Why is it worth pre-testing digital advertising?

by Nigel Hollis | September 14, 2015

I continue to be fascinated that so many marketers and agencies feel free to create and launch advertising guided only by their own instinct and understanding of the target audience. This is particularly true in the digital domain, where TV is often repurposed for digital without any research into the likely effectiveness. The failure to conduct research not only makes the outcome inherently unpredictable, but it is totally unnecessary.

One of the biggest arguments for not researching a digital campaign is the time taken to conduct the research, particularly when the launch date is looming. That might have been a valid excuse in the past, but not any longer. If you can launch and get feedback on your campaign in a few hours at low cost using LinkNow for Digital, then why not do it? Maybe that is why so many TV ads get repurposed even when the evidence suggests this is a very risky practice.


Another growing argument seems to be: why pre-test when you can optimize based on click-through or other in-market metrics? However, why risk launching a campaign that could leave all the optimization algorithms working with fractions of one percent in terms of response rates? Why not check that key executions at least have the potential to capture reasonable numbers of eyeballs and encourage click-through (assuming that metric is even relevant to your brand)? It is not as if there is no downside to ineffective or damaging advertising.

Maybe it is because people still doubt that pre-testing can anticipate in-market response correctly? True, the validation for digital pre-testing is fairly new; unlike traditional media where the validation is based on many hundreds of ads in different countries and categories. It is tough enough to separate out the impact of TV from an overall campaign and allow for other differences in the marketing mix, never mind doing that for digital. However, we do have evidence that digital pre-testing does anticipate in-market results. Comparing to results from in-market Brand Lift Insights we find that the top 20 percent of ads tested in LinkNow for Digital raise purchase intent by an average of 1.8 percent, while the average for the bottom 20 percent is no change. So, if your campaign is designed to have a call to action – as most are – why not check ahead of time that the ad will likely achieve its objectives? One of the great things about digital is that there is so much that can be done to change and improve an execution before launch.

Given the increasing amount companies are investing in digital advertising of all sorts, surely it is time to treat digital with the respect it deserves and apply the same discipline to digital that has been applied to traditional advertising? What do you think? Please share your thoughts.


Leave a comment
  1. Bjørn Dahl, September 15, 2015

    Nigel, thanks for this and all other posts. They are greatly appreciated.

    I believe we are seeing the same arguments as for more traditional media earlier - and there are always someone who threaten by ad research.

    In light of the arguments that digital is so much more efficient than traditional media, with a direct opportunity to respond, one should think the opportunity to further optimize and improve the effects would be embraced.

  2. Guy Powell, September 14, 2015



    Actually, we've found the true value of an ad is in what we call the 'seen, but unclicked' impressions.  Typically 75% of the value of the purchased digital media is found in the seen and unclicked impressions.  This means that optimizing to the click-thru rates is probably not the best thing.  If these seen, but unclicked impressions drive that much value we are wagging the dog with our click-thru analytics.  

    With this in mind executing an ad pre-test is critical to getting the most value out of the digital ad.

    Thanks for posting


    www.ProRelevant.com for white paper on this topic

    Leave a comment