B2B eCommerce Show

Bryan House brewed beer professionally before helping lead Acquia, the open source company behind Drupal, to more than $170 million in revenue. He is now CEO of Elastic Path, a B2B commerce platform built for manufacturers and distributors with complex catalogs and pricing.

Justin King sat down with Bryan to ask who owns the intelligence in AI-driven B2B buying. Bryan explains the difference between renting intelligence by feeding product data to models like ChatGPT and owning it by running your own AI-powered search. They also cover why product data quality now decides competitiveness, how AI can bring each customer’s negotiated pricing into search, and why the cost of a new commerce implementation is about to drop sharply.

Key idea: encode every bit of intelligence your company has, so your people can spend their time on judgment.

Trevor Pope is on the strategy and value team at Optimizely, where he has spent more than twelve years helping manufacturers and distributors get more out of their commerce, CMS, experimentation, and personalization tools.

Jason Hein sat down with Trevor to ask a question every B2B vendor has stopped being able to answer with “we’re doing AI”: what does it take to adopt AI in a way your company can still manage six months from now? Trevor’s answer is the AI harness, the layer that connects your applications, permissions, brand guidelines, and audit logs to whichever LLM you choose, so you can change the model without rebuilding everything around it. Jason’s version is that the LLM is the horse, the digital business is the cart, and the harness is what lets you swap horses.

Trevor explains why Optimizely built its own harness instead of adding AI features to existing products, and how it brought in a second LLM provider without changing anything for users, agents, or context. He walks through how a regional wire distributor used agents to run more experiments, how another distributor caught products missing from categories, and how manufacturers are producing more campaign content with small teams. They also get into how a harness differs from an iPaaS, what CIOs actually ask about (audit trails, authentication, and whether their data trains someone else’s model), and why most companies are still pointing AI at internal work before letting customers talk to an agent.

Key idea: the horse will keep changing. The harness is what you build to keep.

Søren Handlos is co-founder of Shopspray, a software company that lets B2B sellers run punchout directly from their own commerce platform. He started in advertising, moved through internet media and telecom, built a commerce department at a telco, and spent years as a commerce consultant before a large Swedish B2B customer’s request for punchout led to Shopspray.

Jason Hein sat down with Søren to question a feature most B2B teams think they already understand. Jason opens with his own punchout work 26 years ago, and Søren makes the case that the usual setup, a separate punchout catalog replicating your data, gets it backwards. If you’ve already spent heavily on a commerce platform with customer-specific catalogs, contract pricing, and CRM connections, why copy that logic into a third-party system when you can punch out of the platform itself?

They get into why procurement systems were never built for custom configurations, contract versus off-contract pricing, or delivery windows, what Søren calls the “integration premium” of bolting punchout on, and why handing the connection to a middleman puts your customer relationship and purchase data at risk. They also cover why Scandinavia is ahead on digital adoption, why EDI and punchout can live side by side inside a single customer, what “Amazon-like” should actually mean in B2B (relevant, real-time, personalized, and hassle-free), and where AI earns its place, which is cutting punchout customer onboarding from days to hours to minutes.

Key idea: you spent millions building the experience, so don’t hand your customers a scaled-down copy of it.

Stacy Hanks is Vice President of Global Digital Commerce at SureWerx, where she has led go-to-market strategy for pure play and omnichannel, the PIM, and the B2B distributor portal since 2021. Before that she led digital commerce at Master Lock, spent six years at Amazon, and started in outside sales and sales management in consumer packaged goods.

Jason Hein sat down with Stacy to ask what “a B2C-like experience” actually means in B2B, and what happens when a launch doesn’t hit the numbers the board was promised. Stacy’s answer is that commerce is commerce, and that digital results are a math problem: when the math doesn’t math, you go back to the inputs instead of calling it failure.

She walks through two launches of the same portal. In Canada, a rebuilt site with better search and PIM-driven product data grew the share of business coming through the channel by 500 basis points in three to six months. In the US, adoption lagged until she stopped pitching the portal as a website and started offering it as a free tool for price and availability, delivered through customer service and inside sales at the moment customers were already calling in.

They also get into why a website can only reflect what a company already does well offline, why Stacy doesn’t believe in digital MVPs, and why Jason’s favorite metric, lines per order, is really a measure of trust.

Key idea: a website can only be a reflection of what your company does well today.

Anna Harkins is the General Manager of CDS Visual, a visual-first engineering and commercial experience platform that helps manufacturing OEMs grow aftermarket revenue and speed up quote-to-order cycles. She began her career in the chemical, automotive, and food service equipment industries, so she knows what it feels like when a line goes down and nobody can name the part.

Jason Hein sat down with Anna to explore how OEMs can win back aftermarket parts sales by making parts easier to find and easier to buy. Anna explains why complex equipment benefits most from visual parts identification, and why many manufacturers are losing a large share of genuine part sales to third-party suppliers.

She walks through how a buyer can enter a serial number, see a 3D rendering, identify the failed component, and order it online without making a phone call. Manufacturers with only 2D documentation can offer the same experience by converting legacy PDF manuals into drawings with clickable hotspots.

Anna and Jason also discuss how dealers and distributors fit into the picture. Partner Exchange gives partners pricing, inventory, and shipping details in one place, while the complexity stays behind the scenes. The conversation closes with a practical look at the data manufacturers already have, and why retiring experts make this the right time to capture parts knowledge in systems.

Key idea: Make it easy to find the part and easy to buy it, and your aftermarket revenue stays with you.

Ethan Giffin has spent nearly two decades helping mid-market manufacturers and distributors modernize how their customers buy. He founded Groove Commerce in 2007, relaunched it as The B2B eCommerce Agency in Baltimore, and wrote Closing the Digital Revenue Gap.

