Search Advertising's Effect on In-Store Sales

ConversionKalyanam, K.; McAteer, J.; Marek, J.; Hodges, J.; Lin, L. · 2018Quantitative Marketing and Economics

Topicssearch advertising·cross channel·offline sales·field experiments·bayesian meta-analysis·roas·retail

The search budget arrives with a clean online-conversion case, while store sales sit in a separate reporting system. Before approving the plan, decide whether the business case credits in-store sales affected by search advertising. For broad, non-branded category terms, the evidence says it should. The board case still needs a margin-based profitability test, because the in-store return averaged just below break-even even though several retailers cleared it.

Search ads can create store sales that online attribution misses.

For large multichannel retailers, broad, non-branded category search should be evaluated as a cross-channel growth investment. Category search advertising lifts in-store sales in the advertised categories and total store sales, so the gain is new business for the retailer, not a shift between categories. Counting in-store sales alone, the average return on ad spend fell just short of break-even at a typical 35% gross margin, though several retailers cleared it and online sales were not counted.

Data chart

Search advertising lifts advertised categories and total in-store sales

The advertised-category sales gain persisted at total-store level, showing net new retailer sales rather than a category reshuffle.

Action guide

Require search business cases to include in-store sales.For large multichannel retailers, broad category search created net new store sales that online-only attribution would miss.
Test profitability against your own margin structure.In-store return was positive and close to break-even on average, and cleared it for several retailers, so whether the search investment pays depends on your gross margin rather than on a single industry answer; online and future sales were not measured and would raise the return.
Govern broad and branded search as separate investments.The causal result applies to broad, non-branded category terms in top search positions; branded search was outside the study.
Build enterprise benchmarks from repeated field tests.The effect appeared across many experiments and categories, while category-level differences remained uncertain.
Apply this to large multichannel retailers running broad, non-branded category search.The data are 15 field experiments across 13 U.S. retailers; test separately before extending it to branded search or pure online retailers.

Evidence

In response to category search advertising, in-store sales in the advertised categories rose 1.27% and total store sales rose 1.18%, so the gain was new sales rather than a shift between categories.
Counting in-store sales alone, return on ad spend was 2.50 against a 2.86 break-even at a 35% average gross margin; online and future sales were excluded, so the figure understates the full return.
All 18 broad product categories tested showed a positive estimated sales lift.

Key takeaway

Broad category search can create net new store sales, so evaluate it across channels and test profitability against margin.

Source

Kalyanam, K., McAteer, J., Marek, J., Hodges, J., & Lin, L. (2018). Cross-channel effects of search engine advertising on brick & mortar retail sales. Quantitative Marketing and Economics. Experiments ran on Google.com and one co-author was at Google; the protocol committed in advance that neither Google nor the retailers could drop an experiment from the analysis on the strength of its result, and non-significant experiments were retained.

Evidence strength: Strong (15 randomized market-level field experiments across 13 well-known U.S. multichannel retailers and 76 categories; strongest for similar retailers using broad category terms in top search positions). The study measured in-store sales only; online sales and future sales were not counted.