Positioning and protecting
brand stories in AI-answers.
Decode
AI knows your name.
It doesn't know what you mean.
You've spent years building a brand. The codes that signal prestige, heritage, trust and category authority. Harvard Business Review research found that LLMs reliably process the explicit signals: your name, your price, the word "premium." They consistently misread or flatten everything else.
Most AI content strategies are built blind. Before we produce a single piece of content for your brand, we need to know what AI currently says about it. mentioned. decode is the diagnostic that makes everything else. Ccontent production, monitoring, implementation work in the right direction instead of amplifying the wrong one.
What AI does to brand signals — illustrative
What your brand intends
Heritage & provenance
Four generations of craft. A specific place. A founding story that earns trust.
​
Scarcity as authority
Limited production signals mastery, not shortage.
​
​
Spatial & material cues
The texture, the silence, the weight of the object. Prestige you feel.
​
Artistic association
The collaborations, the references, the cultural positioning.
What AI
actually renders
Brand established in [year]
Known for quality products in its category.
Scarcity as authority
Not detected
​
Spatial & material cues
Not rendered
​
Artistic association
Partially detected. Misattributed
FLATTENED
INVISIBLE
DISTORTED
| Mentioned.
// 00.01° | Mentioned. decode
The gap no one
is measuring.
McKinsey finds that 90% of CMOs are experimenting with AI. Less than 10% have captured real value. Harvard Business Review explains why: the AI your customers use to make purchase decisions doesn't understand your brand the way your customers do. The codes that took decades to build are, right now, largely illegible to the models mediating an increasing share of consumer choice.
Most organisations respond by optimising for visibility, being found more often. Almost none are checking whether what AI says about them, once they are found, is actually accurate to their positioning. Hyper-personalised delivery of a misrepresented brand is noise at scale.
90%
of CMOs are experimenting with AI. Fewer than 10% have scaled it or captured value across marketing workflows.
McKINSEY, 2025
≈0%
are measuring brand description accuracy in AI. Organisations track mention frequency. Almost none track whether the description is correct.
​
Mentioned.decode research, 2026
HBR
LLMs consistently misread implicit brand codes. Scarcity, heritage, artistic association, spatial prestige cues — flattened or invisible.
HARVARD BUSINESS REVIEW, 2025
| Mentioned.
// 00.02° | Mentioned. decode
More visibility for
a misrepresented brand
is not a win.
Content production and AI monitoring are powerful tools. But they operate on whatever signal architecture already exists. If AI currently describes your brand incorrectly — wrong category, flattened heritage, stripped positioning — producing more content on that foundation does not correct the distortion. It scales it.
You end up with a brand that is highly visible in AI answers, consistently described as the wrong thing. The mention rate goes up. The brand equity goes down. Decode prevents that from happening by establishing what AI currently says before any content is produced.
Found
is not enough without described correctly
| Mentioned. monitor already tracks four signals: found, mentioned, described, chosen. Decode is what makes the described signal real — it gives you the baseline without which description accuracy cannot be measured.
Every
piece of content produced without a decode is a risk
If the signal architecture is distorted, content produced to increase visibility reinforces the distortion. Decode identifies the problem before the sprint begins.
​
​
One
audit. One time. Before anything else.
Decode is a diagnostic, not a subscription. Four weeks, five dimensions, 200+ prompts. The findings feed directly into every subsequent content and monitoring decision.
​​​
​
| Mentioned.
// 00.03° | Mentioned. decode
Five ways AI
misreads your brand.
We find all of them.
| Mentioned. decode tests your brand across five implicit signal dimensions. The codes that AI systems consistently flatten, distort or miss. Each dimension is scored, benchmarked against competitors and translated into a concrete remediation plan.
01
01
Implicit signal dimension 01
Heritage & provenance fidelity
Does AI accurately represent the depth, specificity and authority of your origin story? We test whether founding narratives, geographic provenance, craft lineage and generational expertise survive the translation into AI-generated answers, or collapse into a generic "established in [year], known for quality."
For brands where heritage is the primary differentiator, this is often the most damaging gap. We score it, benchmark it and fix it.
40+
PROMPTS TESTED
PER DIMENSION
5
AI PLATFORMS
BENCHMARKED
02
Implicit signal dimension 02
Scarcity & authority signals
Scarcity in premium brands does not mean shortage — it means mastery. Limited production, by-appointment-only, waiting lists: these signal authority, not unavailability. LLMs frequently either omit these signals entirely or reframe them as a customer inconvenience.
We test whether AI correctly reads the authority dimension of your scarcity signals and build the structured content required to make those signals legible. The difference between "hard to get" and "worth waiting for" is entirely in how AI frames it.
Scored
AGAINST YOUR OWN
POSITIONING BRIEF
Fixed
WITH STRUCTURED
ENTITY CONTENT
03
Implicit signal dimension 03
Category positioning accuracy
Is AI placing your brand in the right competitive set or has it categorised you alongside brands that share your price point but not your positioning? Category misassignment is one of the most common and most damaging forms of brand distortion in AI-generated answers.
We test how AI frames your competitive context, which peer brands it associates you with, and whether that framing supports or undermines your positioning. Being in the right category, correctly described, is the foundation of AI-driven purchase intent.
