Artificial Intelligence
Dec 14, 2017

Crossing the chasm: Reflections from the AI Summit in New York

Twenty-six years is a long time in the tech world. Although Geoffrey Moore wrote Crossing the Chasm back in 1991 (and generations of new technologies have since lived and died), this book is still relevant in 2017. After attending the recent AI Summit in New York, I see many parallels between this seminal book and the state of the AI industry. This blog describes where we are regarding the AI solutions market and what we can expect moving forward.

First, let's describe some of the key concepts of Crossing the Chasm, often referred to as the bible for helping industry practitioners understand how to make disruptive high tech products become runaway successes. At the risk of over-simplifying, disruptive innovations initially take root with early adopters—that is, visionaries and technology enthusiasts—but most disruptive products die in the chasm before they cross over into the mainstream market. That's where the pragmatists reside and—more importantly—the larger market opportunity.

The key to successfully crossing the chasm is for the solution provider to have priorities. It should focus on a narrow target segment, build a “whole product" offering which meets the needs of targeted pragmatist buyers, and ensure that client references, the appropriate marketing strategy, and distribution channels exist. Net/net: much easier said than done.

Why the AI solutions market is now in the chasm

After listening to several company representatives describe their work at the AI Summit, I believe the AI solutions market is now in Geoffrey Moore's “chasm." AI Summit speakers are the visionaries and early adopters, but the huge majority of mainstream businesses have NOT yet implemented AI solutions. Recent research by Genpact and FORTUNE Knowledge Group – Is your business AI-ready? – reveals that while companies have bullish expectations about AI, they see it as still being in an early stage of market evolution. Respondents were, however, interested in exploring what AI leaders are doing to succeed.

The AI Summit provided several key takeaways:

  • AI is a disruptive technology: It offers companies great promise for transformative impact, but successful project implementation is hard. Similar to other early markets, there are no turnkey AI solutions yet. Enterprise decision-makers should NOT believe vendors who tell them AI projects will be easy.
  • It is still an early market: There is much confusion over where to begin/how to best proceed with AI. Many people discussed how to simply explain to executives what AI is and how it can best impact their businesses. AI-related terminology is not yet standardized, and there are no clear market leaders.
  • Early adopters are often enamored with new, cool technology: But this will change as the market evolves. Although it is easy to get caught up in the AI technology hype, wise leaders should focus on business problems, not on the technology. Don't put the cart before the horse, and recognize that AI is a new tool—although it is an exciting one—to help solve business problems.

How AI solutions to cross the chasm can

Although AI solutions are now in the chasm, I have no doubt that some will soon cross the chasm into mainstream enterprise usage. For this to happen – and for the AI solutions market to evolve successfully – we need to see the following:

  • AI providers focusing on narrow AI use cases, working with partners and rapidly developing “whole products" that solve specific client pain points in targeted industries. The higher the pain, the better. The more focused, the better. General-purpose AI solutions, while they may garner media interest, will not cross the enterprise chasm for a long time.
  • AI providers creating whole products through ecosystems. Successful AI providers exploit open source tools and partner with complementary product or technology suppliers to offer whole AI solutions. That said, traditional waterfall development techniques are, indeed, a thing of the past. Leading AI providers innovate like lean start-ups, co-innovate with clients and develop minimum viable products before full industrialization and scale.
  • AI providers generating client references. They do this through successful early implementations to interest pragmatic mainstream users. These clients are NOT interested in whiz-bang AI technologies. Rather, they want to see proven ROI and/or real business outcomes. Pragmatic AI for enterprise solutions will be first to cross the chasm. These are ones that can also achieve regulatory compliance by avoiding such issues the black box problem, where companies cannot see or trace how AI solutions make decisions. These solutions do NOT need huge data sets, which take more work and lengthen project times, to train the AI system.
  • Enterprises recognizing that technology is NOT the main challenge. Changing mindsets, culture and business models is harder. Getting change management and governance right is critical; these items will be part of the whole product delivered by successful AI providers. And winning enterprises will start now, work to develop an experimentation culture, and ensure continual learning.
  • Most important, enterprises remember that people and domain expertise matter greatly. Successful AI solutions that cross the chasm will do so because of the innovative coupling of deep process and domain understanding WITH creative use of AI technology. This will drive greater operational efficiencies, improve end customer experience, and present new revenue opportunities. It is only through people that AI solutions will cross the chasm.

I am bullish that select AI solution providers and enterprises will follow these recommendations, so I expect to see more whole products and pragmatist buyers at 2018 AI Summits. This will indicate that for certain high pain point/high value-added AI solutions, the chasm will be in the rear-view mirror.

About the author

Dan Glessner

Dan Glessner

Vice President, Digital

Follow Dan Glessner on LinkedIn