Personal Selling Keeps Lifting Sales Longer Than Advertising

Marketing AnalyticsKöhler, C.; Mantrala, M.; Albers, S.; Kanuri, V. · 2017Journal of Marketing Research
Topicscarryover effect·marketing communications·advertising·personal selling·duration interval·meta-analysis·budget setting

Most carryover benchmarks come from mass media advertising studies collected before the mid-1980s. Carryover is the part of a communication's total sales effect that keeps working after the spending period ends, and its duration is the number of months until 90% of that total effect has arrived. Imagine your team is splitting next year's budget among television, targeted mail and email, and a field sales force. Which of them earns credit for later sales, and for how long?

On average, most of the sales effect of marketing communication arrives after spending ends, more so for personal selling and targeted advertising than for mass media.

Data chart

Most of the sales effect arrives after the spending ends, most of all for personal selling

Share of a channel's total sales effect that arrives after the spending period, in percent (average of study estimates); the dashed line at 61% is the average for all marketing communication, not a channel.

All marketing communication: 61%Half the effect arrives laterPersonal selling68%Targeted advertising64%Mass media advertising52%
  • Personal selling keeps the largest share for later.

On average most of the sales effect of marketing communication arrives after the spending period, and personal selling and targeted advertising keep more of it for later than mass media advertising.

Averages pooled from 918 estimates in 114 studies, 1962 to 2015; exact values .607 overall, .684 personal selling, .642 targeted advertising, .523 mass media advertising, as shares of the total sales effect. The overall average also includes other kinds of advertising, which have no bar. After accounting for other study differences, personal selling is the only channel shown whose gap over mass media advertising is clear.

Data chart

Sales calls keep working for about a year; advertising for a few months

Median months until 90% of the total sales effect of one period's spending has arrived.

Personal selling12.6 monthsMass media advertising3.4 monthsTargeted advertising2.0 months
  • Plan advertising on shorter cycles than personal selling.

The sales effect of personal selling takes far longer to play out than the effect of mass media or targeted advertising.

Medians from 863 estimates of how long marketing communication effects last, 1962 to 2015; medians are used because a few very long estimates distort the averages. Targeted advertising's median rests on 52 estimates and its average share on 62; the paper does not explain why it ranks high on share but lowest on duration.

Action guide

  1. Credit your communications with the sales that arrive after the spending period.Ask your finance and analytics teams to judge marketing communication budgets against the total sales effect, because on average more of the effect arrives after the spending period than during it. A budget set on the immediate sales response alone understates every channel's sales effect.
  2. Give your personal selling and targeted advertising their own carryover benchmarks.On average both keep a larger share of their sales effect for later than mass media advertising does. In a budget illustration, using an earlier advertising benchmark for both sets their budgets too low.
  3. Plan your advertising on shorter cycles than personal selling.Personal selling's sales effect takes far longer to play out than the effect of mass media or targeted advertising. Mass media advertising's effect also plays out faster than an earlier review found, so schedule it more often than older benchmarks imply to hold the same sales response.
  4. Treat your personal selling's larger carryover as a strength, except on new nonpharmaceutical products.There personal selling keeps a smaller share for later than mass media advertising, so check those launch budgets before crediting the sales force with more carryover.
  5. Ask your analytics team how they converted model output into carryover.Carryover figures depend on the model form and on how its output is converted, so require the conversion that matches the model before your team compares channels or benchmarks.

Evidence

  • On average, about 61% of the total sales effect of marketing communication arrives after the spending period, so the later effect is more than one and a half times the immediate one. After accounting for other study differences, such as product, channel and model type, the share is about 10 points lower in North America than in Europe. The share does not change reliably in recessions.
  • About 68% of personal selling's total sales effect arrives after the spending period, against about 64% for targeted advertising and about 52% for mass media advertising. After accounting for other study differences, personal selling's gap over mass media advertising holds; targeted advertising's is not reliably different from zero.
  • An earlier carryover benchmark of about 51%, from reviews of advertising studies, sits below the new average of about 61% for all marketing communication. In a budget illustration, using it for personal selling and targeted advertising sets their long-term budgets about 36% and 27% lower than budgets built on their own carryover shares would be.
  • The median time until 90% of the total sales effect has arrived is 12.6 months for personal selling, against 3.4 months for mass media advertising and 2 months for targeted advertising. An earlier review put mass media advertising at six to nine months.
  • After accounting for other study differences, personal selling keeps a larger share of its sales effect for later than mass media advertising on established products and new pharmaceutical products. On new nonpharmaceutical products it keeps the smaller share.
  • How analysts convert a model's output changes the carryover figure. In a widely used model, effects multiply rather than add and last period's sales carry forward. Reading that model's raw carryover number directly as the share of the total sales effect that arrives after the spending period gives about 50% on average. The correct conversion gives about 37%.

Key takeaway

Customers keep buying long after a campaign or sales call ends, especially after personal selling.

Limitations

The studies cover mass media, targeted advertising and personal selling, mostly in North America and Europe from 1962 to 2015, and track sales, not profit. The budget comparisons are illustrations built from average figures, not measured returns on spending.

Source

Köhler, C., Mantrala, M., Albers, S., & Kanuri, V. (2017). A Meta-Analysis of Marketing Communication Carryover Effects. Journal of Marketing Research, 54, 990–1008. https://doi.org/10.1509/jmr.13.0580.

https://doi.org/10.1509/jmr.13.0580

Exact values behind the rounded figures: shares of the total sales effect arriving after the spending period of .607 for all marketing communication, .684 for personal selling (.687 in the paper's abstract), .642 for targeted advertising (.650 in the abstract) and .523 for mass media advertising; a share .104 lower in North America than in Europe after accounting for other study differences; the budget illustration substitutes the earlier advertising benchmark of .510, which sets long-term budgets 35.51% lower for personal selling and 26.94% lower for targeted advertising; .502 read raw against .367 correctly converted for the multiplicative model that carries last period's sales forward.