When Information Stops Being the Business Moat

When Information Stops Being the Moat

When basic information becomes easier to find and synthesize, a professional-services firm's advantage moves toward the context others cannot reconstruct quickly, the judgment required to make a decision, the integration needed to execute it, the accountability attached to the result, and a service relationship that continues to earn its economics.

Information does not become worthless. The weak claim is that access alone remains a durable reason to pay one firm instead of another.

The client was never buying a document

A market report, legal memorandum, underwriting file, offering memorandum, valuation, or strategy deck is the visible artifact. The client is usually buying a better decision.

The document might reduce uncertainty, make an opportunity legible, create support for an internal approval, surface a risk, coordinate several specialists, or make a transaction possible. Its value depends on what happens because the document exists.

That distinction was central to Dalton Anderson's E081 conversation with Luke Tatman. Luke described commercial real estate firms that historically assembled data, built models, and presented the result. If gathering and assembling some of that information becomes cheaper, the client can see more clearly what the firm adds after collection.

Luke's conclusion was not that data disappears. Better data still matters, especially when small errors compound across a large lease or investment. His sharper point was that the advantage moves from simply possessing information toward using it to tell the right story and support the right decision.

Which information is getting cheaper

The broad phrase "information is commoditized" hides several different activities.

Research has found meaningful gains on bounded tasks.

An NBER field study followed 5,179 customer-support agents during the staggered introduction of a generative AI assistant. The authors reported a 14 percent average productivity increase, with larger gains among novice and lower-skilled workers. They offered suggestive evidence that the system helped diffuse the practices of stronger workers. Generative AI at Work

In a separate experiment, Shakked Noy and Whitney Zhang studied college-educated professionals completing incentivized writing tasks. Participants with access to ChatGPT completed the work faster, and the study found changes in output quality and inequality across participants. Experimental evidence on the productivity effects of generative AI

Neither study establishes that an entire profession has become interchangeable. They show that some production advantages can narrow when a tool makes practiced patterns easier to apply.

That is enough to challenge a service whose only distinction is that experienced staff can produce a standard artifact faster than inexperienced staff.

The frontier is jagged

The value shift is not a smooth transfer from professional to model.

A field experiment with consultants introduced the idea of a jagged technological frontier. Workers using AI performed better on tasks that fell within the tested system's capabilities. On a task outside that frontier, AI use could make performance worse. The work has since appeared in Organization Science. Navigating the Jagged Technological Frontier

That result matters because professional engagements are not one task. A workflow can contain easy retrieval, ambiguous classification, numerical modeling, local knowledge, negotiation, regulated judgment, confidential data, and an executive decision.

The model may be strong in one step and unreliable in the next. The human team may not know which step crossed the frontier until the output has already shaped the conclusion.

flowchart LR
    A["Access and retrieval"] --> B["Assembly and first-pass synthesis"]
    B --> C["Context and judgment"]
    C --> D["Integration and decision"]
    D --> E["Execution and accountability"]
    A -. "Often easier to substitute" .-> B
    C -. "Harder to substitute when consequence rises" .-> E

The valuable firm is not the one that declares every step automatable or protects every step as craft. It is the one that can locate the frontier for the current task, design the handoff, and make errors visible before they reach the client.

Six places the advantage can move

Proprietary context

Generic access is different from relevant access.

A public model may find broad market facts. It may not have the firm's licensed datasets, private transaction history, client constraints, negotiated terms, building-specific conditions, internal definitions, exception history, or the reason a decision was rejected last year.

Proprietary context is defensible only when it changes a decision. A large archive with weak provenance, conflicting definitions, and no retrieval path is storage, not a moat.

The useful test is whether another capable team could reconstruct the context in time to serve the client. If the answer is yes, ownership of the files provides less protection than the firm assumes.

Judgment

Judgment is the disciplined selection of what matters when evidence is incomplete, objectives conflict, and the cost of error is uneven.

It includes choosing an authoritative source, recognizing that a comparison is not comparable, rejecting a polished but implausible result, understanding which uncertainty needs escalation, and deciding when more research will not improve the decision.

Calling all of that "human judgment" can become an excuse for undocumented intuition. Strong judgment leaves an audit trail. It identifies assumptions, alternatives, evidence, uncertainty, and the person authorized to decide.

AI can support that work. It can propose scenarios, locate contradictions, test calculations, and expose missing fields. The professional advantage lies in making the decision process more reliable, not in keeping the process mysterious.

Integration

Many client problems are difficult because information has to move across systems, teams, incentives, and approval boundaries.

The answer may require finance, legal, insurance, operations, engineering, market research, and executive sponsorship. Each group can produce a correct local answer while the organization fails to act.

Integration includes translating definitions, resolving ownership, sequencing work, building the approval record, and turning analysis into a change someone can operate.

A commodity report can describe the problem. An integrated service carries the answer into the client's environment.

