Investing in industrial technologies: looking beyond the hype

Digital transformation, especially in industry, has become a buzzword over the past 5 years. Technologies such as IoT, blockchain, AI and robotics are meant to transform how businesses operate, driving a race for capital towards them. Gartner’s latest CIO survey suggests that adoption of AI technologies has grown 270% over the past 4 years, with 37% of enterprises deploying the technology. IDC, for its part, predicts worldwide IoT spending to reach USD 745bn in 2019, led by manufacturing.

Given these numbers, ever wonder why startups in these sectors remain small and there are no real industrial technology unicorns? Two years ago, I met with the CEO of an IoT platform startup in the Bay Area. The company, which targeted the consumer and electronic appliances market, had approx. 4 million devices on its platform — a record at the time. Yet, that number had been achieved after an incredibly long time and the investors and CEO had grown impatient. The company ended up being acquired soon after and likely not for the heady multiples seen at the peak of the IoT hype cycle.

Look beyond the top line projections and a different reality emerges. Adoption is not mainstream and most companies do not fully understand the implications of these technologies to their businesses nor how they can benefit from it. Progress is patchier still amongst smaller companies, with adoption of emerging technologies up to 10 times lower.

Investors that are familiar with B2C startups and the fast adoption curve of consumer technologies are successively confused and disappointed with the pace of growth of B2B — and particularly industrial startups. This can lead to frustration and eventually disillusionment, if one does not enter the space with a completely different perspective.

How then does one identify where to invest against this backdrop of slower adoption rates? Having seen hundreds of companies over the last few years, we have learnt to identify the markers for impending growth.

It is important to start with the correct perspective. Rather than looking at specific industries, one must look at specific customer segments and use cases within those industries. Equally, it is important to understand that adoption will not be linear. Indeed, there is a good reason to expect that adoption will accelerate over time and the adoption curve can be improved through sometimes minor course corrections.

The starting point has to be patience. While this is generally a pre-requisite for any VC or PE investing, its need is perhaps more pronounced here given the longer lead time between technology availability and customer validation. As a result, the proverbial valley of death is substantially longer for enterprise startups. Over the past few years, as some of these technologies have matured on the hype cycle, more investors have targeted both the early- and late-stages of VC investing. In Q4 2018, the share of low-value deals below USD 10mn grew to 70%, even as the number of high value deals above USD 100mn also reached a record high, according to GlobalData. Increasing average deal sizes is generally indicative of more mid-growth startups having to fight for a shrinking set of financing opportunities.

A big reason for slow adoption rates is friction in the adoption process. This friction occurs for many reasons, including a poor conversion of the technology into an industrial value proposition, having a capex-heavy business model that requires a higher BU-level approval vs. an opex-model that can be approved by managers lower down an enterprise hierarchy. Removing a small amount of friction can lead to large increases in adoption.

A second reason for slower adoption is lack of faith in technology or risk avoidance. Enterprises, keen to avoid risk in general, look for proof points and momentum on new solutions. For a variety of reasons, the burden of proof is usually very high. This challenge can be overcome through gathering overwhelming proof points — not via pilots but through demonstrating real-world, mainstream deployments within market leading companies. As that happens, enterprises worried about being left behind will inevitably join, while others will be convinced of the benefits from interacting with their peers.

Finally, there is a self-reinforcing cycle that will develop as a technology achieves greater proof points and deployments. While companies initially only adopt technologies for limited rollouts on their highest-value cases, the benefits of these technologies become more obvious as they become pervasive and converge, with greater opportunities to monetize or save money. As that happens, adoption rates will accelerate. Thus, what starts out as condition monitoring of equipment, through deployment of IoT sensors, can lead to optimizing maintenance schedules across a site or enterprise once a large proportion of assets are monitored. Combining IoT with AI, and integrating additional datasets, can subsequently lead to optimizing production schedules on a factory floor. Finally, integrating with robotics and video analytics for quality assurance, one has the building blocks for a “lights-out factory”.

Even when solutions offer obvious financial benefits to companies their adoption can be slow. It is no different for the current status of IoT, AI and AR technologies, whose use in the enterprise is far from pervasive. However, it would be a mistake to base future outlook on the past either in the aggregate or when evaluating specific companies. Instead, it is important to look for the right markers that indicate whether a specific technology or a specific company may be poised correctly and could grow rapidly enabled by the right approach.