Walk into any large retail store, and there will be some sales on large brands—discounts that are usually funded by the manufacturer. For retailers, these promotions can be a mixed bag: the lower prices may drive people in store or increase sales of that particular item, however the promotion may hurt sales of their competing own-brand products, which offer higher margins. A national retailer in the US decided to conduct experiments to determine how it could preserve their own market share when competing brands were on sale. Six conditions were created —a control and then five experimental discount levels that ranged from zero to 35% for the own-brand items. The retailer divided its store network into six groups, and the treatments were randomly assigned to these. This meant each store had a mixture of the experimental conditions distributed across the different products in the study. For example, in Store 1 own-brand sugar was discounted 20%, and own-brand mascara was full price, whereas in Store 2 mascara was discounted, but sugar was not. This additional randomness meant the retailer was able to allow for variations in sales because the store groups were not identical. The test revealed that matching the large brand promotions with moderate discounts on the own-brand products generated 10% more profits than not promoting the own-brand items at all. As a result, the retailer now automatically discounts their own items when a competing large brand comes on sale. Once they established their proof of concept, the retailer further tested various types of promotions. For example, it discovered that a “Buy One, Get One for 50% Off” promotion on a large brand should be matched on the own-brand product, rather than applying a straight discount. There are two key reasons these experiments were successful: The interventions were easy to implement The results were easy to measure. The products used, stores involved and length of experiments were all limited for ease, and the methodology used for implementing the discounts followed standard operating procedure - store employees often didn't know they were even helping to implement an experiment! This was because the retailer had previously run experiments, and found that if they changed too many things at once, the stores could not handle the implementation, causing long delays and additional cost. At times temporary staff had to be trained to go into the stores, find the products, and change the prices and shelf signage. Then, if the experiment extended beyond a week, problems arose as shelves were constantly rearranged and new signs applied. A maintenance program was required to monitor store compliance - too much work. In some ways the retailer had experimented on experimentation itself— and learned how to design studies that it could analyze more quickly and implement more easily. Sources: Harvard Business Review: A Step-by-Step Guide to Smart Business Experiments