A retail business ran a controlled experiment to test the impact of three different types of promotions - "% discounts," "buy 2 get one free", "$ coupons" - on a number of its products over a six month period. During the third month of the experiment, a major competitor introduced some new, directly competing products in the market. The business found that this intervention negatively affected the % discount promotion, but the other two promotions were not affected. The competitor's action allowed the business to test not only the impact but also the robustness of its promotions. This is an example of:

A. slicing the data after the experiment

B. starting with a proof of concept test

C. a natural experiment

D. measuring everything that matters