B2B eCommerce Leader.
18+ years helping Brands, Manufacturers & Distributors launch and scale.
Featured in Forbes
Chicago | Atlanta | Miami | Paris
Master B2B Nexus | B2B eCommerce Association Ambassador
Sonepar
Goodyear
Kate Spade
Coach
Ruggable
Your customer doesn't gain anything by switching
— unless you give them something the phone, the rep, and EDI can't.
The five reasons, in the order I see them:
If your adoption is stuck, it's one of these five. Not 'we need more AI.'
It's hierarchies, permissions, contracts, blanket orders, approvals, credit limits, PunchOut, EDI, projects, payment terms.
If your portal is 'wholesale pricing on a B2C theme,' you've built a B2C site for buyers who weren't asking for one.
- Personalization is important — you already know what's on their contract
- Search matters less than catalog completeness
- The job is removing reasons to not switch
Stop optimizing the homepage. Audit the friction.
9 plays. We'll go fast.
- 4 customer experiences, each tuned to a different buyer segment
- Checkout: from 3 steps to 1
- Not because we 'simplified UX' — because we matched what the buyer actually does
Audit every field. Each one is either earning its place or costing you adoption.
If your checkout has more steps than the phone call, you've already lost.
The deepest version of 'match the buying process.'
A construction or industrial buyer doesn't order for a company. They order for a job site.
If your portal forces them to order at the company level, they'll call the rep.
Applies to: construction material distributors, electrical and plumbing distributors, industrial supply, rental and equipment suppliers — anyone whose buyers manage multiple active projects.
- Buyer selects the job site before adding products to cart
- Every order tagged to a project — no manual sorting after the fact
- Order history filtered by job, not just by date
- Delivery addresses tied to job sites, not company accounts
- Spend by site, real-time
Can your portal do project-level, or only company-level?
Every step the phone doesn't have:
- Login walls before pricing
- Captcha
- 'Confirm address'
- Re-typing a PO number
- Pricing visible only after add-to-cart
- Order status without a phone call
- Expected delivery date in real time
- Named rep on the account
- One-click reorder from invoice
- Blanket order burn-down
The phone gives you 'I'll check and call you back.' The portal should give you the answer in one click.
Two internal teams kill or save portal adoption: sales and customer support.
- Comp alignment — self-serve orders count toward rep quotas
- Name the rep in the portal — the customer still belongs to them
- Train reps to demo the portal cold
- Weekly meeting with the call center. Co-build features with the people answering the calls.
- 3-email rule for technical support. If it takes more than 3 emails, switch to a live meeting. Always.
Reps and CSRs are the two adoption levers nobody tracks.
Adoption is field work, not a UX project.
Train customers on their turf, with their rep beside them
Dedicated team that visited branches and job sites
Embedded inside the management platform Quebec electricians already used every day
Meet customers where they already work. The portal comes to them.
You'll learn more in a day at a client than in a week brainstorming with your team.
Zero-result queries are your roadmap.
- What they searched and didn't find = SKUs to add or rename
- Top searches on items with no images = your highest-leverage merchandising
- Failed searches by repeat customers = a phone call you're about to lose
Most distributors have this data. Almost no one looks at it weekly.
There's a cottage industry telling you to optimize for AI. Add FAQs. Add schema. Get cited. All of that is fine. None of it is enough.
That's the GEO trap. You optimize to be found, and you get found — as one of fifteen interchangeable options.
A buyer, an AP clerk, a branch manager, and an end-user need different views.
Husky's 4-experience setup is the canonical example. Build for the role, not the company.
This is where the portal beats the phone.
The rep can't tell you in 5 seconds:
- Your full purchase history across 18 months
- Your blanket order burn-down
- Your contract pricing on a SKU you've never bought
- What you ordered for this job site last quarter
The portal can. And AI makes it scalable:
Fed by product docs + FAQ + resolved support tickets → deflects 30–40% of incoming volume
UserPilot-style guides buyers through the first 3 orders
Every resolved ticket trains the system → next customer gets the answer in 5 seconds
This is your AI moment. Not 'AI everywhere.' Specifically: scalable support and contextual recall.
What's the ROI when integrating a new customer. Not all integrations drive adoption equally:
ERP integration → drives the biggest accounts (real-time pricing, stock, order status)
- Mobile App → drives field and branch buyers (Growth lever)
- PunchOut → privileged path for enterprise procurement
- EDI → last resort. Offer it when you've lost the fight to keep them in the portal.
EDI is not the sophisticated answer. It's the defensive one.
Stop tracking 'portal sessions.' Track:
If you can't show movement on these in 90 days, the playbook isn't being run.
The portal itself is changing.
Ordering on behalf of buyers
The portal stops being a UI for humans and becomes an API for agents
If your PIM is messy and your product data is incomplete, you're invisible to the next generation of B2B buyers
The next adoption fight is GEO/AEO — getting your products surfaced in ChatGPT, Perplexity, and Copilot when buyers ask.
Read it. There's a roadmap in there.
Listen for 'I just called instead.'
The answer will surprise you.
That's it. Don't buy a new platform. Run the plays.
I write about this every week. Practical stuff, no hype.
DM me on LinkedIn
RudyAbitbol.com