Agustín may have been a Spanish, retail icon but past successes do not always ensure present successes. What I mean is just because he had renewed a few fashion companies before he was made president of Emília does not mean that he has the secret ingredient to renew Emília. Agustín does not employ design thinking even in the loosest sense of the word. His strategy was to come into the company, call its practices outdated, and try a new approach based off his industry knowledge. This approach did not work. What Agustín failed to do was to be patient and understand his current customer base, the new customers he wanted to bring to the retailer, and the ways of the retailer itself too. As was mentioned later in the blog, he did not perform any user experiments or any kind of prototype. He convinced the board to listen to him without data or user input, and for that he lost the company two hundred million euros. If he had engaged with his market and his resources, he could have easily discovered that his ideas to increase foot traffic and divest from a reduced price strategy would not be met as wholeheartedly as he thought.
If I were hired to consult Agustín, I would definitely advise him to consider the tools I mentioned above; namely, performing user interviews, running prototypes, and not rushing into overhauling the old system without understanding it. More specifically, I would advise Agustín to talk to current customers to see what makes them loyal or what leads them to leave the store without buying anything. Then I would approach competitor stores and do the same. This may only give understanding of a certain geographical location but he could try this in a few other selected areas and make greater learnings. If his initial ideas were validated and needs were found, I would advise him to implement them, but not in the way that he did in the blog. He overhauled all the stores at once. Instead I would suggest trying one store or locale and focus marketing there. This wouldn’t cost much and should already be a success given the previous research. Maybe something unforeseen comes up to disprove his idea, but at least not that much of the retailer’s capital was lost in the process.
Agustín did not once utilize the customer feedback to development iteration loop. He chose a direction, shot his arrow, and hoped he would hit a target. As machine learners may understand, optimizing a model (like in gradient descent) does not mean going straight for the optimal solution. It means making a hypothesis, taking a small step in that direction, re-assessing the hypothesis, making a small step in the new direction, and so on and so forth. Gradually you will arrive near an optimal solution.