Full
COMPETITIVE SET
ANALYSIS
Cross
PLATFORM
COMPARISON
04
Implicit signal dimension 04
Tone & register fidelity
Your brand has a voice. A specific register — authoritative but not cold, precise but not clinical, confident but not loud. AI systems summarising your brand rarely preserve that register. The result is a flattened, generic description that sounds like a press release written by someone who has never met you.
We test whether AI-generated descriptions of your brand match your intended tone, and build the signal architecture needed to close the gap. Register is identity. If AI gets it wrong, first impressions are wrong.
Tone
SCORED AGAINST
BRAND GUIDELINES
Before
AND AFTER
MEASURED
05
Implicit signal dimension 01
Cultural & artistic association
The collaborations you've made. The references that anchor your brand in a cultural conversation. The designers, artists and movements you're associated with. These associations took years to build and define your brand for a buyer who already knows you. For a new buyer asking an AI system, they may simply not exist.
We audit whether AI accurately renders your cultural positioning, test for misattribution and omission, and build the entity-signal architecture that makes these associations discoverable and creditable. Cultural positioning is either legible to AI or it doesn't exist for the buyers AI is mediating.
Deep
CULTURAL SIGNAL
AUDIT INCLUDED
Entity
SIGNAL PLAN
DELIVERED
| Mentioned.
// 00.04° | Mentioned. decode
Four weeks.
Total clarity.
WEEK 01 // INTAKE
Brand brief
& signal mapping
Does your brand appear at all when buyers ask relevant questions? Tracked per prompt, per platform, per market. The baseline signal — without it, everything else is irrelevant.
TRACKED DAILY · ALL PLATFORM
​
​
WEEK 02 // TESTING
200+ prompts across five platforms
How often are you named versus competitors in your category? Mention rate in AI answers is the new share of voice — and it shifts faster than any metric you currently track.
​
​
TRACKED DAILY · COMPETITOR BENCHMARK
​
​
WEEK 03 // SCORING
Brand
narrative
When AI names you, does it describe you correctly? Heritage, positioning, category, tone — or a flattened, generic version that contradicts everything you've built? This is the signal almost no one is measuring.
​​
ORBIT + APEX · BRAND STORY PROTECTION
​
​
WEEK 04 // REMEDIATION
RECOMMENDATION
RATE
When a buyer asks AI who to contact, who to trust, who is best — does your name come first? The recommendation rate is the commercial signal that connects AI visibility directly to pipeline.
TRACKED DAILY · ALL PLATFORMS
​
​
| Mentioned.
// 00.05° | Mentioned. decode
The brands with the
most to lose if AI
gets them wrong.
// Premium & Luxury
Brands where implicit codes are the entire value
When your category premium depends on signals that took decades to build, a flattened AI description doesn't just reduce visibility. It actively undermines the purchase rationale. Decode was designed for this problem first.
// Heritage & Craft
Brands whose story is their most important differentiator
If your brand's authority comes from where it's from, how it's made and how long it's been made that way, you are especially exposed to AI distortion. These are exactly the signals LLMs handle worst.
// Challenger & Positioned
Brands fighting for the right category placement
Category misassignment by AI is particularly damaging for brands moving upmarket or redefining their space. If AI puts you next to the wrong competitors, you lose the positioning battle before it starts.
// Existing monitoring clients
Brands already on
Ignite, Orbit or ApexIf your monitoring shows increasing mention rates but you have never audited brand description accuracy, you may be tracking the wrong signal. Decode gives your monitoring the baseline it has been missing and tells you whether the content being produced is building on correct foundations or reinforcing a distorted one.
| Mentioned.
// 00.06° | Mentioned. decode
One audit.
No subscription.
Everything else follows.
Decode is a standalone diagnostic. You commission it once, before any content is produced. The findings feed into your implementation sprint, inform your monitoring baseline, and tell every subsequent content decision which direction to build in. There is no bundle. No lock-in. Just the clearest picture available of what AI currently says about your brand and what to do about it.
mentioned. decode — standalone brand signal audit
Everything included. Four weeks. One report.
Brand signal brief and intake session. 200+ prompts tested across all seven AI platforms. Five implicit signal dimensions scored and benchmarked against competitors. Prioritised remediation plan with content brief, structured so it feeds directly into an implementation sprint or an ongoing content strategy. 30-minute senior debrief session included. No retainer, no follow-on commitment.
​
​
One-time · no commitment
€ 2.950
EX. VAT · FULL REPORT IN 4 WEEKS
// After decode
| Mentioned.
The decode report feeds
directly into three things.
OPTION 01 // FROM €2,950
Implementation Sprint
The remediation plan from your decode becomes the brief for an implementation sprint — content, schema and entity signals built to correct the distortions identified. View sprint →
​
​
​
OPTION 02 // FROM €950/MO
LLM Intelligence Monitor
The decode establishes the baseline for the DESCRIBED signal in your monitoring dashboard. Without it, description accuracy cannot be meaningfully tracked. With it, every monthly report has a reference point. View plans →
​
​​​​
​
OPTION 03 // COMBINED
Sprint + monitor
Most clients run the sprint immediately after decode to fix the identified gaps, then move to monitoring to protect the corrected signal architecture and track drift over time. This is the full sequence. Talk to us →
​
​
​
​