Distribution technology expert Klaus Werner helps companies use technology to develop a competitive advantage and outperform in crowded markets.
In this discussion, Ian Heller and Jonathan Bein, Ph.D., talk with Werner about how distributors can leverage artificial intelligence (AI) to improve productivity, quality and profitability in customer service, supply chain, sales automation, personalization and ecommerce.
Ian Heller: Klaus, you will deliver a keynote presentation, “AI Applications and Use Cases for Distribution,” at our upcoming Applied AI for Distributors conference. AI is such a profound and sweeping technology; can you give us an overview of what you will cover?
Klaus Werner: First and foremost, one of the things I want to discuss is that AI is not a new and shiny object. AI has been around for a long time. It’s just now getting press. The technology has reached a point that is going to be truly transformative, and every CEO needs to start exploring how they can use it in their business.
An example of technology impacting operations is in farming. When 80% of the farm labor was people, the truck came into existence and displaced farm workers. It eliminated the mundane but elevated the people’s skillsets.
Google and Yahoo are other examples. They essentially displaced the Yellow Pages to make finding information more efficient. AI will be transformative, as well, impacting you and your people. It will impact every single discipline in your business, including customer service, supply chain, human resources, recruiting talent and ecommerce. I’ll discuss these areas and leave the audience key takeaways to use in their business. It won’t be a silver bullet that fixes everything and will require a thoughtful and disciplined approach.
Jonathan Bein: It might be a series of silver bullets because it’s not addressing just one thing. You mentioned it spans the entire gamut, so where might people start with AI?
Werner: I’ll provide a top 10 list, but let me use an analogy. When you think of cybersecurity, everyone assumes it is the responsibility of the chief information officer, the CIO. It’s not. It is everyone’s problem because the majority of breaches occur through email. It could be in merchandising, marketing or elsewhere in the supply chain. AI is everyone’s responsibility.
If you are the VP of recruiting, how you use AI will change how you identify candidates and move them to the top. If you lead customer service, how you use AI will change how you manage mundane tasks and leave the more complex problems to the experienced representatives. If you are in sales, how you use AI will change how you automate sales processes for increased productivity. If you are in ecommerce, how you use AI will change how you provide customers with a personalized experience that drives revenue and cross-selling opportunities. It’s not the CIO’s problem or the CEO’s problem alone. Everyone can use AI to make the business better.
Bein: There was another hype cycle of AI about 35 years ago. This is not new when we look at the Gartner Hype Cycle peak of inflated expectations, trough of disillusionment and plateau of productivity. Even in the 50s, the founders of AI were doing things with computers that ran at the speed of an abacus. But they were already thinking about these ideas. In that first hype cycle of AI, we dreamed about having enough computing power. That power is here now.
Werner: And it doesn’t have to be incredibly high-tech to be useful. A significant element of AI is advanced analytics which could be as simple as going out there and scoring your entire customer base using usage patterns to predict what each customer will buy in the next quarter. Then, you can set your marketing and CRM systems to rally around that. Using that data, I anticipate you could get anywhere between 7%-15% incremental revenue. You can serve that data to your sales reps so they know where each customer is in their lifecycle and treat them accordingly. You can crawl, walk or run with AI and take baby steps.
You can start getting quick wins on AI simply without spending a million dollars.
Bein: When you say start simply, there is a build vs. buy issue. The companies you worked for most recently are big enough that they could build these solutions and develop their own analytics AI capability. What are your thoughts on the buy vs. build issue for different-sized companies? Where do you start?
Werner: That’s an excellent question. AI is such an emerging technology, no one is an expert. But I look at it this way: It is no different than any build vs. buy analysis you have done your entire career. No matter the company’s size, you should do a business case to determine what you want to accomplish.
How much would it cost if you pick ACME Company to do this? How much would it cost if you build it? Your CIO would do an estimate and then make a deliberate decision on the best solution. The size of the company doesn’t drive it. It’s driven by the size of the opportunity and budgets.
Bein: What capabilities does a company need?
Werner: It all starts with data. What is the quality of your data, and what is the accessibility of your data? Data drives everything. Data is the oxygen that drives businesses and drives business growth. You need your product data, transaction data and customer data to be clean and easily accessible. The cleaner and the easier it is to access the information, the better your AI solution will be.
For instance, I was recently speaking to a retailer about their plans, and they said the quality of their data could have been better. They didn’t have good customer data; they had good transactional data, but they didn’t have details on that customer. That said, their current data would not yield a good output as it pertained to predictive and prescriptive analytics. Data is the first to consider but I am curious about your thoughts, Jonathan. You’ve been in this business for a long time.
Bein: First, everybody is apologetic about their data, whether it’s customer transactions or product data. I’ve yet to talk to a distributor of any size who doesn’t feel something could be better about their data. Second, in some cases, AI can compensate for less-than-perfect data. If you can get good data, that’s going to drive better solutions. But I’ve seen some instances in AI-based cross-sell where all they had were SKU numbers. They didn’t have product descriptions, yet it significantly outperformed compared to before they had the cross-sell solution. One of the things about AI is that it is resilient to imperfections in the data. There will be cases where your data’s is not good enough to develop a model. You may not have enough data, or it’s too messy. There will also be cases where AI will compensate for less-than-perfect data.
Werner: That’s insightful, Jonathan. AI can fill in the blanks. I agree with you. I saw something interesting on Amazon. I was looking at a product, and a product review encapsulated 5,000 reviews into a single review. From those 5,000 reviews, it narrowed down to the top 10 things I needed to know. It saved me a lot of time. If AI can fill in the blanks, maybe we can accelerate the data that may not have been as important 10 years ago to kick off a project. It becomes critical but not mission-critical.
Heller: I want to bring us down a level. You both are experts at this, but if you are someone who hasn’t been exposed to AI, how do you know if your data is good enough? How do you find the tools to fill the gaps you just discussed?
Werner: The question is, what does a CEO do? Do they bring in a consultant? It goes back to business fundamentals. Look at your business, your opportunities and the problems you are trying to solve. AI would be one of the tools in your toolbox. It’s not just saying, let’s do AI. It’s focusing on the problem you want to solve and evaluating if AI can help in that initiative. In some cases, you may need to bring in a consultant. In other cases, your existing vendor may have a solution. Or you can vet three or four vendors to see who brings the best solution to market.
Heller: I want to raise another issue. I just wrote a column about this. There is a need for CEOs to prepare with regard to AI in that it will increase productivity. Certain roles will require fewer people, fewer analysts and fewer customer service. I dislike what I keep reading about on LinkedIn, which is, don’t worry about AI just learn how to use it. Or, your job isn’t threatened by AI, it’s threatened by someone who knows how to use it. That’s happy talk.
Technology costs people jobs and it’s going to displace people. If you are running a distribution company, you are supposed to make the business more profitable, which means you’ll implement productivity-saving technology. You will lay people off. I think people need to be realistic about that and start preparing to do that compassionately and efficiently.
Werner: It will cost jobs. It will take away the mundane jobs. People will be responsible to a certain degree to elevate their skillsets. In some companies, CEOs will invest in their people and help them do that. In other cases, it’s going to eliminate some jobs, but it will also create a bunch of new ones.
