A Q&A with Applied AI for Distributors conference speaker & AI expert Brooks Hamilton
Join us for the Applied AI for Distributors conference, June 4-6 in Chicago. Learn more.
Brooks Hamilton, a seasoned veteran in the tech industry, has done ground-breaking work in AI-driven price optimization and sales effectiveness in B2B companies. Today, he is a sought-after speaker and thought leader in the AI community.
We sat down with Hamilton to get his take on the latest developments in AI and how it’s being used in the wholesale distribution industry. He also previewed his session on AI and profitability at our upcoming 2024 Applied AI for Distributors conference.
Distribution Strategy Group: Where are we today with AI tools and where are distributors in terms of the use and application of those tools?
Brooks Hamilton: When I look out at the vendor landscape and the rate of AI adoption within the industry, what I see is unique in my experience. The research and capabilities in the market are quite impressive, and those capabilities are increasing rapidly, quite literally week by week. That pace of innovation is unlike anything I have seen before in my career.
That said, we know it takes time for broad capabilities to translate into business value. First those broad capabilities make their way into products, then businesses evaluate those products, implement them within their processes, and then evaluate the accrued business value. Today, the industry at large is at the stage where it has decided to invest in AI. Businesses may have picked a few initiatives that are a mix of primarily lower-risk projects and then perhaps one higher-risk-higher-reward initiative. They are evaluating vendors and going through those initial implementations.
What they’re finding is really quite fascinating. What they see is a change, not just in the metrics and the benefits, but they’re also seeing a changing way of looking at the processes and their organizational structures themselves. One executive told me: “We rolled out a customer support tool and within 30 seconds of rolling it out, I realized that I was going to need a team that looked really different from the team that I have today.”
DSG: In what way?
Hamilton: Prior to the implementation and during change management planning, they already knew that they had some process redesign and rethinking to do. What they realized upon seeing the process in action was that they will also have to rethink the roles and structure of their organization. Their inside sales team receives many orders via email. Now they have a solution in place that will read the email, identify the key order information and then populate their order entry application.
There is an oversight step in the middle of the process, prior to order entry, where a member of the inside sales team reviews and evaluates the work of the AI before it populates the order entry application. This shifts the work from the mundane of data translation (e.g. Outlook to SAP) to quality review and process oversight. The profile of an individual with these skills is different from the experience level and skillset of an individual who is primarily engaging in order entry.
DSG: Are distributors seizing this opportunity?
Hamilton: Distributors are definitely taking it seriously, but different organizations are in different places in terms of their understanding of it and their call to action. Among boards, CEOs, CIOs and executive ranks, there is a keen interest in pushing ahead. They’ve allocated budget to understand this. They’re going to conferences and running some initial pilots.
They’re trying to get their arms around what they can do from a benefit standpoint, but also, and I think this is especially wise on their parts, they are evaluating the cultural change required to adapt to this speed and pace of change. It is a tall order to visualize the process and organizational structure changes that the business may undergo over the next five years when these general capabilities make their way into core processes. They understand that their current and future competition intends to use this technology to improve their customers’ journey and deliver more to their bottom lines.
DSG: Tell us about your session at the Applied AI for Distributors conference.
Hamilton: There are three areas that I’m going to be sharing for the conference this year. One is an update on where the technology is. The second is a broader perspective on where we see the AI technology being utilized now that over a year has passed since the initial generative AI applications like ChatGPT were released. And then the third is looking around the corner. There are obviously some transformative technologies that have come out already, but there are two more very significant waves that are going to be fairly impactful for distributors that are around the corner.
DSG: What’s around the corner?
Hamilton: What we’ve seen up until today is fascinating text and image generative AI such as ChatGPT and Dall-E 3 from OpenAI. These AI applications can create highly realistic responses and images in response to our text prompts. That said, those AI applications are limited in three core ways. First, those applications have limited access to tools such as web browsers, Microsoft Office applications and programming environments. Second, they lack the capability to manage sets of tasks and subtasks, verify whether the task was completed successfully and envision new tasks to meet the goal.
Finally, the initial models had limited working memory capabilities. While it may be able to recall the key issues discussed during the Continental Congress of 1774, it does not retain key points and preferences from prior exchanges with you. If we put this in terms of tennis, the current AI models can return serves with the best of them, but they have no hope of running a tournament. That is about to change.
We will see two new entirely different waves of AI capabilities by the end of this year and early next year.
The second wave of the AI transition is the explosion of autonomous agent applications. Autonomous agents are applications where we give them an objective and that agent is able to identify what the tasks are, see where the problems might come up, and then utilize the digital tools that they have access to, like email and web searching, to execute on those tasks.
