This post is by Darren Woolley, Founder of TrinityP3. With his background as analytical scientist and creative problem solver, Darren brings unique insights and learnings to the marketing process. He is considered a global thought leader on agency remuneration, search and selection and relationship optimisation.
It is fairly common for Agency Relationship consultants to recommend that agency fees, and especially performance fees should be linked to the results of the surveys they run. This is a great way to lock the advertiser into a contract and on the surface provides a tangible value to the survey process. But in our experience and based on insights from behavioural economics this is seriously flawed.
In one recent case we heard about, the agency had more than half their bonus based on the score from the survey and because of issues (with fault on both sides) that occurred a week or so before the survey, the marketers appeared to punish the agency with very low scores and an increased ‘gap’ between agency and client perception. The Client Services Director then spent weeks trying to undermine the results to get a score changed in the face of a 50% reduction in bonus.
We recommend performance based components to agency remuneration, where it is appropriate. We just believe linking the agencies bonus to a score on a survey does an injustice to the agencies and the advertiser and here is why.
For advertising agencies, “a happy client means a happy agency CFO”. Because when the client is happy the agency makes money.
So it got me wondering why so many clients use financial incentives to encourage their agencies to be nice to them, when a bigger financial incentive is already built into a positive relationship?
I hear and have seen account management people turning themselves inside out looking for a few extra points on the latest client relationship survey because it will mean they will get a few thousand more in bonuses.
Behaviour Economics Professor Dan Ariely has shown time and again the distorting impact monetary incentives have on human behaviour.
So is that really the purpose of having a relationship management process? To have the agency fixate on the bonus and how it is calculated?
If the relationship is strong, then the marketers will give more work to the agency and the agency will increase their revenue.
This has a positive impact on everyone in the agency. There is more work for strategy, creative, production and account management, making for a happy agency CFO.
But lets look at the alternative.
Having some sort of regular structured feedback system is a positive step. I think all major advertisers should invest in managing their agency relationships. And that’s all of their agency relationships.
But what about the money?
How much is a good relationship worth?
How big does the bonus need to be? 5% of revenue? 10%? Is it 40% of the bonus? 50%?
And what score does the agency need to get the bonus? Points above the global average? Points improved?
Suddenly it gets so murky and petty. And the agency becomes more interested in the score than improving the relationship.
Suddenly the score and the bonus is more important than the relationship.
But important to whom?
Well that is usually only the senior management team at the agency, because almost no one else working on your business gets a benefit from this bonus.
So who is it really incentivising?
If you want good agency relationships, you need to manage them, measure them and look for ways to maximise them. But it is impossible to simply just pay for them.
In our experience, regular relationship reviews (either quarterly or six monthly) are invaluable. But when these are linked to the agency bonus the focus is diverted away from the objective of improving the relationship and makes the agency focus how to maximise the score now.
We believe a more effective bonus system links agency payments to the delivery of tangible results such as improved media value, increased marketing metrics or in the case of direct response, leads and sales.
What is your experience with performance based models? How have you integrated metrics into this process? And do soft scores like ‘relationship’ work or not? I look forward to the discussion.