As part of our profiles of TIM people, let us introduce Sourobh Ghosh of Harvard Business School, which is in the United States. Sourobh, you recently won a TIM award at the 2020 virtual conference. Congratulations! So…
What are your research interests right now?
I study how firms use inexpensive experiments to drive innovation and form strategy. While the phenomenon of firms using experiments isn’t necessarily new, what’s changed is that the cost to firms to design and run business experiments has become relatively cheap. With recent advances in information technology, experimentation is now scalable to test decisions of firm-wide consequence, changing the way firms pursue innovation and form strategy.
I’m primarily interested in how firms manage inexpensive business experiments and the implications that this has for the way firms search for new opportunities. There are many open questions here that I’m working on, such as: how should firms prioritize among multiple, conflicting goals when experimenting? Also, do senior managers help or hurt experimentation?
What do you think is your most exciting contribution to academia?
Business experimentation allows firms to reduce the uncertainty of pursuing new strategic alternatives. As the cost of experimentation continues to fall, we might expect that firms would improve in their ability to evaluate new strategic alternatives—helping them identify paths that yield the highest performance.
In my job market paper, I look at how the cost of testing influences a firm’s ability to design experiments that yield high-performing returns. What I find is the opposite of what practitioners and many scholars might expect—that is, I find that cheap testing may actually harm a firm’s ability to identify high-performing returns. This occurs because as the cost of testing falls, the cognition necessary to design complex changes becomes relatively more expensive for firms. And that’s important because I also show that complex changes in experimentation associate with a greater chance of performance breakthrough and a decreased chance of performance failure. Overall, my findings help demonstrate that cheap experiments are not an effective substitute for cognition in strategy. More broadly, it’s possible that for many firms, experimentation may actually be too cheap to pursue.
At the 2020 Conference you won the best student paper award from TIM. Tell us about the paper and why you think its findings are important.
If many strategic decisions are informed with the help of experiments, then what’s the job of senior managers? With Stefan Thomke (Harvard Business School) and Hazjier Pourkhalkhali (Optimizely), we use proprietary data of live business experiments from the widely-used A/B testing platform, Optimizely, to estimate how increasing management seniority associates with learning and performance outcomes in experimentation across industries. Our findings suggest that senior management’s involvement in experiments is not as simple as generic warnings about executive biases or the prescribed benefits of having more executive support for testing. On the one hand, senior managers associate with bolder changes in experiments that create more significant learning signals. On the other hand, senior managers simultaneously associate with smaller optimization and performance improvements.
Understanding the influence of senior managers in experimentation is important as many firms today continue to scale up data-driven decision-making. While our findings suggest that senior managers associate with lower performing experiments on average, our paper also describes different experimental learning modes in the formation of strategy. Alternating between these learning modes may help decision-makers manage potentially competing outcomes of learning and performance in experimentation.
Tell us something personal about yourself.
These past few months, I’ve been on a quest to make a great cup of coffee at home. It all began innocently enough, with the pre-ground coffee you can get from supermarkets. I’ve since graduated to grinding my own beans and experimenting (no pun intended) with various methods using an AeroPress. While I’ve been coached on proper technique by the hipsters on YouTube, I’m still not happy with the cup I’m able to produce myself. If readers have suggestions for how to improve (or if you can tell me what I’m doing wrong), I’d love to hear from you.
If you (the reader) would like to be profiled for a TIM-troduction, or would like to nominate someone else, please contact us at: email@example.com.