Digital leaders across distribution, wholesale, and manufacturing are turning to AI-powered solutions to innovate, provide better customer experiences, and drive efficiency across their organizations.
Xngage Managing Partners Joe Albrecht and Charbel Abou Chaaya recently sat down with Pacific Plumbing Supply’s Digital Commerce Manager Todd Sisson to discuss how his company is using Xngage’s AI-powered xAgent solution to provide a high-quality customer experience and improve its internal efficiency.
xAgents provide unmatched product expertise for your online customers: answering questions, guiding selection, and closing the knowledge gap that costs you revenue. xAgents are your best product expert because they’re trained on your products, data (including spec sheets, manuals, and other documents), and terminology. Our team integrates the solution into your ecommerce environment and other core systems. The xAgent acts as an always-on product specialist for your customers by answering technical questions, recommending products, and reducing cart abandonment driven by uncertainty.
Charbel Abou Chaaya, Xngage Managing Partner and Chief Digital Officer: Todd, everybody’s been talking about AI, but you’ve really embraced it. Joe (Albrecht, Xngage Managing Partner and CEO) and I talked with you over a year ago about how you can leverage this technology to solve real problems and create efficiencies at Pacific Plumbing Supply. Can you talk to us about the AI strategy at Pacific?
Todd Sisson: As most plumbing suppliers, mechanical contractors, we suffer an issue where we have aging employees with great experience, tons of knowledge, and part of our challenge is how do we replace those, you know? That knowledge comes from years and years and years of investment and training and education. We love the idea of deploying an agent to capture some of that information and knowledge from the existing employee workforce that we’ve got. So, one of our strategies is to work with an agent that has got the intelligence from the documentation of the products that we sell, the installation manuals, the specification sheets, and compile that and provide an intelligent answer to our customers when a question may come up. So, one of our initial thoughts was that it’s a way to boost that level of knowledge that our company has had for many, many years and is currently worried about losing.
Charbel: How is your team – not just the people that are using it, but leadership – supported you in such an initiative?
Todd: Yeah, leadership is very excited. Any way that we can leverage a tool that increases our efficiencies, that brings an answer quicker to a customer – whether it's a customer-facing agent on our website or through a tool that we use on our ERP system – it’s a huge benefit to the sales processes and the sales cycle. So, we look forward to any gain of efficiency we can in any way.
Charbel: Joe, I know you were heading this product, the xAgent, and you worked closely with our AI team to make it production ready. Can you talk to us about what the xAgent is and how it can help a distributor?
Joe: Yeah, so this is the first of many agents we’re working on in our agentic suite. It’s the product expert agent, and it’s an agent that’s really, really good in product discovery. It understands the information, it understands the data, the documentation of all of the products. It sits on top of the product master data. You could call it a findability tool; I call it a product discovery tool. It really helps not just to find a product but to get to the right product for the purpose that you have as maybe a contractor in the field or as a sales rep who has to answer a specific question from a customer. It can be extremely smart about what the particular product attributes might be and tell you exactly what you might need to fit into your given application or the job that you’re doing. For instance, an HVAC unit that needs a new replacement part – you can ask it very detailed questions about the specifications. It further leverages, of course, AI and LLM but many other components and is fully integrated with the website. On the customer side, you could also call it a product recommender because it’s giving the customer one or three or four options to look at and then directly guides them to the product page and then through the normal website experience.
Charbel: Joe, I really like what you said about this product acting as a product recommender. Can this technology have different use cases than just a product expert or finding a product? How can it be used in other ways?
Joe: At the moment, it’s confined to providing product spec recommendations. It can help you compare products. It can basically do anything you want it to do. If you ask it to make you a table with all of the key attributes – you can even say which ones you’re looking for – then it will make you a table and give you the values of all of these attributes. If you say, “Here’s a picture. Do you have something like that?”, it will produce a result and find products that are the same or similar to what’s in that picture. It can do some other fancy things. You can upload an entire document with manufacturer part numbers; it can determine those and give you an entire set of product parts that are internal to your operations. Didn’t you deploy that from a sales rep perspective, Todd?
Todd: We did. It’s an interesting piece. We’ve got a couple of gaps that we’re working through, but it is a fantastic tool for putting a list together, giving us the URL for the products, and putting it into the cart. It’s powerful.
Joe: And actually, there’s this aspect, you call it the spec book, or something?
Todd: It does it to a degree right now, but we’re working on a tool where you drop in a set of plans, it builds a specification book for all of the specs for every one of the products with all of the submittals, if we want a submittal book, or a manual to build different types of output for the customers and the engineering team or the architect, or whoever is involved with the building that would like it. But it takes a job that would take days and turns it into minutes.
Joe: And we actually have that prototyped, but we haven’t launched that yet.
Todd: Fantastic. That’s a big ask for us, so I’m excited to hear that.
Charbel: So, you talked about some efficiencies using the product expert whether that’s to find a product or solve a problem. Were you able to produce some estimates on efficiencies? How is this xAgent helping your team internally?
