Why the Firms That Once Organized Categories Must Now Organize Trust

Why The Next Analyst Category Will Be Built On Trust, Not PDFs

Analyst firms are being mispriced by the market because they are being misdefined by the market.

LLMs are better synthesizers of information than humans. They are faster, cheaper, and dramatically less interested in sleep. If the analyst business were really just “read a lot of stuff, summarize it, and turn it into a chart,” the machines would already have the office keys and half the humans would be updating LinkedIn with the phrase “exploring new opportunities.”

But that was never the real business. Gartner, Forrester, and IDC have always mattered at the moments when raw information stops being enough: When buyers are trying to figure out what problem they actually have, what options are credible, and whether a big decision is about to become a very expensive new regret.

“Analyst firms are being mispriced by the market because they are being misdefined by the market.”

That role matters even more now because the interface has changed. Gartner has warned that traditional search behavior is shifting toward AI chatbots and virtual agents. Forrester is warning that machines are increasingly a primary content audience because they help determine what buyers see, while its current buyer research shows that buyers are using AI heavily but still validating what AI gives them against trusted outside sources. IDC, meanwhile, is explicitly repositioning around workflow-ready intelligence with provenance, citations, confidence signals, and human oversight built into the answer layer itself.

Public markets can see the pressure, even if they do not yet fully understand the opportunity. Gartner remains huge. Forrester is under heavier strain. IDC is private, but its public language has shifted decisively toward embedded intelligence, enterprise AI workflows, and trust as product. The market still seems to be looking at these firms and thinking: “Ah yes, expensive summarizers standing in front of a machine.” That reading is too shallow.

What is actually happening is bigger than a delivery-channel change. The analyst category is moving from category maps and portals to connectable, citable intelligence inside enterprise AI. The firms that win will not just publish research. They will become governed sources of truth that models can call, cite, and defend in the workflow where decisions are being made.

TL;DR: Analyst firms are not dying. They are being forced to evolve. In a world where buyers increasingly discover solutions through AI-mediated answers rather than portals and PDFs, the firms that win will move from category maps to problem frames, from paywalls to protocols, and from static research to something more strategically important: Decision Trust Infrastructure.

Table Of Contents

D3C: From Seller Stages To Buyer Stages

Most go-to-market frameworks are seller anatomy: Funnel, pipeline, handoff, forecast. Useful, sure. Also a little like asking the company to narrate its own autobiography and then acting surprised when it makes itself the hero. What those frameworks often miss is the buyer’s actual experience of moving through uncertainty. That is where D3C comes in.

D3C is a buyer-centric way of understanding how buying actually happens: Discovery, Create Confidence, Commit.

Discovery is the moment the buyer says: What exactly are we trying to solve, and what kind of thing might solve it?
Create Confidence is the moment the buyer says: How do I know this is the right call and not just a beautifully branded mistake?
Commit is the moment the buyer says: Now that we are doing this, how do we make it work, prove it worked, and avoid getting murdered by procurement, finance, or reality?

D3C is useful because it clarifies the outputs as well. Discovery produces a Solution Thesis. Create Confidence produces a Defensible Decision. Commit produces Durable Outcomes. That is a much better way to understand the analyst category than pretending the report itself was the product. The report was the visible artifact. The deeper value was always in helping buyers move through uncertainty.

That matters more in the AI age, not less. Point-and-click is giving way to chat. Search is giving way to answers. Category pages are giving way to synthesized recommendations. But none of that removes the need for trust. It relocates it. The report was never the whole business. It was the receipt for a trust transfer.

Back to top

Discovery: From Category Maps To Problem Frames

Today, analyst firms are strongest and most visible in Discovery, and the mechanism is familiar: Category maps. Gartner has Magic Quadrant. Forrester has The Wave. IDC has MarketScape. Gartner says Magic Quadrant should be used as a “first step” to understand which providers are worth considering for an investment opportunity. Forrester positions The Wave as a guide for buyers considering purchasing options. IDC says MarketScape helps buyers evaluate products, identify vendors that meet their criteria, and confirm investment decisions. These are all Discovery tools. They reduce chaos, define the field, and make the market legible.

