Can AI Boost Firms’ Innovation After An IPO?

Going public is a common way to raise new funds, but while it might be assumed that this extra capital would be used to fund new innovation, research suggests that it can often result in the opposite happening. Indeed, one study, from Duke Fuqua, suggests that as many as 7 in 10 firms suffer an innovation slump after going public.

This is arguably because going public forces firms to focus on extracting profit in the here and now rather than investing in the future. A recent study from Wharton examines whether AI might be used to help mitigate this risk.

AI to the rescue

By deploying AI analytics with the funds raised through the IPO, the researchers believe that firms can mitigate some of the “innovation penalties” they typically encounter.

It’s counterintuitive that innovation often declines after an IPO. One would expect that with the infusion of new capital, a firm would feel less financial pressure and have more incentive to pursue risky, innovative projects. Additionally, the enhanced financial position should help firms attract top talent, boost their bargaining power with suppliers, and signal quality to consumers, all of which should theoretically spur innovation.

This is seldom the case, however, as firms focus on short-term goals while also meeting higher disclosure requirements and also working with the dilution in incentive for managers and employees to pursue innovation after their stakes have been reduced and they cash out their initial holdings.

Remaining innovative

The threat posed to innovation by going public has prompted some firms to choose other options. Ikea, for instance, is one company that has famously remained private, while computer firm Dell went back into private ownership in large part to bolster innovation again.

The researchers argue that deploying AI could prove to be a slightly less drastic option for firms that still want to go public to access the additional finance. They believe that AI can be especially useful in the kind of recombinative innovation that is increasingly dominating the innovation landscape (indeed, research suggests that 40% of patents today combine existing technologies in some way).

AI has been proven to be good at this kind of task. For instance, research from Carnegie Mellon University showed that AI can effectively mine the database of patents and research papers and provide valuable suggestions for how these existing ideas and concepts can be combined in novel and innovative ways.

Maintaining the pace

As such, the Wharton researchers believe that successful deployment of AI can ensure that firms don’t suffer an innovation slowdown after an IPO. Their data found that investment in AI was effective in ensuring that firms going public didn’t experience the decline experienced by similar firms that had also gone public, while maintaining the short-term earnings and higher levels of disclosure common with public listings.

The researchers trawled through patent data from nearly 1,500 firms, over 1,000 of which had gone public, with the remaining firms staying private. It then re-examined those firms a few years later to explore whether they had maintained their rate of innovation.

The researchers found that the role AI can play grew over time as the capabilities of machine learning grew. These gains were especially pronounced in manufacturing, which saw a 61% boost in their innovative output after investing in AI after IPO. Investments in AI were also beneficial for companies where the disclosure requirements of a public listing were especially onerous. For instance, firms in energy and pharma industries have a particular regulatory burden in this regard.

As well as investing in technology, the findings remind us that investing in skills is equally as important. The researchers found that firms not only invested in hardware and software, but also in their human capacity.

Of course, investing in AI shouldn’t be seen as a guarantee that the impact of IPOs on innovation can be overcome entirely. Indeed, the researchers found that such investments did little to change the mindset of executives after flotation, with their gaze still very much on short-term goals rather than longer-term investments in innovation. It can at least go some way towards mitigating that change in emphasis, however.

Facebooktwitterredditpinterestlinkedinmail