It is often said that agencies, if they want to make higher margins, should have more skin in the game, aka more risk.
This increased risk is generally manifested as greater revenue variability, i.e. an element of agency remuneration tied to achievement of the client's marketing objectives and/or the client’s assessment of the agency’s performance. There is nothing wrong with this approach as such, but as often practiced, it may not be ideal.
Firstly, an agency’s finances should be sustainable with just base fee taken into account; variable remuneration should be a success bonus on top, not a pre-requisite to breaking even. The variable element should not come out of base fee - where's the incentive for the agency if the best case scenario is simply to get back to where it would have started?
Additionally, the agency needs to be confident that the measures by which its success will be judged are appropriate and that the bonus will be paid out if targets are met.
The bonus element also needs to be sufficiently motivating for the potential pay-out - a single-digit bonus on a small fee where calculations are arbitrary and / or complex may seem barely worth the effort to achieve.
Another consideration is that revenue variability for the agency also means cost variability for the client, which makes budgeting for both sides more difficult and may explain its decreasing popularity on smaller pieces of business.
A more appropriate approach is to consider value demonstrably created for the client and establish the agency's share of it, rather than a simple top-up bonus metric subject to arbitrary criteria.
The greater issue, however, is that revenue variability is not the agency's only risk - perhaps not even its biggest risk.
As Michael Farmer argues, agency workloads are in general not appropriately linked to fees; the agency agrees a client fee based on staff to be provided, but has not agreed an accompanying detailed, documented scope of work.
In this case, the agency's risk is not just on its revenue line but on its costs, in terms of staff hours needed to complete the work requested by the client.
In this scenario, agencies already have significant skin in the game and are arguably neither charging an appropriate premium for the risk taken nor mitigating their risk sensibly.
What do agencies need to do?
The first step is simply to link workload to fee, by detailing deliverables (be they inputs, outputs or outcomes) and the corresponding fee.
Next, agencies need to price work according to risk - an inputs-only agreement (a set number of hours of certain people's time) is the lowest risk, so should have the lowest profit margin.
Agency guarantees of outputs and / or outcomes need to attract a higher pricing premium. This will involve establishing value created and cost to deliver, then deciding where between these two price should sit - this is what Tim Williams calls the discipline of pricing.
Agencies also need to mitigate their risk by documenting what is included in scope and what isn't - not simply "media plan" or "creative idea", but also the process around delivery; how many rounds of review, deadlines for client approval, how many in-person meetings at the client's offices and so on.
If this feels too bold an approach with clients, remember Blair Enn's findings that Challengers close more sales than Relationship-Builders.
Agencies also need to look at the risk portfolio of their entire client business - how many high-risk, medium risk and low risk clients they have.
They should only add another high-risk client if the business is in a position to take on the risk and manage it.
In addition to managing their overall risk profile, agencies also need to start moving the conversation with clients from an adversarial fight over how the pie is divided up to an alignment in growing the overall size of the pie.
The IPA’s Commercial Conference considered all these issues – the full speeches are available on YouTube and are a masterclass in pricing, selling and negotiation.