When the Proxy for Value Becomes More Important Than the Value

Thought leadership has become remarkably measurable.

We can count followers, engagement, impressions, rankings, mentions, citations, publishing frequency, audience growth, and AI visibility.

Each number can tell us something useful.

But there is a question we should probably ask more often:

What is this metric a proxy for?

Followers may be a proxy for influence.

Engagement may be a proxy for resonance. Search rankings may be a proxy for authority.

AI citations may be a proxy for legibility. Publishing volume may suggest intellectual substance.

The problem begins when producing the evidence of value becomes easier than producing the value it was supposed to indicate.

At that point, we can become very good at optimizing the signal while quietly neglecting its source.

Why Proxies Work Until They Don’t

A proxy is simply something observable that helps us infer something harder to observe directly.

We cannot easily measure “authority.”

But if someone is repeatedly quoted, read, recommended, searched for, cited, and invited to contribute, we reasonably infer that some authority exists underneath those outcomes.

Historically, many signals worked because they were relatively expensive consequences of the underlying thing.

Writing a substantial book usually required enough knowledge to sustain one.

Publishing extensively required considerable time and thought.

Being repeatedly quoted suggested that other people considered your perspective useful.

Building a large audience generally required many people to choose to continue paying attention.

The proxy wasn’t the value.

But producing the proxy often required enough of the underlying value that the two remained meaningfully connected.

Technology is changing that relationship.

When Authority Signals Become Cheap to Produce

AI can help produce a book.

Content systems can generate extraordinary publishing volume.

Professional-looking frameworks can be created in minutes.

Engagement strategies can improve the visible signs of popularity.

AEO techniques can improve performance against tracked prompts.

Beautiful visual systems can make ordinary ideas look remarkably substantial.

(And yes, I say that as someone who loves a gorgeous infographic.)

None of these things is inherently a problem.

The interesting question is what happens when the cost of producing a signal falls faster than the cost of developing what the signal traditionally represented.

A polished framework once provided some evidence that a person had done the thinking required to create it.

Increasingly, it provides evidence that someone can produce a polished framework.

Those aren’t necessarily the same thing.

The signal has become easier to decouple from its source.

The Scarce Thing Moves Upstream

This is one reason I don’t think AI makes substantive expertise less valuable.

I think it does something closer to the opposite.

AI makes fluency cheaper.

It makes polish cheaper.

It makes content cheaper.

And it makes frameworks cheaper.

It makes the appearance of intellectual production considerably easier to create.

So the scarce thing moves upstream.

Toward judgment.

Experience.

Pattern recognition.

Original distinctions.

Intellectual coherence.

A developed way of seeing.

I explore this in Your Intellectual Signature, particularly the idea that recognizable authority increasingly comes from the pattern of thinking visible across a body of work, rather than one polished idea or signature framework.

When almost anyone can reproduce the surface conventions of thought leadership, the more interesting question becomes:

What does this person characteristically see that others don’t?

That is much harder to manufacture.

Style Should Make Depth Legible, Not Substitute for It

This doesn’t mean we should abandon the signals.

Quite the opposite.

Good branding can make excellent work easier to recognize.

A thoughtful infographic can make a complicated distinction easier to understand.

Publishing can externalize expertise that would otherwise remain invisible.

Search optimization can help valuable ideas become findable.

AI optimization can make a substantive knowledge estate easier for retrieval systems to understand and use.

Visibility can expose authority to more people.

The hierarchy matters.

Style should make depth legible. It should not substitute for depth.

Visibility should expose authority. It should not substitute for authority.

Content should externalize expertise. It should not substitute for expertise.

AI citations should provide evidence of retrieval. They should not substitute for the knowledge estate that made retrieval possible.

The problem isn’t optimizing the proxy.

It’s allowing optimization of the proxy to become the work.

The Metrics We Can See Are Often Downstream

This connects to another model I’ve been using to think about AI-mediated discovery:

Knowledge → Externalization → Legibility → Retrieval → Visibility → Citation → Traffic

Notice where most of our dashboards live.

Toward the right.

That’s understandable because those outcomes are easier to observe.

We can count impressions. We can count citations. We can measure traffic. We can track rankings.

But the durable value lives disproportionately toward the left.

Knowledge.

Judgment.

Ideas.

Distinctive language.

A coherent body of work.

The intellectual infrastructure from which those downstream signals emerge.

This doesn’t make downstream metrics unimportant. They provide valuable feedback about whether the underlying estate is successfully interacting with markets, search systems, AI intermediaries, and people.

But evidence about the asset is not necessarily the asset.

That distinction becomes increasingly important when evidence can be manufactured more cheaply.

Ask One More Question of Your Metrics

Suppose you have 10,000 followers.

Wonderful.

A proxy for what?

You received 500 AI citations.

Interesting.

A proxy for what?

Your posts generated 100,000 impressions.

Great.

A proxy for what?

You published 150 articles this year.

Okay.

A proxy for what?

These aren’t cynical questions. They are clarifying ones.

Perhaps those 10,000 followers represent a deeply engaged community that trusts your judgment.

Perhaps those AI citations demonstrate that a substantial knowledge estate has become highly legible across a growing range of questions.

And perhaps those impressions are steadily introducing distinctive ideas to the people most likely to need them.

If so, the proxy is giving you useful evidence about something you genuinely value.

But if you cannot explain what the metric is evidence of, you may have started optimizing a scoreboard whose relationship to the actual objective has become unclear.

When Appearance Stops Being Evidence

There is a deeper consequence to all of this.

As familiar signals of authority become easier to reproduce, people and systems need other ways to evaluate what sits underneath them.

One of those may be lineage.

Can I trace this idea through a body of work?

Does this person keep developing it?

Are there related distinctions, questions, frameworks, examples, and applications?

Does the idea connect coherently with other things they have said?

Is there evidence of judgment accumulating over time?

In other words, when appearance becomes less reliable as evidence, provenance becomes more valuable.

A framework-shaped graphic can be created quickly.

A coherent intellectual lineage is harder to fake.

It leaves a trail.

Don’t Optimize the Evidence at the Expense of the Source

I’ve never really been about maximizing visibility for its own sake. The larger question, for me, has always been around how expertise becomes known.

Metrics help us see pieces of that process.

Followers can tell us something.

Rankings can tell us something.

AI citations can tell us something.

Traffic can tell us something.

But none deserves to become more important than the underlying qualities we hoped it would reveal.

So use the proxies.

Measure them. Learn from them. Optimize them when doing so helps substantive work travel farther.

Just keep asking what sits underneath.

Because as the visible signals of authority become cheaper to manufacture, the value of genuine intellectual substance doesn’t disappear.

It becomes easier to see why we needed the substance in the first place.