OPINION20 August 2010

Pay-for-performance: coming soon to a market research agency near you?


The pay-for-performance model is gaining traction in the ad agency world. Increasingly, agencies are being asked to share in the success – or failure – of the campaigns they create. Research Rockstar’s Kathryn Korostoff wonders: will clients soon expect the same of their research agencies?

Here are four reasons pay-for-performance will never happen with market research agencies:

  1. Markets change. Research is a snapshot in time. A survey that predicts that feature X will command a price premium today could be out of touch with reality next week. A competitor’s announcement, technology breakthrough or economic climate shift could make the research instantly outdated. We can’t hold the research firm to blame if a new product flounders because the key differentiator is no longer relevant.
  2. The research agency can’t enforce implementation. An agency might do a great project that, for example, identifies an opportunity to dramatically improve repeat purchases. But the agency has no control over if, or how well, the client implements the new strategy.
  3. It’s just too risky. Research agencies are already being squeezed on profit margins – it would be foolish for them to take on more risk.
  4. Some projects just suck. Was that too blunt? Come on, we have all been there. Project delays and budget over-runs can occur that are totally outside the researcher’s control. An agency can’t be expected to suffer because of unforeseen circumstances.

So revenue-sharing as compensation for market research agencies does seem unlikely. But, let’s consider the current market research industry before we reach any firm conclusions:

  • Clientside researchers are battling for budgets and are under pressure to show return on their research investments.
  • New tools are making it easier for companies to do their own research, in-house.
  • Barriers to entry in the market research business are ridiculously low. There are hundreds of “market research firms” out there (even if most of the volume is controlled by the top 20 agencies), all competing for business.

So clients want to minimise risk. When they do work with an outside agency, they need to know it is going to be great. And they have plenty of agencies begging for a shot at their business.

Maybe, just maybe, pay-for-performance could help market research agencies compete in this environment.

But how could it happen? In research, I can only envision models that do not assume a revenue-sharing approach. Given this limitation I see three possible models for pay-for-performance in market research:

Model 1: Base plus. In this model the parties agree that a base fee is payable to the agency for all of its out-of-pocket expenses. This may include software, hosting, subcontractors, sample, phone, mail, and so on. However, the staff-time portion would be subject to a satisfaction scale. If the client is satisfied, the agency gets 100%. Not so satisfied, maybe 50%. Scary? Yes. Crazy? No. If research agencies want to make a statement, this could be it. Of course in this model it is important that satisfaction scores come from at least three different people on the clientside. Preferably the primary project lead and two in-house clients of the final research.

Model 2: Cut of savings. For clients with ongoing research, such as trackers or other regular, annual studies, savings are always important. If a new supplier offers an innovative solution to maintain the research while reducing expenses, there could be a payment based on percentage of savings over a two- or three-year period.

Model 3: Right of refusal. Some studies pertain to non-proprietary topics. For example market segmentation, customer needs and brand perceptions are all studies that a client and its competitors could all be conducting. In this model the agency guarantees its work with a right of refusal. If the client is not satisfied with the final research project, they don’t have to pay. But if they don’t pay, the agency is free to sell the research to anyone else, even a competitor or media outlet.

Obviously, this is all outrageously speculative, but perhaps it is worth further discussion. Pay-for-performance has already come to ad agencies and SEO consultancies. Can research really be that far behind?

Kathryn Korostoff is a 25-year veteran of the market research industry, and currently runs Research Rockstar, an independent, online market research training company. She can be reached at KKorostoff@ResearchRockstar.com.


13 years ago

Interesting perspective. I'd like to hear what others think. We have a foot in the healthcare space, where there has been a lot of talk recently about pay for performance in the provider arena. It's interesting to think that it could ultimately change the way healthcare is given/received in this country, and that the model could trickle over into market research. It's definitely a double-edged sword.

Like Report

13 years ago

Perhaps a model where 'fixed costs' such as fieldwork are guaranteed by the client and the 'variable costs' are remunerated by commercial impact? After all, an insight should always add commercial value for a client so we shouldn't be worried about this. How do we measure ROI again? :-)

Like Report

13 years ago

I LOVE the idea of measuring commercial impact, but how to do it? Sigh....

Like Report

13 years ago

Although tempting, this could be very hard to implement due to the subjective nature of any evaluation criteria. What would be satisfaction measures? Service? Outcome? Often when clients don't like the results, they look for faults in the methodology, sample, etc. Research is the preferred punching bag when things don't go according to plan. What constitutes innovative? Something the client never done or a technique never used in research? and how open to innovation are clients? This would vary from organization to organization. I think this would open the door for abusing vendors by clients trying to get "free" research they never intent to pay for. The right to refusal sounds hostile to me. It feels like making clients buy under duresse. Not good for good client-vendor partnerships. I think any of the 3 will give a lot of work to attorneys and defense lawyes when the suit avalance hit us all.

Like Report

12 years ago

In this earlier debate, no one mentioned the compromise between what the client is prepared to spend, compared with what the agency recommends as the most appropriate methodological solution. If the agency is likely to be called to account, then I imagine the gap between what the client is prepared to pay & what the agency believes is necessary to do a good job will increase. Or, the caveats on the results will be so great as to render the findings fairly useless......In the days of TQM, one way of dealing with this issue was for the internal research dept. to ask the client whether the likely financial outcome resulting from a given research project provided a positive business case for spending the money in the first place. However, this can be very difficult to assess. Perhaps the most effective form of accountability is to set firm but realistic objectives to the research, agreed to by all stakeholders, and then assess whether these were achieved. The rest is then up to the client.

Like Report