Pre-testing animatic television spots that have not yet finished the production process can be as useful as testing the final ad if the correct methodology is employed, a study in the Journal of Advertising Research (JAR) has argued.
Thomas J. Reynolds (University of Texas – Dallas) and Joan M. Phillips (Andreas School of Business, Barry University) discussed this subject in a paper, entitled The Strata model predicting advertising effectiveness: A neural-network approach enhances predictability of consumer decision making.
One of the main learnings from their research – which covered 240 television ads in all – involved a comparison of ready-to-air TV spots with animatic versions that are essentially at the stage of a rough draft.
Reynolds and Phillips explained that the “strategic assessment of animatic advertisements was consistent with those of finished advertisements.
“An important implication of this finding is that advertising can be assessed in rough stages before production expenses are incurred.”
To generate these results, over 5,500 people watched television ads, with an average of 46 viewers per commercial. Some 55% of ads were finished, while 45% were animatic versions.
More specifically, every respondent watched two spots – viewing each ad individually on two occasions, as well as watching both ads simultaneously on a further two occasions.
In eliciting viewer responses, the study used the Strata methodology, which draws on a computer-driven interviewing system for pre-testing ads, coupled with a “means–end” questioning technique.
Using this format, respondents are asked two questions focused on product attributes and functional elements respectively, plus two “upper-level” questions that emphasise “psychosocial consequences” and personal values in turn.
The “product-affect measures” used by Reynolds and Phillips addressed whether the spot left the viewer with a favourable view of the company or product concerned, and if it had increased their likelihood to purchase.
In tracking the “advertisement affect”, the questions covered how entertaining the spot was, and how well it held attention for a participant in the study.
This neural-network approach enables the researchers to determine any linkages between the four questions they asked, which are also known as “nodes”.
And by comparing how entertaining an ad was with purchase intent, the study found the “the neural model is substantially more predictive of advertising effectiveness than is a traditional, entertainment-based copy-testing assessment approach”.
Sourced from Journal of Advertising Research; additional content by WARC staff