DSG: Can you give us an example?
Hamilton: Let’s say I need to look across the last four statements that I’ve received from a vendor and check to see whether there’s an opportunity for rebates. It will be able to retrieve those statements, look at your rebating policies, read through the contract, check the business that you’ve done with that vendor, assemble that into a set of spreadsheets, apply the logic and then determine whether you are owed and just how much that rebate is going to be. That is an example of what an agent can do independently.
The third wave of innovation that we will see is lower-cost and more effective robotics in commercial and industrial settings. That, of course, will be important for distributors’ warehousing and logistics operations.
In a matter of 24-36 months, distributors will see each of the three waves impact their businesses. Generative text paired with autonomous agents to drive efficiency related to the cognitive tasks tied to the headquarters functions: planning, administration, customer journey and IT functions. The third wave will impact the physical and cognitive tasks of the warehouse: loading/unloading, storage, picking, packing, etc.
DSG: How can distributors think of these tools in terms of greater effectiveness, greater efficiency and lower costs? In other words, greater profitability?
Hamilton: A fairly common scenario that distributors find themselves in is responding to large bids. They often are unable to respond to all inquiries on time, across all requested parts and in an analytically comprehensive manner. To respond to the bid on time, the bid response team may take simplified approaches such as flat price increase and only analyzing high-velocity parts. The team simply does not have the resource bandwidth to pull together all of the required data (e.g. cost, price, customer volumes, alternative parts, etc.), match the customer part IDs to our part IDs, prepare the analysis, modify the bid based upon suggested feedback from the review process and then transfer it all back to the customer’s requested RFQ response template.
The result is frequently that we missed out on revenue we could have won or we left margin on the table by not focusing enough on pricing and product mix. Some business is not touched because there simply was not enough time or resources so the RFQ is left unanswered.
DSG: So, how can AI help?
Hamilton: AI can do a great job in terms of helping with the tactical aspects of assembling and responding to that RFQ. So, we think about the aspects such as I need to match up each of the part IDs that are in the bid with my stock numbers, and then pull in my pricing. That alone is typically quite a bit of work, and that can last days to more often weeks just to go through that assembly process to understand where we are before we’ve made a single commercial decision.
What we can do today is try to shift where we’re spending time on things from the tactical bid response to considering the bid commercial aspects and the strategy that we are trying to align with our go-to-market strategy.
DSG: When it comes to AI, have there been any surprises for you over the past year?
Hamilton: Yes. So, the surprises that I’ve seen over the last year are:
One: The amount of resources being applied by the major tech vendors is just astounding. It reaches at least into the hundreds of billions in terms of internal and external spend that they’re applying toward this effort.
Two: The pace of research and releases has been really, really fast.
Three: I initially assumed that the AI market would be owned by the major model vendors such as OpenAI, Google, Microsoft and whoever else had enough money to actually go train one of these hundred-million-dollar models. Instead, we’ve seen a significant amount of open-source activity by Meta and others. In addition, there is a push to make these models smaller and more efficient so that they can be run on a computer and, hopefully soon, a smartphone. That would address a lot of the privacy concerns that many organizations rightfully have about how and where their data is being used and stored.
DSG: Are there any misconceptions the market still has around AI?
Hamilton: There seems to be still quite a bit of emotional pushback on what these models actually represent. I think the most common refrain I hear when people talk about ChatGPT is they call it a statistical parrot – it just gives you the next most likely word. I think that very much misses what these models actually are, which is they’ve developed a deep understanding of the facts, trends and causation within the environment in which they have been trained.
In the case of ChatGPT, its training environment is text. It’s great at answering text questions, but poor at answering questions about the physical world. Now we are seeing models trained not just in text, but across multiple modes that more closely match to what a human might experience such as of vision, hearing, and text. In that case, I think we can expect models to be able to respond to us that have a much deeper understanding of the world around them in which we operate as opposed to the initial text models.
DSG: Why should distributors attend the Applied AI for Distributors conference in June?
Hamilton: I think this is an excellent opportunity for distributors to learn about how their colleagues are progressing through this AI transition, as well as getting a feel for where the vendor space is. I think as distributors, we lean quite a bit on our software vendors and our technology vendors because we typically don’t have giant margins to fund R&D departments. It’s really a great opportunity to see where their heads are.
And then finally there’s a really great speaker lineup, a set of industry thought leaders and technology insiders who can provide a broader perspective on what they are seeing, both within distribution as well as other industries.
Join us for the Applied AI for Distributors conference, June 4-6 in Chicago. Learn more.
