Television Advertising Weight and Sales Response

ConversionLodish, L.; Abraham, M.; Kalmenson, S.; Livelsberger, J.; Lubetkin, B.; Richardson, B.; Stevens, M. · 1995Journal of Marketing Research
Topicstv advertising·advertising weight·split cable·grp·advertising elasticity·copy testing·cpg

An annual media plan asks for more television behind an established brand because the planning model assumes extra exposure will increase sales. Before approving the increase, ask whether heavier delivery has produced incremental sales in-market. The answer is uncomfortable: larger increases did not produce a clearer sales response for established brands. New products and genuine strategy changes behaved differently, so one advertising rule does not fit the portfolio.

Adding more television advertising is not a dependable growth plan for established packaged goods.

For established packaged-goods brands, heavier television delivery behind the current campaign did not reliably increase sales. The evidence supports a different leadership model: treat an increase as an in-market hypothesis, distinguish launches from established brands, and require incremental-sales proof before scaling.

Data chart

New-product sales responded more often to added television advertising

55New products33Established brands

New-product television increases produced positive sales effects more often than established-brand increases.

Action guide

  1. Gate television increases for established packaged goods on in-market sales proof.More television behind an existing campaign was not a dependable growth lever in the packaged-goods tests.
  2. Use different assumptions for launches and established brands.New-product sales responded more often, so portfolio planning should not apply one response rule to both.
  3. Require strategic change in the investment case.Added television performed better when copy strategy changed or aimed to change attitudes, not merely reinforce them.
  4. Make in-market sales tests the measurement standard.Recall and persuasion scores did not predict sales reliably enough to carry television investment decisions.
  5. Apply this to frequently purchased packaged goods advertised on television.The data are 389 split-cable TV experiments on low-priced U.S. packaged goods, 1982-1988, established brands and new launches alike; test before extending it to digital media or other sectors.

Evidence

  • Only 33% of 207 established-brand tests showed a positive sales effect at the study's 80% confidence standard, against the 20% that would clear that bar by chance alone; larger television increases did not improve the odds.
  • Across 89 established-product tests, average sales responsiveness to more television advertising was near zero and only marginally distinguishable from no response; the nonrandom sample may make the estimate optimistic.
  • Television advertising is not shown to be ineffective: pooling all 210 established-brand weight tests, the study rejects the hypothesis that added weight has no effect at all. The finding is that a further increase behind the current campaign is an unreliable way to grow, not that advertising does nothing.
  • New-product tests showed a positive sales effect at the same 80% confidence standard 55% of the time, versus 33% for established brands.
  • Changed copy strategy and attitude-changing copy were associated with larger established-brand sales effects, with attitude-changing copy the strongest positive factor measured; the copy-strategy findings come from comparisons across tests rather than a designed experiment, and the study asks that they be read conservatively.
  • Standard recall and persuasion scores had no strong, reliable relationship with sales response and provided no clean success cutoff.

Key takeaway

For established packaged-goods brands, more television exposure is a hypothesis to test, not a dependable sales-growth rule.

Source

Lodish, L., Abraham, M., Kalmenson, S., Livelsberger, J., Lubetkin, B., Richardson, B., & Stevens, M. (1995). How T.V. Advertising Works: A Meta-Analysis of 389 Real World Split Cable T.V. Advertising Experiments. Journal of Marketing Research.

Evidence strength: Strong (pools 389 real-world split-cable experiments for U.S. packaged goods, 1982–1988). Generalizes most confidently to frequently purchased, low-priced packaged goods on television, not digital media, other markets, or long-run brand equity.