A new model for optimal advertising impression allocation across consumer segments

With conflicting recommendations and marketer practices about advertising impression allocation approaches (ie ‘the media strategy’), from approaches centred on reach (‘go broad’) versus targeting (‘get specific’), the debate rages on: ‘Are marketers targeting too much, not enough, or simply targeting the wrong consumers with their advertising?’ This paper interprets the issue of targeting as an advertising impression allocation question and instead of leading with case study evidence which by its nature is parochial, uses a novel mathematical approach to create an ad impression allocation model based on probability of choice. This contrasts with broad reach strategies and is different from other targeting schemes, eg key demographic, high lifetime value consumers, non-buyers for conquest, proprietary segments of interest. The findings suggest that targeting Movable Middles, ie those with a 20–80 per cent probability of choosing the brand of interest, can lead to 50 per cent improvement in return on ad spending (ROAS) versus broad reach media plans. The results are then supported with two in a large scale market case study. The Movable Middle, a segment of category buyers with a 20–80 per cent probability of choosing a brand, are shown to generate 2–23 times more ROAS than other category buyers who are mostly non-buyers of a brand. This pattern was uncovered mathematically but then subsequently verified empirically. By shifting about 10 per cent of ad impressions to audiences that have high concentrations of Movable Middles, a typical 10 per cent share brand can expect a 50 per cent improvement in campaign ROAS and a 13 per cent improvement in converting non-buyers. This leads to better serving brand needs for both quarterly sales and for long-term growth via customer acquisition. This new media strategy is not just limited to digital campaigns; it can be implemented across any media channel, including linear TV, radio and print.