Three people who live in B2B infrastructure every day sat down to talk about ERP-connected commerce:
- Aaron Sheehan from OroCommerce, who works at the intersection of product, partners, and architecture
- Brendan Cameron from Americaneagle.com, who designs and fixes complex B2B stacks
- Christine Martin from Avalara, who lives in the world of tax, compliance, and finance teams.
They weren’t debating whether to connect ERP, commerce, and tax. Their customers are past that stage. The focus was what comes after: where ERP reaches its limits, how tool sprawl erodes trust, and what it takes to build a setup teams can depend on.
Didn’t catch our recent ERP-connected commerce session?
Takeaway 1: Your ERP Can’t Carry Every Idea the Business Has
For years, the default instinct was simple: if the ERP runs the business, it should run everything else too.
Brendan sees the limits of that thinking up close. He described an electrical distributor on a big-name cloud ERP, eager to offer subscriptions for common consumables: gloves, earplugs, disposable gear.
The problem? Their ERP had no concept of a recurring order.
“There was no way for their ERP to say this order should repeat every month… reporting couldn’t tell subscriptions from one-offs,” Brendan explained.
The workaround pushed subscription logic into the commerce layer, which re-submitted each cycle as a brand-new order. The ERP saw identical orders, not a pattern. Questions about monthly recurring revenue turned into manual detective work in order history.
Aaron’s FileMaker example shows the same instinct from another angle. One “system” became dozens of small apps, each built to answer a local request.
“Every department head could ask for a table and a form. Over the years that turned into a pile of micro apps feeding separate databases.”
Reporting across the business became nearly impossible.
Both stories land in the same place: ERP should stay the backbone for finance and operations, not the dumping ground for every new idea, channel, or buying model.
Takeaway 2: Tool Sprawl Kills Trust Faster Than Any UX Problem
You can’t promise “one experience” to customers if every system is telling a slightly different story.
Once teams add CRMs, marketplaces, order management, and tax tools on top of ERP, the work shifts from “having systems” to “keeping them in sync.” That’s where trust starts to wobble.
Christine sees it first inside the business. Tax logic, exemption rules, and reporting all need to live in multiple places at once. ERPs, CRMs, marketplaces, order systems. Keeping that aligned by hand is a full-time job.
Brendan sees the fallout at the edge. A buyer checks the webstore, calls their rep, maybe places orders through punchout. If any of those touchpoints show a different price or tax treatment, confidence evaporates. One bad quote, and they stop trusting the system and go back to email and phone.
Christine added the tax “shortcuts” she still runs into:
- Discounts quietly used in place of tax.
- Exemption certificates that expired months ago.
- IT buying a tax platform without telling accounting, then expecting them to own it during year-end close.
None of this is about design or front-end polish. It’s what happens when you let tool sprawl grow faster than ownership and shared rules. The buyer sees one brand. If your systems don’t, they will feel it.
Takeaway 3: Consolidation Moves When Someone Owns It
You can draw the cleanest future-state architecture you want. If nobody owns the change, nothing moves.
Brendan’s first move with any client is to work out who will carry the roadmap when priorities clash and projects start slipping to “later.” At Wastequip, that role was clear. They switched ERPs while running a strong digital business and launching new sites in parallel – work that usually stalls for months.
“It takes a champion… a lot of the way the organization matured over time wouldn’t have happened without Kevin putting his thumb on the scale.”
From there, Brendan leans on structure rather than gut feel.
- A quick SWOT to surface where the biggest problems and opportunities are.
- Then a change framework like McKinsey’s 7S to force specifics: who owns the data, who approves changes, who benefits, who’s already overloaded.
“It puts everyone in a room together… from who’s owning the data to the process to the approvals… so you’re not surprised six months into a project being like, ‘We’re missing this very critical piece.’”
Christine described a similar pattern at a customer running 7–10 acquisitions a year. They use that same model to drive every integration: Ops, finance, IT, HR all pulled into one approach, with a defined sequence from the month before close through the first ten days on new systems.
Consolidation moves when there’s a named champion, a repeatable way to decide, and clear responsibilities. The architecture only works if that part is in place.
Takeaway 4: AI Makes Your Data Impossible to Ignore
When the panel got to AI, nobody treated it as a shortcut. They treated it like pressure on everything underneath.
Brendan’s rule is simple: “You should always play with AI. You should never trust it.”
He talked about the tradeoff teams forget to price in. If a task takes four hours today and AI can do a first pass in one, that sounds great. But if you then spend eight hours fixing the output, you’ve gone backwards.
His focus stays on structure and data. He expects most serious platforms to plug into things like Model Context Protocol (MCP) and small language models over company data. That only works if the source systems are clean and connected.
“Good data’s the answer to everything. AI is the tool which accesses it.”
Aaron tied it directly back to commerce. Many manufacturers and distributors still hesitate to expose pricing, inventory, or product data because they don’t trust it.
“If you threw an AI on top of that crappy data, it would not make for a better customer experience… that customer journey would still end up in an email or phone call.”
In their view, AI doesn’t fix weak architecture. It just makes the gaps in ownership, modeling, and data quality a lot harder to ignore.
What This Means for B2B Teams
Most of the drag in B2B commerce comes from workarounds, scattered ownership of data, and changes that never quite line up across teams.
The thread running through this panel: ERP-connected commerce works best when you design for clarity (who owns what, which systems are trusted, and how changes flow) before you add more tools or AI on top.
