- Loop
- 6 stages
- Core panel
- 30-50 buyer questions
- First cycle
- 90 days
- For
- B2B and B2C companies
A transparent system from buyer question to measurable demand.
Six stages. Each one inspectable.
Measure, map, fix, expand, re-measure, attribute. The loop repeats on the same fixed panel, so every change can be compared with the run before it.
- 01Measure
Measure real buyer questions
We turn real sales questions, objections, comparisons and search demand into a fixed commercial question panel.
Mine sales calls, support conversations, CRM notes, objections and search data. Group them into 5-10 intent clusters and agree on a fixed core panel of 30-50 commercially meaningful questions. A smaller discovery set is kept separately, so the baseline does not move every time a new idea appears.
- Inputs
- Approved sales-call notes
- CRM objections
- Support themes
- Search Console
- Outputs
- Fixed core panel
- Separate discovery set
- Intent clusters
Question intakeSample- is automated sales tax filing accurate...
- Sales call, objection
- alternatives to spreadsheet bookkeeping...
- CRM, lost-deal note
- how to reconcile Shopify payouts...
- Support theme
- best accounting software for Shopify...
- Search Console
- 02Map
Build the observable Source Map
We record the brands, visible citations, recurring domains, descriptions and competitor gaps around each priority cluster.
For Google AI surfaces it can also record observable or strongly evidenced fan-out subtopics. For closed third-party systems, related subqueries stay hypotheses unless directly exposed. This is an observable source footprint, not a claim that we can see every hidden retrieval or training source.
- Inputs
- Core panel
- Agreed engines and locale
- Competitor set
- Outputs
- Source Map
- Recurring domains
- Owned, earned and proof gaps
Source ledgerSample- Review platform category page
- 9/12 / includes you
- Industry listicle, ecommerce accounting
- 7/12 / omits you
- Competitor A comparison page
- 5/12 / omits you
- Tax compliance publication guide
- 4/12 / omits you
- 03Fix
Fix the owned narrative
We remove technical blockers and build clear category, product, comparison, use-case, proof and answer pages.
Answer first, evidence next: direct answers, clear sections, explicit decision criteria, concrete proof and honest limitations, in language that matches real buyer intent. No padding written to satisfy a word count.
- Inputs
- Owned gap backlog
- Technical audit
- Expert interviews
- Outputs
- Priority pages live
- Entity contradictions removed
- Schema that matches visible content
Page briefSample- Page
- Multi-state sales tax for Shopify sellers
- Answer first
- Who it fits, in two sentences
- Evidence
- Filing accuracy method, customer quote, pricing
- Blocker fixed
- Plan table now server-rendered
- 04Expand
Expand external authority
We pursue legitimate contextual presence across publications, listicles, reviews, partners, podcasts, YouTube and authentic communities.
Context and topical relevance matter more than link count. Paid opportunities are optional, disclosed, approved in advance and carry the right link attributes. No PBNs, fake community accounts, mass spam or undisclosed pay-to-play.
- Inputs
- Earned gap list
- Source Map
- Proof assets
- Outputs
- Outreach log
- Earned placements pursued
- Review collection process
Outreach logSample- Industry listicle, ecommerce accounting
- Editor contacted
- Tax compliance publication guide
- Data offered for update
- Shopify sellers forum thread
- Founder answer, disclosed
- YouTube review, sales tax apps
- Demo account offered
- 05Re-measure
Re-measure the same panel
We repeat the fixed questions under documented conditions and report movement, volatility, unchanged gaps and raw evidence.
Same questions, engines, locale and competitors. Repeated samples reduce the impact of stochastic output. One answer or one personalized chat is never treated as a trend.
- Inputs
- Fixed core panel
- Documented run conditions
- Outputs
- Comparable run
- Measured delta
- Raw answers and citations
Run comparisonSample- Panel
- Same 40 questions
- Engines
- ChatGPT, Perplexity, Gemini
- Conditions
- en-US / logged out / 4 repeats
- Run 01 / Run 02
- Sep 2 / scheduled for day 90
- 06Attribute
Connect visibility to revenue signals
We combine answer tracking with analytics, AI referrals, branded search, self-reported attribution, CRM data and sales feedback.