Todd: Currently, right now, I don’t know the exact numbers with efficiency, but it’s been used probably 500 to 1,000 times a week by our internal team. For example, if I’ve got a question from a customer about an installation for a particular piece of equipment, and we would like to know the clearances from the walls that this model needs to be installed on, in seconds versus minutes – maybe even longer – you get an answer. And what’s more important about that is that you’re getting an answer so quickly, you’re not getting an answer from a book or researching online or anything like that. It’s coming instantaneously to you, and it allows that flow of information to the customer so much quicker. So, you don’t lose those moments, and they go, “OK, I’ll take that piece of equipment, and we’ll buy it from you.” It’s a quicker, more efficient process to close the sale.
Charbel: We were discussing a lot of different topics during the Xngage Client Summit, and one of the things I heard someone say was that if you’re able to get back to your customers quickly, whether on product information or a certain quote, that you’re going to win that customer. If you don’t get back to them within an hour, you’re most likely going to lose them because they are focused on getting the job done. They need to figure out if you have the product that’s going to help them finish the job. In your experience, is that true, and how can an agent help?
Todd: It’s a very real situation. We sell products that lots of different companies sell, so the more confident that the customer is that they can get the information quickly from us – whether we have it in stock or a particular specification question, whatever it might be – the quicker we are to respond and don’t leave that customer hanging, the more likely it is that sale is going to go through.
Charbel: Todd, you’ve always been forward thinking and always want to adopt new technologies. And I know when you find something that works, you start thinking about what’s the next problem you want to tackle. When you came to us and said you think that we can leverage this technology to solve this problem. You were the first to adopt this technology, and we’re very grateful for the feedback we’ve heard from you and your team – it’s been amazing. Do you have plans to continue to adopt AI in your operations at Pacific, in different areas, in different departments?
Todd: Oh, absolutely. We adopt a different agent into our commercial quotation business right now. We drop a set of plans; it takes all of the information from those plans and puts it into a quotation. So, it streamlines the quotation process for our department probably ten-fold in terms of speed to market. Fantastic tool. There are all kinds of ways internally that we’d like to leverage the xAgent, specifically, in terms of helping new employees learn the systems that we have, learn the products that we sell, gain information quicker without having to go ask somebody else in another building or another branch.
One of the ways we’d like to see the evolution of the xAgent is to make it more personalized to our customers. So, it’s a tough ask, and we’re working through some of the process. I’d like to make it more than just a “Claudia bot.” We want to make it a personal friend to our users and our internal team – to make sure that they’re having fun with us is one of our things. Our company’s motto is, “We’re the service people.” So, we want to be the point of contact whether it’s product information, job site information, whether it’s installation questions, whatever it might be. We want to provide a tool that really is an employee, not just for us but for our customers as well. Let them drop in a set of plans and build a PVC pipe list or a water line list or a mechanical takeoff. To do all of that stuff quickly and efficiently through an agent is an incredibly powerful tool. We’re really excited to see the future and where you guys are taking this tool, what’s coming next. And we hope to help provide some information and guidance.
Charbel: That’s amazing, Todd. I really enjoy listening to you because you look at AI as a tool, as an extension of your team, how it can help your team, support them to continue to deliver on the culture of Pacific Plumbing Supply. If customers are dealing with Pacific Plumbing Supply, it doesn’t matter what technology it is, it needs to still feel like this is you, this is your company.
Todd: It has to have our flavor and color.
Joe: I know we’ve covered a lot, but what is one thing that is not AI that you’re excited about, looking forward in terms of your digital solutions or your website?
Todd: We’re looking at several initiatives trying to branch out in ways that are a little unique in our space. Part of the distribution atmosphere is that we’re all very old, rigid companies for the most part – “dinosaurs” is what most people call it. So, we’re looking at ways to take our digital tools, maybe expand outside of our normal boundaries of where we play in the sandbox. And it’s an interesting application, and we will be looking to probably expanding that sandbox in the next year and half, two years.
Learn More About Xngage’s xAgents and Other AI Services
In addition to our xAgent product expert solution, Xngage helps manufacturers and distributors define their AI strategies and implement AI tools that have a real impact on their productivity and customer experiences. With our AI consultation services, we can help your organization identify the internal processes and workflows that can be automated with the help of AI solutions. We make recommendations for AI solutions or build custom solutions for your most complex challenges.
We also offer an AI-powered solution that helps you streamline your product data pipeline. This tool removes errors, duplicates, and inconsistencies that cause issues and inefficiencies throughout the rest of your ecosystem. Product data is standardized by aligning units, formats, and attribute structure across every product and source. The data is mapped to consistent taxonomies and relationships that power search, recommendations, and your other AI solutions. And finally, it is enriched with attributes, assets, and descriptions that drive conversion and discoverability. While this process normally requires a wealth of time and resources, our solution can do it all in minutes.