Ask most small business owners who their best customer is and the answer comes back quickly, based on the figure that is easy to find: who spends the most. That number is genuinely important and no business should be indifferent to it. But it describes only one side of the relationship. What a customer is worth is what they spend minus what they cost to serve, and the second half of that is rarely measured, because nothing in the accounts records it. The cost of serving a customer is made of time, and time spent is not itemised anywhere: the phone calls, the revisions, the visits, the chasing of payment, the work done as a favour, the meeting that could have been an email.
Once that cost is considered, the ranking often changes, and sometimes it reverses completely. A large account can absorb a disproportionate share of the week - because it is large, because it feels too important to say no to, because it has more people in it who can each ask for things, and because a business that depends on it is reluctant to push back on anything. Meanwhile a set of smaller customers who know what they want, accept the first proposal, and pay on time may quietly be earning more per hour than the account that is talked about as the mainstay. The business feels busy and successful and cannot understand why the money is not better, and the explanation is sitting in the difference between revenue and revenue per hour.
Measuring this properly sounds like a project and does not have to be one. Take the last few months, list the customers who account for most of the revenue, and write against each an honest estimate of the hours the business actually spent on them - including the hours nobody bills for, which is where the whole answer lives. The estimate does not need to be accurate to be useful; it needs only to be honest, because the differences that matter are large ones. Divide the money by the hours and the picture appears immediately, and it is usually uncomfortable in a specific and recognisable way: at least one relationship that everybody thought was central turns out to be earning very little.
What to do with that is a judgement rather than a formula, and it is worth saying that the answer is rarely to abandon the customer. Some accounts are worth a low hourly return because of what they lead to, or what they say about the business, or because they cover the fixed costs and let the rest of the work be chosen freely. The point of measuring is not to fire anybody; it is to stop deciding in the dark. A relationship that is known to be unprofitable can be repriced at renewal, or bounded by being clearer about what is included, or made more efficient - and if none of that is possible, it can at least be a deliberate choice rather than something the business drifted into because nobody had done the division.