AI And Corporate Innovation: Learning Lessons From Digital Transformation

Artificial Intelligence is a moment in the economy like few others, on the magnitude of the shift from steam to electricity. The promised trillions of dollars of new value from efficiency and productivity gains make it an enticing prospect. Digital twins will make production development faster and cheaper with simulations. Marketers can use consumer data to customize offers with unrivalled precision. Gen AI will change human machine interfaces.

Understandably, corporations are concerned that they will be too slow and risk averse to embrace the innovations that such advances may represent. Many are confident of using it to optimize existing products and operations, helping to improve margin performance. However, there is no guarantee that they can capture the prize of using AI to generate revenue growth.

I spoke at an IT conference recently. Delegates were thrilled by the potential of AI and at the same time overwhelmed by the demands business leaders were placing on them. The velocity and urgency of requests coming at them is hard for many to manage. Concerned about this risk, corporations are investing heavily in AI. Managers are determined to overcome the reputation for being slow and risk averse – “this time we are not going to miss out!”

They are right to be concerned. Corporates often miss out when innovation involves a fundamental shift in the market, even when they are ahead in the technology. Think back to Polaroid’s failure to commercialize its early lead in digital photography (it had the world’s first megapixel camera in the market). Or Nokia’s slow reaction to the threat of iPhones, despite having all the core technologies at its disposal.

We often ascribe these failures to risk aversion – and that is certainly an issue – though more recent examples show the reverse tendency. Failure comes not because corporations spend too little on innovation, but rather than they spend too much. I am thinking here of the billions GE spent pursuing its goal of being a “top ten software company,” before the strategy crashed, and the CEO was fired in 2017. We can also look at the over $4 billion Goldman Sachs lost pursuing its goal of setting up a digital only consumer bank. Advertising agency Havas spent only millions on its push to become a “digital agency,” though the scale of the losses still cost the CEO his job and set the company back for years.

What unites these examples is not risk aversion but ill-discipline. In all of these cases, executives committed to investing in innovation without having identified a clear customer problem to solve, nor had they taken the time to validate that their answer was what customers wanted. That these were decisions motivated more by hubris than customer insight is clear from the public justifications made by Goldman Sachs CEO David Solomon. In one television interview he said, “We are a big bank with a big balance sheet, we’ve got a big capability to invest in technology and therefore invest in disruption.”

Bringing this back to our main question of innovation and artificial intelligence, is being the biggest and the boldest the right basis for these investments? The trap of over investing ahead of learning about the commercial opportunity seems to be very real and most of us don’t have billions to waste in the manner of GE and Goldman Sachs.

In the non-AI world before making these investments, we would want to see evidence of a clear customer problem to solve. Develop clear hypotheses about what needs to be true for our solution to be adopted by customers. Be willing to be wrong more often than right in these experiments, so that we know we are genuinely learning. AI is a different sort of shift, its speed, its magnitude, the potential that the technology gives us. Does this make a difference to how we approach it or is there the risk of the same as in “digital transformations” – lots of outlay without tangible benefits.

In the digital business craze firms like GE, Goldman Sachs, and Havas lost their way through ill-discipline not because they were risk averse or slow to act. How can we know that investments in AI are not going down the same path? That’s the question I am researching over the next few months. Please reach out and let me know what you think.