First off, is payment by results (PBR) ethical? For example, you don't expect to turn up to the theatre and say to the person in the booth: ‘Yes, I know all the actors, stage hands and writers need to be paid but let me watch the play, see if it works for me and then I'll decide whether to pay you for the ticket.’ You wouldn't do the same for a private medical operation, either, and even if the outcome was very bad, your private medical insurance or estate would still have to pay up. If agencies are professionals then you are paying for their expertise and judgement as much as you are for a doctor or actor.

Ah, but then there's the no-win, no-fee lawyers. Surely, that's payment by results? Indeed it is, but there the result is binary and clear to see.

In the more complex world of marketing communications, subtle (and not so subtle) influences can lead to or prevent a sale, drive a charitable donation or trigger a behaviour change that is completely outside the control of the agency.

Establishing cause and effect

Moreover, the 'line of sight' from input to outcome (cause to effect) is muddied, so how do you get to a formula that everyone will be happy with?

For example, a well-performing campaign encouraging people to prepare for retirement began to perform badly after 9/11, when people started to think there might not be a future to save for. And consider the recent election campaign that got an unexpected but welcome boost from the candidates’ TV debates.

But pay per click (PPC) is the ultimate payment by results, isn't it? Well, no. Even this is flawed because it assumes that the last click was the cause of the transaction and ignores the quality of the click – does it actually convert to a sale, for example?

And what about integrated campaigns, when multiple agencies contribute to one objective? How do you apportion cause and effect and thereby reward appropriately? There’s also a scenario where you haven't followed the agencies’ recommendations to the letter, but have amended – and in many cases improved – the work.

Evaluating results

To make PBR a success you need agreement over the different levels of results, a solid set of evaluation methodologies (a fusion of research, direct response attribution and econometrics) and buy in from all parties to the findings. Then add to that a graduated scale of reward (and possibly disincentives).

Creating such a robust evaluation infrastructure is best practice, but the cost of it could exceed the small percentage of profits most agencies will be prepared to risk as part of PBR – especially as they don't have full control of the campaign. Is it actually worth the effort to set the system up and run it?

Should clients share the risk?

And, returning to my first point, if you've piloted or pre-tested and everyone feels it’s the best work and right thing to do, when you roll out, shouldn't the client be taking the risk alongside the agency?

Given that the client was briefed and bought the professional output, shouldn't they pay the market rate for it whether it works or not? Or do we – and the agencies – really believe we'll get better work when a percentage of the agency’s fee is at stake? Why aren't they giving us their best work right off the bat, when being paid in the usual way?

A working group within the COI is currently investigating the pros and cons of PBR to assess how it might work more consistently in a government context.