Your Storefront Can Take the Order. It Cannot Help the Customer Decide What to Order.

The AI budget went to a pricing model, a demand forecast, a center of excellence, and the storefront your customers use every day is still search-and-list. If you run commerce or IT at a distributor, that sequence probably describes the last two years of your own roadmap.

Most of the largest distributors in North America have told their investors they are investing in AI. Far fewer have put any of it where a customer would notice. McFadyen Digital’s 2026 AI-Guided Buying Maturity Study scored 37 of them on how far AI has reached into the buying experience itself: 59% of the companies actively promote AI investment in investor materials and earnings calls, while 16%, six companies, have built AI meaningfully into how customers find, evaluate, and purchase (McFadyen Digital, 2026 AI-Guided Buying Maturity Study). Roughly four in ten distributors talking about AI have none of it visible to their customers.

You do not need to be one of the 37 for the finding to apply. The five dimensions the study scored are the same five any distributor can hold its own storefront against.

What the index measures

The study assessed each company May to June 2026 on a 25-point index: five dimensions, each scored 0 to 5, from public evidence only (storefronts, investor materials, earnings transcripts, trade coverage, vendor case studies). Anything behind a customer login was scored on announced or reported evidence.

Dimension	Plain-language reading
A. Assistant presence	Is there a product-aware AI assistant in production for customers?
B. Product discovery intelligence	How well do search and recommendations understand products and the customer?
C. B2B workflow support	Contract pricing, quick order, quoting, punchout, approvals: the plumbing of B2B buying.
D. AI and digital innovation signaling	What the company says and shows about AI investment aimed at customers.
E. Enterprise readiness	The platform, data and integration foundations underneath the other four.

Scores sort into five tiers.

Tier	Score band	Companies in the study
AI-Guided Buying Leader	21 to 25	6
Assisted Commerce Emerging	16 to 20	9
Modern B2B Commerce	11 to 15	12
Basic Digital Self-Service	6 to 10	8
Traditional Digital Catalog	0 to 5	2

The average score was 14.2 of 25, which lands in the Modern B2B Commerce tier (McFadyen Digital, 2026 AI-Guided Buying Maturity Study). Ingram Micro (24), W.W. Grainger (23), and Watsco (22) lead the index; Thermo Fisher Scientific, Sysco and Ferguson are tied at 21. Those scores are a snapshot as of May to June 2026, built on public evidence. They are structured judgment, not an audited benchmark, and the companies named did not review or endorse them.

The plumbing is finished. The judgment is not.

The gap is sharpest between two of the five dimensions. Of the 37 companies, 62% scored a 4 or 5 on B2B workflow support, while 14% scored a 4 or 5 on assistant presence (McFadyen Digital, 2026 AI-Guided Buying Maturity Study). Roughly one in seven has a product-aware AI assistant in production.

The paper’s summary of that spread is the line we would ask you to remember: distribution has digitized its transactions without digitizing its judgment. The systems can take an order. They cannot yet help a customer decide what to order.

Follow the AI dollars and the pattern explains itself. One electrical distributor committed $500 million to a multi-year digital program and another created an executive role for AI and digital, both aimed at internal quoting tools and data backbones (McFadyen Digital, 2026, citing trade coverage). In building products, AI is almost entirely dynamic pricing and demand forecasting. In metals, the frontier question is still whether a customer can transact online at all.

Both tiers tell investors they are investing in AI. Only one tier’s customers can feel it.

Two rankings measure two different things

Distribution Strategy Group published its AI Top 25 in July 2026, ranking distributors on verified production AI anywhere in the business. Ingram Micro and Grainger are top tier on both lists. Wesco International and Fastenal are top tier in DSG’s ranking and score 14 of 25 in ours, under the same snapshot caveat as every score above. Fifteen companies in our bottom two tiers do not appear in DSG’s ranking at all (McFadyen Digital, 2026 AI-Guided Buying Maturity Study). Being an AI leader in distribution today does not yet mean being an AI-guided buying leader. The second question is the one your customers are asking.

Buyers see the same gap from the other side

Deloitte’s B2B agentic commerce study surveyed 530 US buyers and 530 US suppliers and found a mirror image: 72% of suppliers describe their sales processes as mostly or highly automated, and 47% of buyers agree (Deloitte, 26 June 2026). In the same survey, nearly 40% of buyers already use agentic AI in purchasing, against 24% of suppliers using agents in sales (Deloitte, 26 June 2026).

Suppliers grade their automation from the inside, where the ERP integrations and the quoting tools live. Buyers grade it from the storefront. The two views are 25 points apart, and buyers are already bringing agents to the transaction.

What we think you should do with this

Our reading of the data: a distributor does not need a chatbot. It needs an AI buying layer that understands products, customers, pricing, inventory, contracts, availability, and purchasing workflows. At most of the 37 companies, the missing piece is the connection between AI and that commerce workflow. The models exist. The plumbing exists. Six companies have wired one to the other where the customer can feel it. Most have not, and the reason is more specific than ambition.

The connection is made of foundations the paper lists as Phase 0 and calls “not optional”: product attributes complete by category, live pricing and inventory APIs, and a commerce platform built to be extended, typically three to nine months before any customer-facing AI is attempted (McFadyen Digital, 2026 AI-Guided Buying Maturity Study). It stays unmade because of ownership. The most common failure the paper describes is an AI team shipping operational wins indefinitely because the commerce backlog belongs to someone else (McFadyen Digital, 2026 AI-Guided Buying Maturity Study): the pricing model ships, the forecast ships, the storefront waits. The third and fourth posts in this series take up the foundations and the ownership question.

The gap has a price on the record. US Foods, at 20 of 25 and one tier below the Leaders, reported that embedded merchandising tools in its MOXē ordering platform raised average order size by more than 1.5 cases since launch (Digital Commerce 360, 10 June 2024). At the top of the index, Ingram Micro reported that by the second quarter of 2026, opportunities supported by its intelligent digital assistant were converting at nearly four times the rate of traditional quotes (Ingram Micro Q2 FY2026 results, 30 July 2026). Two units, two tiers of the index, and a measurable return in both.

One action. Before you fund the next pilot, score your own buying experience on the five dimensions above, honestly, from the customer’s side of the login. The study gives you the rubric and the tier definitions. If you land where most of the 37 did, strong on workflow support and weak on discovery and assistance, the next post in this series covers where to start (embedded AI over visible AI). Our earlier argument about AI tourism made the same case without the measurement. This study is the measurement.

McFadyen Digital, which published it, is a B2B commerce consultancy: we assess product data readiness, architect the integration layer between AI and commerce systems, and build guided buying experiences for distributors.

Read the study: The AI-Guided Buying Gap. 32 pages, ungated, with the full index, sector averages, six design patterns, and a phase-by-phase sequencing framework.

If you would rather have the first read done for you, the McFadyen AI Commerce Readiness Audit is a free, structured look at product data readiness and commerce architecture extensibility, the two foundations the study says every later phase depends on.