99% of AI projects add value (really)Posted November 22, 2018 by
This short blog will callout some important points from a November 2018 survey analysis on AI adoption by McKinsey&Company and add commentary to them.
AI Adoption is Ramping Up
McKinsey&Company report that the number of companies with at least one area of AI usage has grown from 20% to 47% in about a year. This is a significant increase, probably indicating that AI for business is breaking out of early adopters and heading towards mainstream usage. This is also supported by a large increase of mentions of AI and its relative terms like machine learning during earnings calls for public companies.
AI Follows the Money
In each industry, AI is most likely to be found in the most financially important area of the business. The image below shows the concentration of AI projects by industry and function of the company. Retailers are using AI in sales & marketing, manufacturers use it in assembly lines, and so forth. This fully aligns with how we see AI for Trucking at TNX. The core of logistics providers, especially transport brokers, is the matching of a buy and sell of capacity to lock in gross margin. This is the functional area where AI can have the most leverage on a logistics company’s bottom line, and as McKinsey&Company report it is therefore the most likely to be a first user of AI.
Source: McKinsey&Company https://www.mckinsey.com/featured-insights/artificial-intelligence/ai-adoption-advances-but-foundational-barriers-remain
AI Blends Buying & Building
The survey responses show that companies are in an all-out grab for AI staff & resources. They hire their own staff, contract with consultants, buy solutions from large corporate vendors (IBM, etc), and from niche AI start-ups like TNX. The lack of available staff in particular changes the balance of the make-vs-buy arguments. AI is a competitive differentiator. But getting there appears to require an “all of the above” approach to resourcing that allows for systems or at least sub-system components to be purchased even by companies usually focused on custom in-house builds.
AI Never Misses
Perhaps the most important statistic reported by McKinsey&Company is this: only 1% of those who are piloting or using AI had no or negative results. So 99% of AI pilots or use cases are delivering value above their cost. At TNX we see large value being created, but wouldn’t have been able to predict such an overwhelming and consistent result. This is one reason I love 3rd party research like what McKinsey&Company puts out. It extends and validates our own experience.
For those interested in all the detail, take a look at the full article on AI Adoption by McKinsey&Company