Every week: we test 30 prompts across ChatGPT, Perplexity, Gemini, and Google AI, find what's blocking your citations, produce 8 on-site and 8 off-site content assets, and measure your share of AI answers again.
No invented visibility score. No ranking guarantees. Just a weekly loop that builds compounding evidence for AI systems to trust and cite your brand.
Illustrative — how the weekly loop works
Buyer prompt
"Best [category] vendors in [market]?"
Your brand: absent
Competitor A: cited
Gap identified
No third-party corroboration for [category] + [market]
Type: authority gap
Week's output
+ 1 on-site article
+ 1 off-site asset
Measured again
Appearances
6 → 11
feeds next week
What the loop covers
AI platforms tested
30
prompts tested weekly
5
AI platforms covered
8+8
assets produced monthly
weekly
re-measurement cadence
On-site assets are articles and page updates your team publishes. Off-site assets are placements, citations, and structured data contributions that build corroboration. Both feed the next week's prompt measurements.
Ranking is not the same as being cited. Most AI visibility problems fall into one of three failure modes — each requiring a different fix.
Buyers ask AI which vendors handle [category] in [market]. Your brand isn't in the answer. A competitor with weaker SEO is cited because AI systems have more corroboration for them.
Prompt returns: no mention of your brand
The AI knows your name but states your service wrong — wrong geography, wrong category, outdated positioning, or conflated with another brand. Every mention works against you.
Prompt returns: your brand with wrong attributes
Your brand shows up occasionally but loses ground week over week. No third-party sources confirm what you do. Ranking pages don't make it into AI answers. The citation rate doesn't compound.
Prompt returns: your brand, inconsistently
Week-over-week measurement
The loop is only useful if it's measurable. Each week's report shows citation changes against the frozen baseline, not generic impressions.
Weekly Report — Week 6
Measured vs. frozen baseline
Appearances
11/30
was 6/30
+83%
Citations
2
was 0
new
Share of voice
14%
was 7%
+7pp
vs baseline
↑
positive trend
directional
Trend data is directional — AI citations are not deterministic and vary by query, session, and model version. Illustrative numbers only.
The weekly loop
Each cycle compounds on the last. The baseline is frozen at week one so improvement is always measured against where you started.
Baseline the gap
We run 30 prompts across 5 AI platforms, map your citation rate against competitors, and identify access, understanding, and authority blockers.
Prompt matrix + gap table
Isolate the blockers
Each gap is typed — missing corroboration, poor extractability, access failure — and ranked by estimated citation impact.
Prioritised gap report
Create the evidence
We produce 8 on-site content assets and 8 off-site placements targeted at the gaps found. You publish; we produce.
8 articles + 8 off-site assets
Measure what moved
We re-run the same 30 prompts and compare citations, share of voice, and appearance rate against the frozen baseline.
Weekly delta report
Feed the next cycle
New gaps found this week become the next week's targets. The baseline stays frozen so improvement is always measured against week one.
Feeds back to stage 01
Scope per engagement: one brand, one market, one language, one category.
What we produce
AI systems cite brands that have both strong on-site clarity and off-site corroboration. The monthly output targets both.
On-site
Client handoff
We produce the content and page changes. You review and publish. We don't push to your CMS without your approval.
Off-site
Why off-site matters
Review sites, comparisons, and forum mentions often drive more AI citations than the best page on your own domain.
Why not just buy monitoring software?
Monitoring tools stop at measurement. Agencies produce without measuring. Uygen does both, weekly, at a transparent price.
| Feature | Monitoring tool | Uygen | Agency |
|---|---|---|---|
| Content production included | partial | ||
| Transparent monthly price | |||
| Weekly measurement cadence | |||
| AI-specific prompt testing | |||
| Off-site asset production | partial | ||
| Stops at "recommend to fix" | |||
| Produces + measures again |
Deliverables
The loop produces concrete artifacts, not slide decks. Every deliverable connects to a measurable outcome.
