Can online search ads impact offline store sales? The connection seems counterintuitive, but the answer is a firm "yes." A milestone meta-analysis of experiments with 13 top U.S. retailers, conducted by Applied Predictive Technologies, proves that search ads drive incremental offline sales. Results vary by industry and advertiser, but the research offers guidance for multi-channel retailers: Create a seamless connection between your channels and use digital to efficiently drive sales across the board.
Retailers have long relied on search advertising to increase online sales. It’s an obvious part of the online shopping process, as a search ad is only a few clicks away from an easy purchase, for both buyer and seller.
What’s not so obvious is the impact online search ads have on offline sales in brick-and-mortar locations. Multi-channel retailers use a variety of traditional and digital channels to connect with consumers, and it can be difficult to see if and how their digital marketing investments translate to in-store activity and sales. Yet, it’s an important question, and the answer could determine how marketing dollars are allocated and measured.
But let’s step back for a moment. The reality is that almost all retailers intuitively know that people shop digitally before going into the store, and shopping trends support that intuition. As many as 88% of consumers are researching items online and then buying in a physical store, according to Accenture.1 And this increasingly digital consumer behavior is having a growing impact on retail. A recent study by Forrester estimates more than 50% of U.S. offline retail sales will be influenced by the web by 2017; and in this same period, those web-influenced in-store sales will rise to $1.8 trillion (from $1.2 trillion last year).2
These trends show a clear link between online researching and offline purchasing, which should be encouraging to digital marketers and multi-channel retailers. But even so, marketers are looking for direct, causal evidence that search ads directly impact offline sales. And now they’ll have it.
Between 2008 and 2011, Applied Predictive Technologies, an independent third-party firm specializing in matched-market testing and research, conducted a series of search advertising and retail studies with 13 major Google advertisers. These participating brands are top, household-name retailers, with more than 1,000 U.S. stores each, on average. They represent a variety of staple categories, including apparel, baby products, electronics, toys, cosmetics, sporting goods, furniture, pet food, and home improvement.
These retailers participated in 15 real-world experiments in the U.S. to measure how their online desktop search spend impacted their in-store sales. In these geo-based experiments, retailers increased their search spend in test markets, while holding constant their regular search spend in comparable control markets, and resulting in-store sales were measured.
I took the findings from those 15 experiments and examined them in a comprehensive meta-analysis, and here’s what I found.
Search ads show clear impact on in-store sales and ROAS
Woody Allen famously said, “80% of success is showing up.” This couldn’t be more true as it applies to search advertising and store sales. Investing in search ads, even at a modest level, increased incremental offline store sales by an average of 1.46% (see fig. 1). Since the compound annual growth rate for the retail industry was 1.55%3 for the period of the study, being able to attribute a 1.46% increase to online search campaigns is significant.
The sales return on search ad spend was significant as well — between two and 14x, according to the meta-analysis (see fig. 2). For example, one participating retailer invested $466K in search ads over a six-week period and saw an in-store sales increase of $5.6 million.
Advertisers not only saw an increase in in-store sales, they also experienced an increase in category level sales, with test categories seeing an average sales lift of 5.8%. Each retailer chose specific categories to test, and heavied up its spending on key category terms (for example, flooring or flat-screen TVs). Sales lift per category ranged from 1.5% to as high as 16.9%.
These are significant findings for digital marketers and multi-channel retailers. How best to take advantage of them?
Play to your strengths
Unlike pure-play competitors, multi-channel retailers can connect with consumers in both the physical and digital worlds. Create a seamless cross-channel experience that puts online search advertising to work driving in-store and online purchases. It’s also important to note that the experiments I looked at were only testing online desktop search. We can expect the impact of search ads on offline sales to likely increase substantially if we account for constantly connected consumers, and the mobile search opportunities they bring with them when they shop.
In closing, we already knew that investing in search ads is a proven way to boost digital purchases. But if the effectiveness of digital ads is only being tracked through online sales, multi-channel retailers are likely missing a piece of their ROAS.
And that’s important to correct. This study demonstrates that the impact of search ads on offline sales should be accounted for in your attribution models for media spend allocation. While results will certainly vary based on industry and specific methodologies used, retailers who understand the full value of search in driving both online and offline sales can make smarter spending decisions, know which channels to invest in, and realize bigger returns on that investment.
Kirthi Kalyanam, Professor of Marketing and Director of the Retail Management Institute at Santa Clara University, conducted a meta-analysis of 15 field experiments performed by Applied Predictive Technologies with Google advertisers to gauge the success of online desktop search ads on in-store sales.
The experiments on which the meta-analysis is based were conducted between August 2008 to November 2011 in the United States, with 13 retail brands participating. The retailers, whose names were hidden, had an average of 1,115 stores across the United States, and nine of the retail brands are in the National Retail Federation’s Top 100 list.
The average number of test markets for each experiment was 29, while the average number of test stores was 157. Search spend was increased in test markets and held constant at normal levels for comparable control markets. Retailers spent an average incremental $287,031 on search ad budget per experiment. Total spend for each experiment exceeded $5 million. The testing period for each experiment lasted an average of four weeks. The average number of categories featured in each test was four, but 54 unique category types were tested overall.
 Accenture, Accenture Seamless Retail Study, April 2013
 Forrester Research, Online Retail Forecast 2012 to 2017, February 2013
 NRF Retail Insight Center, Annual Retail Sales, September 2013