Deepam Mishra

Background

In the previous article I had argued that the US corporate sector is best positioned to deliver on breakthrough innovation and that without it, there is little hope of getting ahead of massive and fast moving challenges of climate change, and insatiable human desires. Only innovation can move faster than the rate of decay that human existence is faced with.

Today, we will discuss that for the near future, innovation and AI are nearly synonymous.

What is different with data and AI?

In short perhaps everything. While this is not ‘just another article on data how AI is changing the world’, I want to make some important observations on how 2 very different groups of corporates are leveraging this tectonic shift.

Incremental vs Breakthrough Innovation

Evidence suggests that a majority of businesses see data and AI as a means to enhance their existing products, eliminate costs such as back office or manual effort. Or worse, just window dressing their strategy and product features with incremental machine learning features.

A 2021 MIT Management Review and BCG survey, found that AI strategic benefits accrued to companies that use AI to explore new ways of creating value rather than cutting costs. Those that used AI primarily to create new value were 2.5 times more likely to feel that AI is helping their company.

Is there financial evidence of the transformational potential of AI? The one example that etched in my head is shown in the figure below? What is fundamentally different between Tesla and Ford? Why do market valuations of these two amazing US corporations differ so much, especially when Ford can also make EVs as well as other vehicles too? You’d probably also conclude, like me, that the market thinks that Tesla is somehow able to consistently grow on a growth curve that cannot be emulated by all traditional leaders? In other words, Tesla is an AI-anomaly. Lets call such corporates AI-first Corporations.

So how do data and AI first companies think about managing new product innovation? Can conventional companies bring similar processes in their culture, or are they just doomed to extinction? To discuss this interesting topic, let us break down the question into multiple parts.

Isn’t there More to Innovation Than AI?

Today, the answer is perhaps no. While surely there are many additional technologies driving innovations, AI is different. Think of it this way. What is innovation? My definition of an innovation is a new method of solving an unmet customer need, at a price point that generates a profit. The size of the innovation is a product of how big the need is, how high the competitive barriers are (which largely determines how much margin can you justify) .

Innovation Impact = (Size of Customer Need) x (Height of the Competitor Moat)

It sounds simple, but understanding an unmet + profitable customer need is one of the most difficult innovation tasks. Steve Jobs famously said something like “People don’t know what they want until you show it to them.” In my early startups, one of the biggest mistakes I made was to ask customers open ended “need” questions, and every time I convinced myself that I was hearing what I wanted to believe. The reality was that without showing customers something specific and without asking them to pay for it, such discussions are always useless. Embarrassingly, in hindsight I often found that the persons I had talked to, were not even our target customer!

As I will argue in the next few sections, AI is the sharpest weapon in the innovator’s arsenal for answering this question. For this reason alone, investments in AI are likely to drive more results than any other.

To understand how, let us see how AI-first corporations operate, which is different from others.

AI-Second-Corporation

A close friend of mine, is the CTO at an e-commerce business. During the Covid-19 pandemic and the subsequent supply chain crises, their customer behavior changed dramatically. Their sales started to plummet and inventory costs shot through the roof. They partnered with a leading Cloud vendor who helped the company build a connected data-lake, by bringing in data from CRM and Warehouse Management Systems -something the business had never done before. They then trained AI-models to better forecast product demand (down to the last SKU and each store level) and over time, started to align their inventory plans and pricing accordingly. They found that AI helped increase their forecasting accuracy by over 40% and directly increased their bottom-line by 3% to 4%. This was a transformational result, that helped save the day. My friend who led that project for over 15 months, was hailed as a savior and superstar. In this company, the old analytics tools (Excel models) have been replaced by an AI-driven forecasting system.

What is an AI-First-Corporate?

Contrast that with a highly successful AI startup that started with the simple task of automating data entry into CRM systems. The company was started by a senior Sales leader who knew that sales teams often do not enter data into CRMs on time, and often make a lot of mistakes in data entry. For the leadership team, these delays come in the way of managing the business. Her startup built a solution to convert all hand written notes, voice calls and even chat sessions into CRM-ready text, that could be updated with a single OK button. However, she did not just build that product. She built a system that recorded every piece of data that it could find, from almost everyone who used it. She allowed sales staff to upload raw images of their hand written notes, which often had significantly more information than the deal. She set multiple reminders and requests for additional information to track the deal-flow patterns. She recorded every click and detail that the Leadership team made to review results, and even sought yes/no feedback from them. All such data was routinely organized and stored, by persona, timeline and business context (e.g. month end closing). Over the next 3–4 years, she kept finding pattern after pattern of customer issues that neither she, nor the customers could explain clearly. As a result, she was able to parallelize her innovation cycle and created over 6 new product lines in 4 years, taking the startup to a category defining role and a valuation of over $1B.

So what defines an AI-first company is that assumes that data is critical for customer insights, and builds all products and processes centered around this belief. It does not treat AI as a feature, but as the core of their innovation.

How do AI-First Corporations Create More Value?

AI-Inside(TM) Innovation Cycle

AI-First Corporates create greater value by identifying and capitalizing opportunities faster. And they do so on a consistently.

A simple example is to compare benchmarks and key performance indicators used by data first companies versus conventional. In cloud companies or data first companies the average time to bring a product from concept to launch is measured in days not months. Further, each product team is often launching more than hundred product releases in a single ear. This can only be possible if technology developers can move at the speed of business thinkers.

Can Any Corporation Become AI-First?

Yes and no, in my opinion. The investments and effort required are not that formidable, or unknown. However the problem is harder than that. The good news is that many determined businesses are succeeding and are establishing a pattern that can be followed. But that is the subject of the next article!