Last June to August I got an opportunity to teach the course “Strategic management of technology and innovation” again at IIM Bangalore. I have been teaching this elective for the last five years. The class is a mix of part-time MBA (PGPEM) and full-time MBA (PGP) students. Every time I learn a lot through the process of preparation and class interaction. In this article, I present three things that stand out in my reflection and two questions where my gut feeling was significantly different from the class.
Starting a business, Dunzo, and the art of iteration:
During the first half of the course, we explored the question, “How do I build
innovation stamina systematically?” One of the frameworks we used was 2-loops
of innovation – the idea-to-demo loop validated feasibility and desirability assumptions
while the demo-to-cash loop validated scalability and profitability
assumptions. We used the 2-loops lens to look at the starting of various
businesses like Kodak, Dunzo, Ather Energy, Husk Power Systems, etc. For each
iteration, we analyzed four parameters - speed, cost, quality of feedback, and cognitive
biases, especially confirmation bias.
Dunzo turned out to be interesting on multiple fronts. It
was more relatable as compared to Kodak (hardly anyone had used Kodak camera)
and Husk Power (little experience with off-grid villages). For the first six months,
until it grew to a few thousand customers, Dunzo was running on WhatsApp. Later it adopted mobile apps, cloud, and analytics. Dunzo also experimented
with drone delivery. It was a good example of how a business can start low-tech
and iteratively become more high-tech. Founders spoke the language of
hypothesis testing and customer focus. Customer
and delivery partner experience was also improved over the years. And yet,
Dunzo remained in the news for cash crunch and market share erosion throughout the
course duration. When asked, which is easier to fix – a broken customer
experience or a broken business model? Almost the entire class felt a broken business
model was easier to fix. However, we found it easier to find counter-examples
for the latter – Dunzo, WeWork, Micromax, and Kingfisher – all had decent customer
experience but struggled to fix their business model. In case you have any
examples where a broken customer experience couldn’t be fixed, happy to learn
from you.
Titan, smartwatches, and the surfing of technology waves:
In the second half of the course, we shifted focus to enabling and management
of innovation. We restricted ourselves to listed companies and used only
secondary sources for discussion such as annual reports, CXO interviews, and quarterly
earnings call transcripts.
Surfing a technology wave is not easy, too early and you
might create technology debt, too late and you risk losing to the competition. We
explored how various companies responded to technology waves – IBM-Internet, Titan-smartwatches,
Vimeo-video marketing, Amara Raja Batteries-Li-Ion, Lego-sustainability, Amazon-speech recognition, and AMD-data centers. For
example, we asked, “Was Titan late in responding to the smartwatch wave?” This was
interesting because 75% of the class wore smartwatches and none had a Titan. So,
on the face of it, the answer was obvious. However, as we dug into the reports,
we found the answer to be much more nuanced and the game is far from over. We
used the "real-win-worth
it" framework to guess investment, no-investment, and divestment decisions. For
example, we asked, “Was it real, win, or worth it criteria that may have led to Ford
Motors divesting in Level-5 autonomous car startup, Argo AI?”
AI, creativity, and artificial insight: When the
course began in June, ChatGPT hangover was still lingering. I didn’t change the
nature of assignments or projects. However, referencing norms became stricter. The
fact that AI is going to be a powerful force going forward was given. The
challenge for me was to show how to see through the fog and hype. This is where
guest lectures helped. Ravi
Aranke showed how to use ecosystem tracking – users (individual and paid),
startup investments, enterprise adoption, regulatory bodies, expert conversions
(experts shifting their opinions), and professionals (marketing, lawyers, doctors, CAs,
recruiters) to create one’s view. Sunil
Mishra showed how one can build a local chatbot using Python’s langchain
library and highlighted generative AI’s banking uses-cases.
We looked at the surprising move 37 during AlphaGo vs Lee
See dol 2016 Go game and asked, “Was move 37 creative?” It was one of the
moves which the professional Go players thought as a mistake at the first sight
and then realized it was part of an intentional strategy. Almost the entire
class felt that the move 37 was not creative. Personally, I felt the move was
creative, but it also created an opportunity to learn about what creativity
means to different people. I also tried to give a glimpse of how Karl Friston shows curiosity
and insight can be simulated by synthetic agents using active inference
framework. Of course, active inference is primarily used to explain natural
intelligence but he and other researchers have also put forward a proposal for how it
may lead to distributed super-intelligence.