That model made perfect sense when the buyer’s first question was a product question: Which CRM should I buy? Which endpoint platform leads the market? Which ERP vendors belong on the shortlist? In that world, the document was the product, the category map was the interface, and everyone politely pretended the PDF was a natural endpoint of human progress.

But increasingly, that is not how buyers start. Now the first question is often a business problem asked inside an AI system: How do I reduce churn? How do I improve seller productivity? How do I unify customer data? How do I govern AI safely? In one motion, the machine begins to do category formation, vendor filtering, and reputational shaping. Amy Bills at Forrester says answer engines such as ChatGPT, Gemini, Perplexity, and Microsoft 365 Copilot are now among buyers’ first stops, while Emma Mathison at Gartner argues that marketers now have to optimize for both AI-driven answers and traditional search because people are researching longer, considering more options, and demanding more specific, trustworthy content.

That is what I mean by Machine-Mediated Reputation: The composite picture of credibility produced when an AI system blends expert research, peer reviews, public narratives, structured data, and its own inferential machinery into an answer. In the old world, buyers went to the map. In the new one, the map has to show up inside the answer.

“In the old world, buyers went to the map. In the new one, the map has to show up inside the answer.”

“What Magic Quadrant, Wave, and MarketScape did for category-era buying, analyst intelligence now has to do for problem-era buying.”

IDC’s newest public moves make that argument far stronger than it was a few months ago. IDC now says it is redefining its role “from a destination for research and data to an embedded intelligence capability that delivers trusted insight directly into the workflows where decisions are made.” It introduced IDC Quanta as “the technology intelligence layer for the AI economy,” says it is grounded in 60+ years of data and built with help from 65+ customers, and describes it through five design principles: Embedded, Contextual, Secure, Aware, and Rigorous. That is not just a nicer portal. It is the map moving into the answer in product form.

“Leaders don’t need more noise. They need intelligence that shows up in the moment, grounded in data they trust, and ready to use.”


Lorenzo Larini

Chief Executive Officer, IDC

Gartner and Forrester are moving in the same direction, each in its own accent. Gartner’s AskGartner positions itself around citation-supported, source-grounded answers. Forrester AI Access promises trusted advice fast, while Forrester’s content strategy work explicitly says machines are becoming a primary audience because they increasingly decide what buyers see. Category maps are not going away. But they are no longer enough. Discovery is shifting from category-led navigation to problem-led, AI-mediated reputation formation.

Back to top

Create Confidence: From Analyst Validation To Living Trust

If Discovery is where analyst firms are most visible, Create Confidence is where they have often been most valuable. This is also the part of the category that outsiders most consistently misunderstand. Buyers do not care about analyst firms because they can summarize. Buyers care because these firms attach methodology, comparison logic, peer evidence, and institutional credibility to a high-stakes choice. They manufacture confidence.

Gartner makes that layer explicit through Gartner Peer Insights, which it describes as a public platform for verified, first-hand reviews from experienced IT professionals. IDC’s TechMatch integrates peer reviews with IDC research and comparison workflows. Forrester’s Buyer Insights leans less on a giant public review marketplace and more on analyst-backed evidence and buyer data, drawing on benchmarks from more than 17,500 global buyers. Not all analyst firms own the same kind of review asset. But all three are, in different ways, in the confidence business.

This is why peer-review data may become one of the most strategic assets in the next phase of the category. If AI-powered word-of-mouth becomes part of how answer engines decide what to recommend, then peer-review systems stop being sidecars and start becoming infrastructure. But infrastructure needs structure. A glowing review from 2022 is basically an archaeological artifact. In software years, it may as well be etched into pottery. It should not count like a 30-day-old implementation note from a verified user at a similar company, in the same industry, using the same deployment model, for the same use case.

“If AI-powered word-of-mouth becomes part of how answer engines decide what to recommend, then peer-review systems stop being sidecars and start becoming infrastructure.”

That is not a fussy data-modeling detail. That is the product.

You can see where the market is going in the scramble to assemble review data at scale. G2 announced in January 2026 that it would acquire Capterra, Software Advice, and GetApp from Gartner, saying the combined network would reach more than 200 million annual software buyers and nearly 6 million verified customer reviews. Gartner still operates Gartner Peer Insights, while G2 is assembling a broader review network with obvious AI-era ambitions. When companies buy review assets like that, they are not buying comments. They are buying structured trust.