Treated as contributory attribution, never as proof that one AI answer caused one sale. Adding a “How did you hear about us?” field catches AI-influenced demand that otherwise lands in direct traffic.
- Inputs
- GA4 and Search Console
- Self-reported attribution
- CRM opportunities
- Outputs
- Business-signal report
- Documented decision: scale, revise or stop
Attribution fieldSample- Form question
- How did you hear about us?
- Options
- Google / AI assistant / Review site / Colleague
- Routed to
- CRM source field
- Reported with
- AI referrals, branded search, sales
Owners and meeting cadence.
Who owns what
- Strategy leadThroughout the program
- The question panel, Source Map interpretation, priorities and the decision at the end of every cycle: scale, revise or stop.
- Execution teamScoped by plan
- Technical fixes, page briefs and expert-led content, source-gap outreach, reporting.
- Your subject-matter expert60-90 minutes per month
- Validates which questions matter commercially, supplies product evidence and expert input, approves claims.
- Measurement recordEvery run
- Stores the fixed prompts, raw answers, visible citations and run conditions, and preserves every comparison.
Cadence by plan
- Website Audit & Fixes
- SEO + AI Growth
- Regular monitoring in Rankry*, with a monthly report and one working call
- Market Leader
- Regular monitoring in Rankry*, with a monthly report and two working calls
How we measure AI visibility.
Three layers, reported together: what the engines say, what visitors do, and what the business sees.
Layer 01
Visibility
- Traditional non-brand and commercial rankings
- AI mention rate: share of tracked answer runs that name the brand at least once
- Recommendation rate: runs where the brand is presented as a viable choice, not mentioned in passing
- Top-3 presence: runs where the brand is among the first three explicitly ordered or clearly grouped options
- Average position: only on answers with an ordered or reliably inferable list; absence reported separately
- Citation coverage: runs with at least one visible citation to the brand's owned or earned footprint
- Competitive share of voice
- Description accuracy against an agreed factual rubric: category, capabilities, pricing, audience
- Positive positioning rate from a documented rubric, not an unexplained sentiment score
- Source overlap: domains that repeatedly shape answers
- Source-gap count: recurring relevant sources that include competitors and omit you
Layer 02
Engagement
- Organic and AI referral sessions
- Engaged sessions on commercial and evidence pages
- Branded-search trend
- Direct traffic trend, treated only as a weak proxy
- Returning visitors and assisted journeys
Layer 03
Business
- Demo requests and qualified conversions
- “How did you hear about us?” responses
- AI- and search-influenced opportunities
- Pipeline and closed revenue where CRM data is available
- Sales-cycle or education changes reported by the sales team
Measurement rules
- 01Derive the first question set from approved sales calls, support conversations, CRM notes, Search Console and stakeholder interviews. Redact personal information and get access approval before processing transcripts.
- 02Keep a stable core question set for before and after comparison, and a separate discovery set for new opportunities.
- 03Keep each product and market panel focused: normally 30-50 money questions, and no more than 100 fixed questions without a documented reason.
- 04Record engine or model, date, locale, language, account and session condition, search mode, raw output and visible sources.
- 05Run repeated samples. One answer or one personalized chat is never treated as a trend.
- 06Separate “mentioned” from “cited”.
- 07Separate “mentioned” from “recommended”.
- 08Distinguish visible citations from URLs merely surfaced in a retrieval interface, and from sources that are not observable at all.
- 09Separate branded questions from non-brand commercial discovery.
- 10Report uncertainty and model volatility openly.
- 11Never imply perfect attribution.
The scorecard format
Methodology sits next to every number. Later columns fill only from saved runs.
| Metric and definition | Baseline | Day 30 | Day 60 | Day 90 |
|---|---|---|---|---|
| Brand mentionedShare of tracked answer runs that name the brand at least once. | 7/40 | Not yet measured | Not yet measured | Not yet measured |
| Recommended in Top 3Runs where the brand appears among the first three explicitly ordered or clearly grouped options. | 3/40 | Not yet measured | Not yet measured | Not yet measured |
| Average position when presentCalculated only on ordered answers. Absence is reported separately, never dropped silently. | 3.8 | Not yet measured | Not yet measured | Not yet measured |
| Citation coverageRuns with at least one visible citation to the brand's owned or earned footprint. | 9/40 | Not yet measured | Not yet measured | Not yet measured |
| Share of voiceBrand mentions as a share of all tracked brand mentions in the competitor set. | 11% | Not yet measured | Not yet measured | Not yet measured |
| Accurate/positive positioningScored against an agreed factual rubric for category, capabilities, pricing and audience. | 72% | Not yet measured | Not yet measured | Not yet measured |
- Brand mentioned
- 7/40Baseline
- Share of tracked answer runs that name the brand at least once.