30 prompts run across 5 AI platforms. Every query, every result, every citation gap — documented.
Which competitors appear where you don't, and what sources they have that you're missing.
Extractability and entity clarity for the pages most likely to drive citation improvement.
AI crawler access, Bing indexing, robots directives, schema, and rendering — all in one pass.
8 on-site articles and 8 off-site assets per month, production-ready for your team to publish.
Citation rate, share of voice, and appearance count compared against the frozen week-one baseline.
Responsibility boundary
Uygen doesn't push to your CMS, post on your behalf, or publish without your review. The split is clear from day one.
Uygen does
You do
Time commitment from your side: under 30 minutes per week to review the report and approve drafts.
Scope and fit
Tight scope is what makes weekly measurement and compounding evidence possible.
The loop runs for a single brand in a single geographic market, in a single language. No multi-brand bundles. Scope clarity is what makes weekly measurement meaningful.
You have SEO spend, a content team, or an agency. You're not starting from zero. The loop amplifies what's working and fixes what AI systems can't yet use.
You can see the gap in ChatGPT, Perplexity, or Gemini. It's not a hypothetical problem. The loop makes the gap measurable and gives you a weekly path to close it.
Not the right fit if:
Pricing and terms
Three months is enough time to baseline, produce, and show measurable movement. After that, continue month to month.
AI Visibility Loop
Week one is the audit baseline. Every subsequent week closes the loop: find gaps, produce, measure again.
Terms at a glance
Output metrics vs outcomes
We commit to output metrics: prompts tested, assets produced, reports delivered. Outcome metrics — citation rates, share of voice — are observed and reported, not guaranteed. AI citations are not deterministic.
What We Audit
We do not treat AI visibility as one vague score. We isolate whether the problem is access, understanding, authority, or a mix of all three.
Can AI systems reach, read, and index the pages that should represent your brand?
Can models extract the core facts about who you are, what you do, and who should choose you?
Do enough independent sources confirm your brand for AI systems to trust and cite you?
Prompts are fixed per engagement and run identically each week. The baseline is frozen at week one so trend data is always directional against the same starting point.
Full methodology →FAQ
From the blog
Two real-world signals show why ranking in Google does not guarantee accurate mentions or citations in AI answers.
Read →AI Overviews rarely send a click. See the current zero-click data, which content types lose traffic, and how to check your Search Console for the gap.
Read →Diagnose why ranking content is missing from Google AI Overviews using Access, Understanding, and Authority checks.
Read →AI engines parse your DOM and schema, not your rendered page. A real case where corrupted JSON-LD cost a brand an AI citation, and how to check your own site.
Read →Citation share counts how often AI engines attach your URL as a source — not just name your brand. Now trackable for free in Bing Webmaster Tools.
Read →A step-by-step method to measure competitor citation share and diagnose why rivals appear in AI answers.
Read →A 30-minute manual method to see if Perplexity mentions and cites your brand, read the Sources panel, and diagnose the gap.
Read →5 steps to check if AI systems can find, parse, and cite your site — from Bing indexing to a live citation spot-check.
Read →Ranking is not the same as being cited. Five on-page tactics that move a ranking page into Google's AI Overview.
Read →Strong SEO doesn't guarantee AI citations. Here's what an AI visibility audit checks and whether your brand needs one.
Read →The honest guide to what llms.txt does, who reads it, and whether your site needs one.
Read →The metric that makes AI citation visibility measurable — definition, formula, and tracking guide.
Read →A manual playbook for measuring AI citation rates and diagnosing visibility gaps before buying software.
Read →A practical prompt-set and citation check for ChatGPT brand visibility.
Read →A diagnostic guide for brands missing or misrepresented in Gemini answers.
Read →A diagnostic guide for brands missing from Perplexity answers and citations.
Read →A diagnostic guide for brands missing from Google AI Overviews.
Read →Ready to start
Book a call and we'll walk through the loop, scope the engagement, and show you what week one delivers. $700/mo. 3-month start.