Justin King sat down with him to dig into a question many manufacturers haven't asked themselves: if the factory floor runs on sensors and AI, why do orders still come in on notepads, spreadsheets, and fax machines?

They get into revenue blind spots, the friction that quietly sends customers to suppliers who are easier to buy from. By Ethan's estimate, a phone order can cost $50 to $70 to process, while the same order placed online costs $3 to $5. Ethan explains why he gets IT, marketing, finance, operations, and sales into one room before any technology is chosen. He also shares why customer service leaders often tell stories the CEO has never heard. The conversation closes on AI, where Ethan points to real returns in catalog building and product search, and Justin makes the case that AI can't be a separate strategy.

Key idea: digital commerce is a revenue channel that touches every department, and it works when the whole company is aligned before the build starts.

Slava Kravchuk is CEO of Atwix, a B2B eCommerce agency that has grown from a one-man open-source shop into a fifty-person team across ten countries, implementing platforms and building AI and automation solutions for B2B companies across the US, Canada, and Europe.

Jason Hein sat down with Slava to unpack why choosing an eCommerce platform is only half the decision, and why the implementation partner might matter more. Slava breaks down what actually separates a good SI from a bad one: real experience with your industry, your product, your customers, and your ERP, not just a logo on their website. They dig into why every technology pitch now comes wrapped in AI hype, how to cut through vendor noise that's increasingly AI-generated on both sides of the table, and why the smartest move for most B2B companies isn't a big-bang replatform but a fast, focused release that proves value before you scale.

They also tackle the system nobody wants to touch: the legacy platform kept alive by one person who understands how it's wired, and what happens when that person, or the vendor's support, disappears. Slava makes the case for facing end-of-life risk head-on rather than waiting for a crisis to force the decision.

Key idea: The platform can be perfect and the project can still fail. The partner is the variable that actually determines the outcome.

Sean McAuliffe spent 20 years in wholesale distribution, building a car-key and locksmith supply business from a couple hundred dollars into an eight-figure company before selling to private equity in 2019. Along the way he ran sales, purchasing, picking, packing, and marketing, and later served as CMO for the company that acquired his business.

Jason Hein sat down with Sean to challenge one of B2B's favorite excuses: "my customers won't respond to that." Sean makes the case that the person placing a B2B order is the same human who shops for themselves after 5pm, and that consumer tactics like flash sales and urgency work just as well in wholesale distribution, sometimes better. He walks through the Thursday flash sales he ran that trained an entire customer base to watch their inbox, why the cutoff had to be real ("real fake scarcity"), and what happened to average order size once buyers were on the site.

They also get into why every recorded sales call is an untapped FAQ database, how to turn scattered application knowledge into something AI can actually use, and the simple math that shows how many people a website can reach compared to a sales team, no matter how good that team is.

Key idea: if you're going to manufacture scarcity, it has to be real fake scarcity.

Kris Harrington is CEO and co-founder of GenAlpha Technologies, where she helps manufacturers and dealers move from static, physical support to connected digital experiences, spanning product configurators, eCommerce, warranty management, and interactive manuals. Her path started in the US Navy as a machinist mate, then took her through Bucyrus International (later acquired by Caterpillar), moving from financial analyst into parts, service, and equipment sales.

Jason Hein sat down with Kris for his first episode hosting the B2B eCommerce Show to dig into a question most B2B websites still can't answer: what happens when a customer knows their problem but not their solution. Kris breaks down why the industry has mastered transactional parts support while still struggling to help buyers who show up with a symptom instead of a part number, and how AI and connected data can turn scattered manuals and tribal knowledge into something a technician can query in real time.

They also get into what happens to experienced salespeople once routine work is automated away, and why aging-out institutional knowledge is quietly one of the biggest risks in B2B today. Kris closes with a reflection on her own career: the years GenAlpha spent building a strong product without marketing it, and why she wishes she'd raised her hand sooner.

Key idea: the part number was never the hard problem. Helping someone find the part number they didn't know they needed is.

Dylan Telford spent more than 20 years across store, digital, and supply chain operations at major omnichannel retailers before moving into consulting, where he now leads Order Operations at Aries Solutions, a systems integrator based in Columbus, Ohio, with teams spanning the US, Canada, India, and Latin America. 

Rutvee Shah has spent 14 years building and scaling SaaS platforms for eCommerce and order management, and now leads product at Kibo, a B2B commerce platform built for manufacturers, distributors, and wholesalers running complex, high-volume operations.

Justin King sat down with them both to work through a question that comes up constantly in manufacturing and distribution: when a company's ERP starts straining under the weight of order routing, fulfillment logic, and inventory decisions, does the answer really mean ripping it out, or does it mean rethinking what the ERP should own in the first place?

They get into why the ERP remains (and should remain), the system of record for finance, why decades of stacking operational complexity on top of it quietly drives up a company's cost in people rather than software, and what a modern order management system actually does once that complexity moves out, from real-time inventory visibility to rules that protect priority accounts when supply runs short. 

Dylan and Rutvee also walk through the fulfillment challenges that rarely come up in a B2C conversation, including multi-location consolidation, contract-specific shipping requirements, and minimum shipment thresholds, along with the reverse logistics side of B2B that most platforms weren't built to handle. The conversation closes on AI, where Rutvee draws a hard line between AI as a recommendation layer and AI as a decision-maker, and Justin pushes back with a case for phasing in trust the same way you would with a new employee.

Key idea: the ERP still works fine, but companies have been loading it up with jobs it wasn't built to do. Moving order management into a separate system fixes that.