Accountability

Accountability becomes more visible as production gets cheaper.

The American Bar Association's Formal Opinion 512 provides one regulated-profession example. It discusses lawyers' duties involving competence, confidentiality, communication, supervision, candor, verification, and reasonable fees when using generative AI. It says lawyers need a reasonable understanding of a tool's capabilities and limitations and must apply an appropriate level of independent review. ABA Formal Opinion 512

The opinion does not govern every profession, and model rules are not identical to binding law in every jurisdiction. The example still makes the value shift concrete. Faster document creation does not transfer the lawyer's obligation to the tool.

Similar distinctions appear wherever a licensed or accountable person must stand behind a filing, valuation, recommendation, coverage decision, engineering conclusion, or transaction.

Accountability is not a decorative sign-off. The responsible person needs access to the evidence, enough competence to challenge the output, and authority to stop the workflow.

Service design

Clients experience the path to the answer, not only the answer.

They notice how quickly the team understands the problem, whether requests are repeated, whether assumptions stay visible, whether revisions create confusion, whether specialists contradict one another, and whether the final recommendation can survive internal review.

AI can reduce cycle time while making service worse. A firm can generate more drafts, create hidden checking work, multiply inconsistent files, or send clients polished output before the team agrees on the facts.

Strong service design uses the saved production time to improve clarification, review, scenario work, communication, and execution. It does not simply raise the number of deliverables.

Relationships

A professional relationship can be a real economic asset.

The advisor knows how the client makes decisions, which risks require escalation, how executives prefer to receive evidence, where prior projects failed, and who must be involved before an answer can move. That knowledge reduces search, translation, coordination, and execution costs.

The relationship is not an exemption from comparison.

Dalton challenged the idea that a long relationship survives any difference in price, speed, information, or service. Luke agreed that people continue to work with people they like, then preserved the economic boundary: the firm still has to offer the services the client needs at a price the client will accept.

A relationship remains defensible when it improves the result and lowers the total cost of achieving it. It becomes complacency when it merely delays a client's reevaluation.

Commercial real estate makes the shift visible

Commercial real estate is useful because a deal combines fragmented data, physical assets, local markets, financial models, documents, negotiations, and relationships.

NAIOP's 2025 report on AI in industrial real estate describes a move from fragmented and static information toward more connected data and decision workflows. It also discusses implementation barriers involving data, systems, talent, and organizational readiness. NAIOP industrial real estate report

The report concerns industrial real estate, not the entire commercial market. It supports a narrower point: faster analysis depends on the quality and connection of the underlying information.

Henry provides a current product example. The company describes software for commercial real estate research, underwriting, property and market analysis, and production of broker opinions of value and offering memorandums. Its Compass Commercial case describes faster BOV and OM workflows. Henry Compass Commercial case

Those are vendor and selected-customer claims. They do not prove a market-wide productivity result.

They do illustrate the pressure on a familiar document. If research, analysis, and initial production of an offering memorandum become faster, the firm's durable contribution has to be clearer. Source selection, assumptions, positioning, factual review, buyer understanding, distribution, negotiation, and deal execution do not disappear because the first draft arrives sooner.

Run the moat test

A claimed moat should survive four questions.

The test should be applied to one service line, not the firm in the abstract.

Begin with the deliverable the client sees. Trace the inputs, transformations, judgments, handoffs, approvals, and actions. Mark which steps a capable client or competitor can now perform more easily. Then identify which remaining steps change the outcome.

If the only defensible answer is "we know where to find the information," the service is exposed.

If the answer is "we hold private data," test whether the data is current, lawful to use, correctly defined, discoverable, and materially predictive.

If the answer is "clients trust us," identify what the trust allows the client to do better. Trust that does not improve coordination, candor, speed, risk control, or execution may be reputation without operating value.

Redesign the value proposition

The old statement often describes effort: the firm researches the market, assembles the information, applies experience, and creates a report.

The stronger statement describes a decision and the responsibility around it.

For example, the service may help a client choose a location using governed sources, expose the assumptions that drive the recommendation, coordinate legal and operating constraints, document the approval, and remain accountable through execution.

AI can reduce the cost of parts of that promise. The value proposition becomes stronger when the firm passes the efficiency to the client, reinvests it in a better decision, or both.

The weakest response is to preserve the old price and process while quietly reducing production effort. That approach invites the client to compare the artifact with a cheaper substitute and conclude that the relationship was hiding the economics.

What the client would still pay for

Imagine that the generic research and first draft are free.

What would the client still pay the firm to know, decide, integrate, verify, communicate, negotiate, and own?

The answer should name a consequence, not a tradition.

Information remains part of the service. It stops being the moat when access and assembly are the easiest part for someone else to reproduce.

The durable firm turns information into an accountable decision and carries that decision into the world.

AI assistance was used for research organization, drafting, and validation. Publication remains unauthorized.