The Future of AI in B2B Distribution

by | Feb 12, 2019 | General BI, Manufacturing & Supply Chain

Reading Time: 4 minutes

The B2B distribution industry was not prepared for the entrance of “Amazon Business” in 2015. The industry has been disrupted by Silicon Valley because it was not technologically equipped and cannot keep up with the heavy automation and artificial intelligence of Amazon.

Tweet: The future of AI in B2B distribution

This difference is nothing new as, “many technologies pioneered in retail have ‘trickled down’ to distribution over time,” but the threat against B2B distributors is more ominous than ever before and companies will have to innovate to stay relevant in our rapidly digitizing age.

The challenge distributors face

A clear indicator of the threat Amazon poses to B2B distribution can be seen in a recent article from MDM’s Ian Heller. In his fictional piece, Heller fantasizes about a super-conglomerate of the largest distributors banding together to compete with Amazon. The radical, and almost defeatist attitude is a strong warning sign of how far ahead Silicon Valley is technologically.

It’s intuitive that many mid-size distributors would not have the technological capabilities of large organizations, but the lack of sophistication is often staggering. Many distributors don’t have an e-commerce site, or they still rely on fax to fulfill orders. By refusing to modernize, distributors have left themselves open to disruption.

How distributors should respond

Speed is the most important factor for distributors in the battle against Amazon Business. Heller and his colleagues at MDM believe many distributors today, “don’t have the scale to match the strength of Amazon but there’s no excuse not to be swift.” These distributors have the advantage of previously laid groundwork through industry relationships and product familiarity so they will likely be able to turn AI into strong insights if they can modernize in time.

Given that distributors can respond rapidly, the technology that is implemented needs to be sufficiently powerful to keep up with Amazon. To be considered a truly “intelligent” solution, the technology must have an automated feedback system able to learn from, and improve upon, performance over time. Furthermore, the system must be able to deliver actionable insights that buoy bottom lines or the experience for the end user.

Hotelbeds Group: What lies ahead

In a talk at ITBBerlin, the world’s largest tourism conference, Sergi Mesquida explained how his company is leveraging artificial intelligence in B2B distribution for hotel goods. Mesquida is the Head of Innovation and New Ventures for Hotelbeds Group. He breaks down the company’s use of AI into three groups: robotics and cognitive automation, cognitive insight, and cognitive engagement. These technologies are similar to the descriptive, predictive, and prescriptive AI classifications discussed in my first post on this topic. However, Mesquida’s B2B approach is an interesting new angle.

Robotics and cognitive automation

The first of these three, robotics and cognitive automation, does not employ machine learning, but instead seeks to increase speed and efficiency of jobs. With the implementation of these tools, his company can better utilize staff, decrease operating costs, and increase operating efficiencies. For Hotelbeds Group, this type of AI is used in automated reconciliation of invoice and automated replacement of loyalty cards for hotels, both of which save labor costs.

Cognitive insight

A second step up is cognitive insight, which leverages data to make predictions. These solutions place events into models that tell whether something is a problem or opportunity but require human intervention after the identification stage. Hotelbeds Group uses this software to detect fraudulent transactions and aid in creating personalized targeted advertising.

Cognitive engagement

The most advanced form of AI for Mesquida is cognitive engagement. Surprisingly, for a B2B distributor, this step is entirely aimed at improving the experience for the end user. Hotelbeds Group uses cognitive engagement for recommending activities to travel agents or predicting demand to improve contracting. These solutions recognize a problem or benefit and carry out the necessary action automatically.

For Mesquida, the process of implementing AI for B2B distribution begins with identifying what information the company wants to gain from a solution. Following this crucial step, the company should spend time scouring the internet looking for use cases in all industries that are similar to their goal. AI for distribution should be aimed at knowing the customer and being better equipped to deliver the right product, at the right price, in the right space. Looking towards the future, Mesquida hopes to achieve a “marketplace of one” through data-driven information to create completely personalized targeting options. But distributors are very far away from that now and should focus on the three styles of AI to chip away at the technology gap between them and Amazon.

Related reading

Teddy Craven

You may also like