Methodology

How Heft® measures what AI actually does with your brand

Traditional trackers stop at “were we mentioned?” Heft® goes deeper: we decode the motive structure inside every answer, score story intensity, and tell you exactly why you're winning — or losing — each buyer moment.

Layer 1

AI doesn't answer — it dramatizes

Every AI response constructs a small story: what's happening, who's asking, what action they're taking, what tools solve it, and why it matters. Heft® maps this motive structure and shows where your brand sits inside the explanation — not just whether your name appeared.

This is the measurement layer rank trackers miss. Two brands can both be “mentioned.” Only one is cast as the recommended guide.

Motive structure · five elements

Situation

What world is the shopper in?

Shopper

Who is asking?

Action

What decision are they making?

Means

What product or path solves it?

Purpose

Why should they care?

Dominant driver: which element leads the explanation — and whether your brand is connected to it.

Layer 2

Four signals on every observation

Heft® scores every AI answer against your brand with four independent layers. Mention without recommendation is a diagnosis, not a win.

01

Mention

Did AI say our brand at all?

02

Cite

Did AI point to our URL as evidence?

03

Recommend

Did AI tell the user to choose us?

04

Role

Hero, supporting option, commodity slot, or missing?

The citation trap

High cite rate with low recommend rate means AI uses you as a source — but recommends someone else. Heft® surfaces this pattern so brand teams can fix the story, not just the sitemap.

Layer 3

Three measurement programs, one readout

Different questions need different probes. Heft® runs structured programs and rolls them into a single strategist view: position → diagnosis → prescription.

P1

Baseline visibility

If someone asks AI once, do we show up? Seven prompt shapes × multiple engines. Mention, cite, recommend, role — per observation.

Isolated check
P2

Conversation survival

Does our recommendation survive follow-ups and pushback? Multi-turn threads test whether your story holds when the shopper digs deeper.

Thread stress-test
P3

Gap harvest

What sub-questions does the market still need answered? Surfaces FAQ branches, comparison angles, and content gaps the engine expects but can't find on your site.

Content intelligence

Layer 4

Diagnosis: why visibility breaks

Heft® doesn't stop at scores. Every gap is typed — so brand and content teams know exactly what to fix.

Situation story weak

AI doesn't frame the buyer's context the way you chartered it.

Motive misfit

Your brand appears but isn't tied to the dominant driver in the answer.

Proof gap

Competitors have editorial or third-party evidence you lack.

Story intensity low

Messaging is generic — clarity, intensity, or emotion scores fail the vividness bar.

Fan-out unmet

Follow-up questions branch to topics your site doesn't cover.

Accuracy risk

AI describes your brand incorrectly — wrong category, invented claims, competitor confusion.

The flywheel

Measure → sharpen → prove

Heft® closes the loop. Charter a buyer moment, run locked probes, read the dramatization, fix the story, re-measure. Role movement and motive fit delta — not vanity traffic.

AgreeHypothesizeMeasureAuditActProve
How Heft® Works — AI Dramatization & Story Measurement