Manufacturers Don't Go Direct. Catalogs Do.

Distribution Strategy Group and McKinsey are arguing about channel economics — inventory pooling, safety stock, who picks up the phone. That argument is a decade old and mostly settled. The one that isn't settled: when a buying agent asks "what replaces this part," who gets to be the answer — and most distributors are unwittingly answering "the manufacturer."

The argument everyone's already having

Distribution Strategy Group's rebuttal to McKinsey makes a fair point: McKinsey's direct-to-consumer data mostly captures field sales and phone orders that were "always sold direct," not a new wave of disintermediation, and distributor shareholder returns since 2020 don't look like an industry being hollowed out. DSG's own math is inventory pooling — volume across brands and customers grows faster than volatility, so a distributor can run leaner safety stock and still hit service levels no single manufacturer can match alone. That's a real, durable advantage, and it's the same thesis DSG has been running since at least "Positioning: Value Matters" and "Bridging the Value Gap" in 2013: distributors win by adding value a manufacturer can't replicate operationally.

Both McKinsey and DSG are still fighting the 2013 war, though. They're arguing about who ships the box. The buyer increasingly doesn't care who ships the box — they're asking a model who ships the right box, and the model answers from whatever text it can parse fastest.

Disintermediation now happens at the answer, not the channel

The mechanism is different from a manufacturer opening a webstore. A multi-source 2026 analysis found 73% of B2B buyers now use AI tools like ChatGPT or Perplexity during purchase research, and Forrester's 2026 buyers' journey survey puts the figure closer to 94% of buyers using AI somewhere in their most recent purchase. Traffic from AI answer engines is converting at roughly 14.2% versus 2.8% for organic Google traffic — a five-fold difference that tells you these aren't casual lookups, they're near-decided buyers hitting a page an AI already vetted for them.

Here's the part the going-direct debate misses: an AI agent doesn't route a query to a company. It routes to whichever document answers the question in the fewest inferential steps. Ask it "what replaces this part," and it doesn't consult an org chart of who's the "real" seller — it finds the page with the cleanest structured answer and cites that page. If that page happens to sit on manufacturer.com, the agent has no reason to mention the twelve distributors who also stock the part. The manufacturer becomes the transaction endpoint not because it built a better funnel, but because it published the parseable answer first.

The self-inflicted version of this problem

Now the uncomfortable part: distributors are frequently handing this outcome to manufacturers voluntarily. Pull up ten distributor PDPs for the same SKU and it's common to find identical copy on all ten, sourced straight from the manufacturer's spec sheet — the kind of verbatim duplication SEO analysts have flagged for years as a ranking liability, and one 2026 ecommerce content guidance now frames as needing roughly 60% SKU-specific dynamic content against 40% fixed manufacturer boilerplate just to avoid being filtered as a duplicate.

Search engines used to just demote the loser of that duplicate-content fight. Answer engines do something worse: they pick a canonical source and cite it as the source of truth. When the content is identical, the canonical source is definitionally the manufacturer — it's their copy, their spec sheet, their catalog of record. A distributor running verbatim manufacturer descriptions isn't neutral in that outcome. It is training the model, page by page, that the manufacturer is where the real answer lives, and that the distributor is a redundant hop.

What a manufacturer structurally cannot publish

The fix isn't better copywriting on the same content. It's publishing the layer a manufacturer has no institutional reason — and often no ability — to produce:

That's not a marketing angle, it's a structural asymmetry. A manufacturer optimizes one brand's parseable answer. A distributor sitting across dozens of competing brands is the only party positioned to answer "what replaces this" honestly, across the whole category — which is exactly the question buying agents are increasingly asking first.

This is measurable, and most distributors aren't positioned for it

This is the argument behind Anglera's Digital Readiness Index — four pillars, fourteen signals, measured directly off each distributor's live site rather than self-reported. It scores exactly this gap: whether a distributor's product pages carry structured, comparative, substitution-aware data an agent can actually parse, or whether they're a manufacturer's PDF converted to HTML. Across the Top Distributors 2026 index, the distributors clustering in the strongest archetypes aren't necessarily the biggest by revenue — they're the ones whose catalogs say something a single manufacturer's site structurally cannot.

We didn't build Anglera to fight the going-direct debate on channel economics; your PIM already handles pooling, pricing, and fulfillment logic fine. We built it because turning "here's the spec sheet" into "here's what replaces it, and why, across every brand you carry" is a data-enrichment problem, not a sales-strategy one — and most catalogs aren't structured to answer it yet. That gap is closeable in weeks, not a re-platforming project, and it's the difference between being cited and being skipped.