Forrester’s answer-engine work reinforces the same logic from another angle. Its analysts argue that reviews, testimonials, customer stories, and community activity are exactly the kind of third-party evidence AI systems prioritize. In other words, confidence in the AI era will not come from expert opinion alone. It will come from expert opinion fused with fresh, structured peer reality.

“Trust isn’t just a feature. It’s the foundation.”


Michael Facemire

Chief Technology Officer, Forrester

That is where Trust as a Service comes in. When an AI system recommends a vendor, category, or architecture, the buyer increasingly wants to know: Where did that answer come from? How current is it? What evidence supports it? What peer signals reinforce it? What assumptions shaped it? How sure should I actually be? IDC is already using that language through provenance, citations, confidence signals, and human oversight. Gartner’s AskGartner is moving in the same direction with source-grounded answers. Michael Facemire says the quiet part out loud: Trust is not just a feature; it is the foundation.

That is why the next step is not simply better analyst chatbots. It is the productization of trust itself. Analyst firms should expose services that work as verification layers for both people and machines: Confidence scores, evidence trails, freshness indicators, methodology metadata, dispute flags, source lineage, and eventually agent-to-agent trust endpoints. The need for that layer is already obvious. The Tow Center found that major AI search tools were wrong on more than 60% of citation-related queries in its study and frequently failed to provide reliable source attribution. In a world like that, trustworthy verification stops being a nice-to-have and becomes a business.

Back to top

Commit: From Vendor Selection To Durable Outcomes

Commit is the least developed part of the analyst-firm story, which is exactly why it may be the most interesting. Today, the category still tells its story mostly as “Help me choose.” Gartner’s BuySmart already shows how far that has evolved: It supports research, shortlisting, evaluation, negotiation, and proposal review. IDC’s sourcing advisory and Tech Buyer tools stretch into benchmarking, deal review, and negotiation support. Forrester is less branded around a single Commit product, but its Buyer Insights and advisory work clearly extend past awareness into proof, alignment, and selection discipline. So Commit is not absent. It is just not yet the emotional center of the category.

That is the whitespace, because the harder question is no longer just, “Which vendor should I buy?” The harder question is, “Six months later, how do I know I was right?” This is where the old analyst model starts to look oddly unfinished. It helped buyers choose, but it did much less to help them prove, improve, and defend the decision after the ink dried. In an AI-mediated buying environment, that gap matters more, not less. Forrester says buying groups are getting larger, procurement is becoming more influential, and trials are becoming essential to reducing risk. Buyers are leaning on AI, but they are also validating AI’s output with trusted outside voices. That is a giant neon arrow pointing at the next value pool: Not just pre-purchase guidance, but post-purchase confidence.

That opens the door to a much larger Commit layer: Implementation checkpoints, adoption benchmarks, value-realization scorecards, renewal-risk indicators, pricing sanity checks, peer communities, executive roundtables, and agentic assurance services that monitor whether the deployed choice is actually performing the way Discovery and Create Confidence said it would. Discovery produces the Solution Thesis. Create Confidence produces the Defensible Decision. Commit has to produce Durable Outcomes. If it does not, the category has stopped too early.

Back to top

Business Model: From Paywall To Protocol

This is where the story stops being just a positioning essay and becomes a business-model essay. A pure paywall model assumes the user comes to the portal, downloads the report, and reads it. That assumption is weakening. Not disappearing — the PDF survives, just with less imperial authority. It still exists. It just no longer gets to act like Louis XIV. The more strategic asset is not the page. It is the structured, licensable intelligence behind the page: Taxonomies, evaluation criteria, use-case maps, benchmarks, buyer signals, peer evidence, vendor attributes, and analyst judgment that can be accessed in machine-readable ways.

That is why I think the category is moving from Paywall to Protocol.

The old licensing question was, “How do we protect the PDF?” The better question now is, “How do we make the intelligence callable?” The strategic asset is no longer just the report sitting in a portal. It is the firm’s ability to become a live, governed, citable source of truth inside the systems where decisions are being made.

