Marketing Effectiveness After a Product-Harm Crisis

Product & Brand ManagementVan Heerde, H.; Helsen, K.; Dekimpe, M. · 2007Marketing Science
Topicsproduct-harm crisis·brand crisis·advertising effectiveness·price elasticity·error-correction model·cross-brand effects

Eta, a packaged-food brand, returns to shelves after a safety recall. So does its larger sister brand, Kraft, pulled as a precaution. Their unaffected rival, Sanitarium, stayed on shelves. Imagine your team proposes the old advertising plan to win back sales. How should you divide a recovery budget when sales may respond differently to advertising and price?

In this product-harm crisis, baseline sales and marketing effectiveness fell while links between brands changed.

Data chart

Advertising has a weaker sales effect after the recall

Percent change in total sales over time for a 1% change in advertising spending.

EtaBefore the crisisAfter the crisis0.32%0.06% (n.s.)KraftBefore the crisisAfter the crisis0.34%0.24%
  • Damage also reached the uncontaminated sister brand.

Advertising's total effect on sales fell for both the contaminated brand and its uncontaminated sister brand.

Tables 3 and 4, pp. 238 to 239, around the 1996 Australian peanut-butter recall. Exact responses: Eta 0.32% before, 0.06% after; Kraft 0.34% before, 0.24% after, per 1% more advertising. Eta's later response is not clearly different from zero; both Kraft responses are above zero.

Data chart

Old advertising responses understate the recovery budget

Weekly advertising needed to bring sales back to the precrisis level, in Australian dollars.

EtaUsing the old responseUsing the new response69782373578KraftUsing the old responseUsing the new response2226832444
  • Eta spent A$0 weekly after returning.
  • Kraft spent A$44,177 weekly after returning.

The weaker postcrisis responses imply a higher weekly recovery budget for both brands.

Table 5, p. 242: modeled permanent weekly advertising to restore sales, using the final 4 weeks before and first 4 weeks after the crisis. Exact requirements: Eta A$69,782 old versus A$373,578 new; Kraft A$22,268 old versus A$32,444 new. Eta's new requirement uses an advertising response not clearly different from zero. Actual spend averages cover the first 4 weeks after return; the calculation does not establish profitability.

Action guide

  1. Require new advertising estimates before setting recovery budgets.Have your team update them weekly once your brands return, and revise spending as more data arrive.
  2. Put more recovery spending behind the brand with the stronger new advertising response.A sister brand pulled as a precaution can lose effectiveness too.
  3. Require new price estimates before restoring premiums.Ask your team to measure Kraft's new, far higher price sensitivity first.
  4. Add competitors' responses to your recovery forecast.Your brands may become more vulnerable to rivals, while your own marketing takes less business from them.
  5. Compare pricing with advertising when a rival returns.If your brand stayed on shelves, ask your team whether pricing would now take more sales from rivals than extra advertising would.

Evidence

  • Eta's total sales response to a 1% rise in advertising fell from 0.32% before the crisis to 0.06% afterward, not clearly different from zero. Kraft was not contaminated, yet its response fell nearly 30%, from 0.34% to 0.24%.
  • In the model, Eta needed A$373,578 of weekly advertising, in Australian dollars, to restore precrisis sales using its uncertain new response, against A$69,782 using its old one. Kraft needed A$32,444 using its new response and A$22,268 using its old one. In their first 4 weeks back, Eta spent A$0 and Kraft an average of A$44,177 a week. The model supported backing Kraft over Eta.
  • Kraft's sales became far more sensitive to price after the crisis. Their immediate response to price went from about -0.50, not clearly different from zero, to -3.00.
  • Eta and Kraft lost baseline sales in the model and became more vulnerable to competitors. Their own marketing also took less business from rivals.
  • Sanitarium raised its weekly advertising 36-fold during the recall, yet its sales responded less to advertising than Kraft's did after the crisis. More advertising would have added few sales.
  • After the crisis, Sanitarium could take more sales from rivals through price, but it raised prices. Kraft's estimated response to Sanitarium's price was not clearly different from zero.

Key takeaway

In this recall, old plans overstate advertising's effect and understate Kraft's price sensitivity.

Limitations

The study measures weekly store sales, not profit, margins, or ROI, for three peanut-butter brands around one severe 1996 recall in New South Wales, Australia. These response estimates are specific to this crisis.

Source

Van Heerde, H., Helsen, K., & Dekimpe, M. (2007). The Impact of a Product-Harm Crisis on Marketing Effectiveness. Marketing Science, 26(2), 230 to 245. https://doi.org/10.1287/mksc.1060.0227.

Exact values behind the rounded figures: Kraft's total sales response per 1% more advertising was 0.34% before the crisis and 0.24% afterward.