The Economist (May 16th-22nd 2009) published an article titled Clock-watchers no more, about Coke moving away from the traditional billable hour compensation model for advertisement agencies and instead using a by performance payment model:

“Coke, however, thinks it can do just that [assessing a campaign’s value]. Its new model guarantees to cover advertising agencies’ costs, plus a bonus of up to 30%. The bonus depends on a number of metrics, including the agency’s overall performance, and the sales and market share of the products being advertised. Coke insists that its aim is not to cut costs but to inspire creativity and efficiency.”

Now, I am no expert, and the article is fairly basic and is not delving much into the details of this “by performance” payment model. But to me this sounds like a bad offering, that is, seen from an advertisement agency point of view obviously.

First, a traditional brand-marketing campaign is not like an online contextual advertisement campaign: it is hard to obtain complete viewership information and there is no direct consumer action that can be recorded (there is no URL link to click on in a newspaper or on the TV).

Second, this scheme that Coke (and others are pushing) is simply an attempt to spread out the risk involved in product development, such that the advertisement agencies involved also bear some of this risk. This practice is unfair in my opinion. Would a subcontractor building car parts accept to share the risk in making the car, by getting paid only cost for producing the parts and hoping for a return if car sales are successful? I highly doubt it: large corporations can not expect the double whammy benefit of outsourcing (and saving on costs) and spreading the risk to those they outsource (and thus saving on product development cost too).

Thirdly, there is already a mechanism in the advertisement world that ensures that risk is somewhat spread to advertisement agencies: it is called bidding for a contract. Bidding can involve hours of work that are not billed by a large number of agencies, and the client is given the opportunity to pick the best campaign. I do not see why, once the client has made a decision, the selected advertisement agency has to further pick-up the tab for clients; and all the loosing agencies are left with nothing but a financial loss.

Fourthly, a client builds and sells a product, whereas the advertisement agency simply promotes and pushes this product. If the client fails at making a great product, which is then unsuccessful in its sales, there is absolutely no reason to punish the advertisement agency that created the product’s campaign. Again, I repeat that this is nothing but a spread of the risk and losses away from the client and towards the agency.

Finally, a trial involving lawyers not billing by the hour but instead billing according to the outcome of the case is a completely different situation than a marketing campaign. In a trial, the service (the product) is entirely provided (produced) by the lawyers, it’s not the defendant that argues their cases. In other words, in a trial it is the lawyer who produces the product being sold. In an advertisement campaign, the product is made by the client and not by the marketing agency, so the analogy to layers can not hold in any way.

I could go on and on about why this is a bad idea.

It could also be a good idea, but only if this risk taking by advertisement agencies is also accompanied by a higher rewards. Unfortunately, a 30 % bonus does not qualify as a higher reward.