That reorders the hierarchy of licensing opportunities. One-time training deals may still happen. Content syndication may still happen. Vendors will certainly continue quoting analyst sentences with exactly the degree of selective enthusiasm you would expect from vendors. But none of that looks like the strategic center anymore. The strategic center is live, entitled, protocol-based access.

Back to top

Connectors: From Content Licensing To Connectable Intelligence

The cleanest licensing model may not be “sell the PDF to the model.” It may be entitlement-aware connectors.

“The cleanest licensing model is not ‘sell the PDF to the model.’ It is ‘let the model call the firm live, with permissions, provenance, and citations intact.’”

In that model, the analyst firm becomes a governed, authenticated data source inside enterprise AI environments. A customer opens Claude, ChatGPT Enterprise, or another enterprise model, connects its analyst subscription, and from that point forward the model can retrieve live, entitled analyst intelligence inside the workflow where the decision is being made. The customer is not asking the model to remember the analyst firm. The customer is authorizing the model to call the analyst firm.

That is no longer just a thought experiment. IDC has now publicly announced that it is building an MCP server and collaborating with Anthropic to bring IDC intelligence directly into Claude workflows. Through that model, organizations will gain entitlement-based access to IDC’s proprietary research, data, and methodologies natively within Anthropic environments via MCP and plugins. IDC says the result is an embedded intelligence layer, not a separate destination.

The connector pattern is now visible on both major model platforms. Anthropic’s custom connectors let organizations connect Claude directly to external tools and data sources via remote MCP. For Team and Enterprise plans, owners add the connector, users authenticate individually, and Claude only accesses tools and data that the individual user is already authorized to see. Anthropic also says that when Claude uses connected sources, it provides clear citations linking directly to the original material. On the OpenAI side, company knowledge in ChatGPT lets organizations use connected apps to answer company-specific questions with clear citations back to the original sources, while OpenAI’s enterprise release notes say company knowledge respects existing company permissions and only gives ChatGPT access to what each user is already authorized to view.

OpenAI’s MCP guidance makes an important product-design detail explicit: If you want company knowledge compatibility, your app should implement search and fetch and return canonical URL values for citations. That sounds technical because it is technical. It also happens to be strategically profound. A serious analyst connector should not merely return text snippets. It should return a provenance envelope.

  • Who Said It: Analyst name, firm, and research type.
  • What It Came From: Report title, data asset, benchmark, event session, or methodology page.
  • When It Was Published: Date, vintage, and freshness signals.
  • Why It Should Be Trusted: Methodology, confidence notes, and data lineage.
  • Where It Lives: Canonical URL, entitlement tier, and click-through source.

That is the difference between retrieval and evidence. It is also the difference between scraping and presence.

“Scraping creates approximation. Connectors create accountable presence.”

Scraped web content can be stale, decontextualized, partially quoted, stripped of rights context, or repeated by vendors in ways that somehow always make them look unusually heroic. A connector model is different. It preserves freshness because the model retrieves live content. It preserves rights because access is governed by entitlements. It preserves attribution because the analyst brand stays visible inside the answer. It preserves revocability because the connector can be re-scoped, updated, or turned off. And it preserves auditability because every answer can carry back-links, source metadata, and methodology context.

That suggests a much richer revenue stack than the old “sell access to a portal” model.

  • Subscription As Entitlement Object: The subscription becomes the rights system that determines what the model may retrieve.
  • Workspace Connector Fees: Enterprises pay to make the firm available inside Claude, ChatGPT, or other approved AI environments.
  • Premium Data Layers: Forecasts, benchmarks, peer data, or sector-specific corpora can sit above baseline access.
  • Governance And Provenance Premiums: Admin controls, traceability, audit trails, and regulated-workflow features become billable.
  • Human Escalation Paths: Analyst access, working sessions, and high-value advisory rise in value because the machine layer makes the human layer more leveraged, not less.

That pricing stack is an inference, not a published IDC price sheet. But it follows directly from how enterprise connectors, MCP servers, and company-knowledge systems now work. The winning model is not generic scraping. It is connectable intelligence.

