200 years of talking

Here is one thing nobody puts in the Cliff Notes on ecommerce. The reason it won is boring: it gets to watch you.

In 2006 Tobi Lütke, Daniel Weinand, and Scott Lake turned a snowboard shop's homemade storefront into Shopify. Nothing in it was new – Carts, checkout, product pages – all of that existed. What they did was hand a guy selling three snowboards a week the same dashboard Amazon was using to scale billions in revenue. He could see a shopper hover on a board, add it, swap the size, and leave at shipping. He looked at a few hundred of those and moved a button. Sales went up a little. He did it again on Tuesday. Twenty years of Tuesdays and online retail takes more than a sixth of US retail.

At some point in the early 2010s the pressure got high enough that staying offline started costing a shop customers. In a very short period of time, from roughly 2010 to 2015 Shopify's store count was 35x. That forcing function came to be known as the Shopify moment.

B2B still hasn't had it.

Buyer. For centuries buyers have described what they want in a sentence. No system could ever read that sentence, so the order cannot become data until a person types it. Now AI reads the sentence, so an order becomes structured the moment it is said.

Supplier. Every brand keeps its own portal and spreadsheets so nothing can compound. There is no universal standards, shared data shape, or common system of record in wholesale trade. Because of this baseline, there is no way to test optimize and improve.

Rep. Oversight on a fifty thousand dollar order is rational, and the rep is the one who can turn a half-formed request into a correct order. Nobody was going to replace that with a cart. The cost is that reps spend about 28% of their time selling and the rest on admin. Keep the rep, remove the admin, and their judgment becomes pure leverage.

The common thread is that nothing before the PO gets recorded.

Consumer retail runs on three levers, customers, frequency, and AOV. When Shopify helped sellers optimize those three things ecommerce doubled in 5 years. Wholesale has its own three, doors, reorder cadence, and order size, and no brand has ever been able to optimize them. Why? Because the thing that moves them – what the buyer asked for and did not get – cannot be recorded.

That is what the sentence being readable changes. AI has 200 years of talking it's first recorder. The question is an event now. The substitution is an event. The "actually make it 24" is an event. Wholesale gets the loop consumer retail has been running for twenty years, and the buyer does not change any behavior to use it.

Forcing Function

When you study the numbers, 2025 looks a lot like the Shopify moment. Wholesale's back office has already modernized: cloud ERP went from a quarter of installs in 2014 to two thirds today, portals just crossed 20% of suppliers, and live inventory and pricing is standard. Retail's moment was two things landing at once. Shops started losing customers to competitors with a checkout, and a storefront cost a hundred dollars a month, so the shop moved and the moving got watched. Wholesale had the tools for fifteen years and nobody moved, because no one was losing accounts over an order form. Now they are.

This is today:

53%of buyers say they have switched suppliers over a bad ordering experience.
83%now prefer to order digitally.
39%will place a $500,000 order self-serve, up from 28% two years ago.

Staying offline has started costing accounts, and the portal cannot absorb that. It asks the buyer to stop talking and start clicking, which is the one behavior wholesale buyers have never done.

I founded Brio because an AI model can now read "same as last year, one more run in medium, drop the tan" and check it against the ERP while the buyer is still speaking. Wholesale gets watched without anyone learning a new screen.

Drop your email here to check it out, I demo Brio every Thursday.

Charlie

Longer version, with the numbers:
The Bottleneck in B2B Trade ↗