first time in most professional firms.

Presenting the result

Carefully. A profitability analysis naming a client is read by whoever owns that relationship as a judgement about them.

Present the method first, the whole ranking rather than the worst case, and the options from the section above alongside the finding. Otherwise the discussion becomes about the arithmetic.

Fixed price against time and materials

The comparison most worth running, and the one that most often surprises. Fixed-price work carries the estimating risk and frequently the better margin; time and materials carries less risk and frequently less upside, and the balance differs by business rather than in general.

Only the record answers it for yours.

A short summary

Count all hours and all people. Use loaded cost rates on realistic hours. Include servicing and the cost of winning. Pick an overhead method and keep it. Look by client, service line and project shape. And treat an unprofitable client as a delivery question before a relationship one.

One line to carry

Revenue is visible monthly and cost is not, which is why the unprofitable client looks fine for years.

What this part has argued

That the record only becomes money through a series of judgements: what to bill, how to round, what to concede, what to write off, what to charge, and what any of it cost.

Each is a decision somebody takes, the record informs all of them, and none of them is made by software.

Where to start

Rank last year's clients by profit including all hours. An afternoon, and it changes at least one decision in most businesses.

What not to do with the result

Circulate a ranked list of clients by profitability widely. It reaches somebody it should not, it is read as a judgement about the people who serve those accounts, and the analysis becomes political before it becomes useful.

Share the method openly and the client-level output narrowly.

A last note

Profitability analysis makes visible what was always true. Nothing in the business changed on the day the calculation was run, which is worth saying to whoever receives an unwelcome result about an account they have run for years.

Also in time and money