A Clearer Position Is Not Proof: What Elevate Is Helping Point & Proof Test Next

A functioning product answers an important technical question.

Point & Proof can connect the steps involved in finding a development, reading the source, preparing a draft, checking it, and arranging publication.

But completing those steps does not automatically create customer value.

Several assumptions sit underneath the workflow:

• The intended customer considers LinkedIn worth some of their attention.
• Research and drafting are meaningful obstacles to publishing.
• Industry developments are useful starting points for their commentary.
• The draft reflects a position they genuinely endorse.
• Review takes less effort than writing the post themselves.
• The result is valuable enough to use again.

Each assumption could be wrong for a particular customer.

Someone may like the draft but see no reason to publish.

Someone else may value LinkedIn but prefer to write from personal experience rather than public sources.

Another person may appreciate the research but spend so long correcting the interpretation that the time saving disappears.

A product demonstration does not resolve those differences.

It is tempting to test the easiest thing to measure.

Did someone finish onboarding? Did the draft generate? Did they click approve?

Those events matter, but they do not necessarily answer the central question.

For Point & Proof, the critical assumption is:

“A suitable expert can reach a post they genuinely endorse with less total effort than their current approach.”

That is a stronger test than whether the AI produces fluent writing.

It includes the work before and after generation: finding the source, checking the interpretation, changing the language, correcting claims, and getting the post ready to publish.

It also includes a boundary.

A faster draft is not a success if the professional would not put their name on it.

The session introduced a practical structure:

“If a condition is met, a particular customer will take an observable action, resulting in a particular kind of value.”

Applied to Point & Proof, a proposed hypothesis could be:

“If an independent advisor receives a relevant source and a draft shaped by their own perspective and voice, they will prepare a post they are willing to publish with less total effort than their usual process.”

That statement can be investigated.

It identifies a customer, a condition, a behaviour, and an expected benefit.

It also makes room for an unwelcome answer.

The source may be irrelevant. The draft may misunderstand the advisor’s point. The review may take too long. The advisor may prefer their existing approach.

A useful hypothesis must allow the test to disagree with the founder.

One lesson from the session was to investigate how people currently work, rather than rely only on hypothetical answers.

The interview guidance emphasizes open questions, recent experiences, existing workarounds, and follow-up questions that uncover why someone behaves a certain way.

For Point & Proof, “Would you use an AI tool to help with LinkedIn?” is not the most useful starting point.

Better questions include:

• Tell me about the last post you intended to publish.
• What prompted the idea?
• How did you find the source?
• Where did the process slow down?
• What did you use to prepare the draft?
• What did you have to correct?
• Did you publish it?
• If not, what happened instead?

These questions leave room for a different problem to emerge.

Perhaps writing was easy, but the person doubted the commercial value.

Perhaps they had a strong idea but worried about confidentiality.

Perhaps they were already happy with a general-purpose AI assistant.

Those answers should influence the product, not be treated as objections to overcome.

A positive reaction to a demonstration is encouraging.

It is not the same as seeing someone complete the work.

The session’s observation guidance makes that distinction practical: let participants navigate the experience, avoid steering their behaviour, and ask clarifying questions rather than performing the task for them.

For Point & Proof, the review stage is particularly revealing.

Does the person understand where the source ends and their interpretation begins?

Can they find the claim they want to check?

Do they recognize their own point of view?

Are they refining a useful draft or replacing most of it?

Do they know what will happen when they approve it?

A founder can explain away confusion during a demonstration. A dependable product needs to make sense when the founder is not there.

One objection I have heard is that AI writing may require so much correction that it saves no time.

That concern deserves a test, not a reassurance.

Measuring only generation time would miss the objection.

A useful comparison needs to include:

• Finding and selecting material.
• Reading enough to understand it.
• Forming the point of view.
• Preparing the draft.
• Checking facts and interpretation.
• Editing the language.
• Preparing publication.

Point & Proof should be compared with the person’s actual alternative.

For one advisor, that may be manual writing. For another, it may be a general-purpose AI assistant. For someone else, the alternative may be not publishing at all.

Those are different comparisons.

If the alternative is silence, the first question is whether the workflow makes a worthwhile activity feasible—not whether it beats a writing process the person never completes.

The presentation included a useful warning:

“Never trust an average without its distribution.”

That matters for a product serving different kinds of experts.

An overall time-saving figure could conceal two very different experiences.

An advisor who already has a clear position may need only a useful source and a first draft.

Another may need help deciding what they believe before drafting begins.

An experienced AI user may move quickly through the workflow. A cautious first-time user may spend longer checking every sentence.

These are possibilities to investigate, not findings I can claim yet.

The lesson is to examine who benefits, under which conditions, and where the workflow fails.

A product that works well for a specific group may have a stronger business than one that produces an unremarkable average across everyone.

An approved post establishes something useful: the user was prepared to publish that particular piece of writing.

It does not establish that the product has become a dependable habit.

The next questions are different:

• Do they use the workflow again?
• Does the second session require less assistance?
• Do they continue to find the sources relevant?
• Do they still need extensive rewriting?
• Is the benefit strong enough to justify paying?
• If they stop, what replaces it?

A first successful session tests initial value.

Repeated use tests whether that value survives outside the demonstration.

Payment tests another part of the proposition.

None should be substituted for the others.

Point & Proof can improve the work of publishing.

It cannot establish that LinkedIn is the right channel for every professional.

A prospective customer may reasonably conclude that referrals, direct relationships, events, or another channel deserve more attention.

If that is the underlying issue, a better draft will not resolve it.

That separates two questions Point & Proof needs to respect:

“Is LinkedIn worth this person’s effort?”

And:

“If it is, does Point & Proof make the work meaningfully easier?”

The product should not require customers to accept the first proposition simply to test the second.

The session organized learning around three questions:

• What did the evidence show?
• Why does that matter?
• What decision follows?

For Point & Proof, that prevents research from becoming a collection of interesting comments.

If sources are consistently too broad, discovery needs attention.

If the source is useful but the perspective is wrong, the product may need better ways to capture the expert’s take.

If checking consumes the time saved in drafting, the review experience needs work.

If people approve posts but do not return, the next investigation should examine recurring value, competing priorities, and channel fit.

These are possible decisions, not changes justified by evidence already collected.

The point is to define what a finding would change before deciding that the research was successful.

The earlier positioning work helped clarify which problem Point & Proof should own.

This session added another discipline: distinguish a clear argument from a tested one.

The next step is not to gather agreement that the product sounds useful.

It is to identify the assumptions that matter most, investigate real workflows, observe where users struggle, and decide what the evidence warrants.

That may support the current direction.

It may narrow the customer further.

It may reveal that one part of the workflow matters more than the rest.

All three would be useful outcomes.

A clearer position gives Point & Proof something specific to test.

The learning begins when customers are allowed to contradict it.

If source-based LinkedIn commentary already has a place in your practice, the useful test is straightforward.

Bring a subject you know, contribute a perspective you genuinely hold, and assess the finished draft.

Does it represent your judgment?

Can you verify what matters?

Does preparing it take less total effort than your current approach?

Those are the questions Point & Proof needs to answer—not simply whether it can generate another post.