Back to top

The Human Layer: From Reports To Interpretation

There is one more implication here that matters a lot. If the machine layer gets cheaper, the human layer does not become less valuable. It becomes more valuable.

Sharyn Leaver writes that the value of a research firm in the AI era is still “largely human,” and Gartner says that by 2030, 75% of B2B buyers will prefer sales experiences that prioritize human interaction over AI. Translation: The machine may be fine for the opening act, but buyers still want a person on stage when the solo matters.

“By 2030, 75% of B2B buyers will prefer sales experiences that prioritize human interaction over AI.”


Gartner

Sales Research, 2025

“The report does not disappear. It just stops being the entire kingdom.”

That has real revenue implications. As recorded insight becomes easier to summarize, direct analyst access, working sessions, scenario labs, peer communities, executive roundtables, and live guidance become more important. The music business did not stop caring about recordings, but the tour got more important once recorded media became less economically privileged. The same logic applies here. The report does not disappear. It just stops being the entire kingdom.

That shift may also change the profile of the analyst. The analyst of the future is probably less pure research academic and more teaching academic: Someone who still does original analysis, but is better at live interpretation, collaborative explanation, scenario coaching, and making buyers smarter in real time. The old analyst archetype was “person behind a report.” The newer one may look more like “person behind a trust system.”

Back to top

Category Claim: From Analyst Research To Decision Trust Infrastructure

At this point, the opening should be obvious. The category race has started, and one of these firms is going to articulate the future faster than the others.

Among the major firms, IDC is now speaking this language most explicitly and most concretely. AI Vision first framed the move around workflow-ready intelligence, provenance, citations, and human oversight. IDC Quanta now gives that move a stronger product shape: Embedded, Contextual, Secure, Aware, Rigorous, and directly connected into enterprise AI workflows via MCP. Gartner and Forrester are also moving in this direction — AskGartner, Peer Insights, BuySmart, AI Access, Buyer Insights, machines as audience, trust-first delivery — but IDC is currently giving the future the clearest public architecture.

Someone is eventually going to call the thing what it is. Better to do it before the market does it badly and a consultant shows up with a laminated framework and a tragic name for it.

I would call that future Decision Trust Infrastructure.

The name matters because it makes the underlying shift legible. The future firm is not just a publisher of research. It is infrastructure for decisions. It does not just produce insight. It operationalizes trust.

  • Problem-First Discovery: It helps buyers solve business problems, not just navigate product categories.
  • Machine-Readable Trust: It turns confidence into citations, provenance, freshness, methodology, and verification signals that machines can use.
  • Protocol, Not Just Paywall: It licenses intelligence into workflows, models, and systems rather than waiting for users to visit a portal.
  • Living Peer Evidence: It treats customer voice as a fresh, structured, continuously useful layer of confidence.
  • Commit-Level Value: It stays valuable after the purchase by helping buyers prove, improve, and defend the decision over time.

“The future of analyst firms is not research. It is trust made operational.”

Seen through D3C, the analyst firm was never just a report factory. It was a Discovery engine disguised as category research, a confidence engine disguised as methodology, and a not-yet-finished Commit engine waiting for the market to ask more of it. The chart survives. The report survives. What changes is their status: They stop being the product and become artifacts of a much larger one.

The firms that win will not simply publish research. They will license intelligence into machines, attach trust to answers, preserve provenance by design, and stay useful after the purchase. What emerges is not just the future of analyst research. It is the future of Decision Trust Infrastructure.

Back to top

Further Reading

Start Here

Go Deeper

Firm Signals

Independent Voices

Answer-Engine Discovery

Trust, Verification, And Protocol

Connectors, Company Knowledge, And Provenance

Human Interpretation And The Revenue Mix

Back to top

Written by

Ever since I was a kid, I’ve wanted to be an astronaut — to explore the unknown, map new frontiers, and make the impossible routine. And since I couldn’t get to space, I’ve done that in marketing instead. Instead of The Right Stuff, I’ve been more focused on The Write Stuff. Across five CMO tours — at Microsoft, Sprinklr, PROS, Pendo, and a few bold startups — I’ve spent my career exploring new frontiers of growth.