It covers the entire survey process, from planning through to recruitment, deployment and optimisation and analysis, making recommendations along the way.

For instance, the guide says that a 15m-impression campaign should be the minimum before a site considers running a study on behalf of advertisers, and that the cost of the research should not exceed 10% of the cost of the overall media buy.

Author Marissa Gluck of Radar Research also writes that media agency staff should be trained and educated on the basics of research methodologies and implementation, and that agencies and publishers should pressure research vendors to undergo independent, third-party evaluations of their statistical methods and recruitment methodologies.

The IAB’s initial report, published last August and based on two-years of research, raised concerns over the reliability and validity of online ad effectiveness studies, citing low response rates, quasi-experimental research designs and a lack of willingness on the part of the industry to fund more appropriate studies.

One particular bugbear is live intercept recruitment. Gluck writes: “With live intercepts falling increasingly into disfavour due to lowered response rates, greater site clutter, and misalignments between campaign delivery and sample, panels are gaining favour with agencies, publishers and vendors.”

Panels are far from a panacea, she says, but if properly validated they can solve for some of the methodological deficiencies of intercept studies – though they are more easily suited to high-volume or multi-platform campaigns and customised ad placements.

The full best practice guide can be downloaded here.