Thanks for your comment.

Youre’ right – the site’s natural conversion rate will impact how sucessfully the PPC agency can deliver sales. A poorly-designed, amateur-looking, over-priced site generally won’t convert as well as a reputable, intuitive, user-friendly site with a strong brand and good value for money.

For that reason, I think any performance model offered to a PPC agency should be relative to the current conversion rate of their site. A decent benchmark using a few month’s worth of data is therefore needed before a performance deal can be created.

Good point about conversion optimisation – I guess with any performance deal (where the agency gets paid based on conversions) it’s in the agency’s interests to work with the client to improve conversion rate through regular multivariate testing.

Very well thought-out post. You obviously know your economics. This shows how hard it is to find a compensation model that accurately aligns the incentives for both company and PPC manager. Further complicating the issue is the fact that the site’s ability to produce conversion could adversely affect PPC efforts. Perhaps conversion optimization would be a better step to take?