- Day 30, 60, 90: not yet measured
- Recommended in Top 3
- 3/40Baseline
- Runs where the brand appears among the first three explicitly ordered or clearly grouped options.
- Day 30, 60, 90: not yet measured
- Average position when present
- 3.8Baseline
- Calculated only on ordered answers. Absence is reported separately, never dropped silently.
- Day 30, 60, 90: not yet measured
- Citation coverage
- 9/40Baseline
- Runs with at least one visible citation to the brand's owned or earned footprint.
- Day 30, 60, 90: not yet measured
- Share of voice
- 11%Baseline
- Brand mentions as a share of all tracked brand mentions in the competitor set.
- Day 30, 60, 90: not yet measured
- Accurate/positive positioning
- 72%Baseline
- Scored against an agreed factual rubric for category, capabilities, pricing and audience.
- Day 30, 60, 90: not yet measured
See exactly what we measure.
Illustrative reporting format. Not client results.- Brand mentioned
- 7/40
- Share of tracked answer runs that name the brand at least once.
- BaselineDay 30 pendingDay 90 pending
- Recommended in Top 3
- 3/40
- Runs where the brand appears among the first three explicitly ordered or clearly grouped options.
- BaselineDay 30 pendingDay 90 pending
- Average position when present
- 3.8
- Calculated only on ordered answers. Absence is reported separately, never dropped silently.
- BaselineDay 30 pendingDay 90 pending
- Citation coverage
- 9/40
- Runs with at least one visible citation to the brand's owned or earned footprint.
- BaselineDay 30 pendingDay 90 pending
- Share of voice
- 11%
- Brand mentions as a share of all tracked brand mentions in the competitor set.
- BaselineDay 30 pendingDay 90 pending
- Accurate/positive positioning
- 72%
- Scored against an agreed factual rubric for category, capabilities, pricing and audience.
- BaselineDay 30 pendingDay 90 pending
Methodology sits beside every number: fixed question panel, documented engines, locale and session state, repeated runs. Later columns fill only from saved runs.
View the raw evidence behind one run+
- Engine
- Perplexity
- Mode
- Default search
- Locale
- en-US
- Session
- Logged out
- Date
- 2026-09-02
- Run
- Repeat 3 of 4
Question
best accounting software for Shopify sellers with multi-state sales tax
Answer excerpt
For Shopify sellers collecting sales tax in several states, the tools recommended most often are Competitor A, which files returns automatically, and Competitor B, which is cheaper for low order volumes. Competitor C suits teams that already use a separate bookkeeper. Your brand is also used by some sellers for payout reconciliation.
Visible citations
- Review platform category page
- Industry listicle: ecommerce accounting tools
- Competitor A comparison page
Coded
- Brand mentioned
- Yes
- Recommended
- No, mentioned in passing
- Position
- 4 of 4
- Brand cited
- No
- Description accurate
- Partly: sales tax filing omitted
The observable Source Map.
Stage two in practice. Select a buyer question to see the brands named, the sources cited and the gaps that become the next actions. The data is a sample.
Buyer questions
Named in answers
- Competitor Aavg #1.411/12
- Competitor Bavg #2.110/12
- Competitor Cavg #3.26/12
- Your brandavg #4.03/12
Recurring visible citations
- Review platform category pageReview9/12Includes you
- Industry listicle: ecommerce accounting toolsListicle7/12Omits you
- Competitor A comparison pageVendor5/12Omits you
- Tax compliance publication guidePublication4/12Omits you
OwnedFix or publish
Owned gaps
Category, product, comparison, alternative, use-case, pricing, proof and FAQ pages.
This questionNo page answers multi-state sales tax for Shopify sellers
EarnedPitch or earn
Earned gaps
Publications, listicles, reviews, partners, podcasts and video where competitors appear.
This questionAbsent from the listicle cited in 7 of 12 runs
ProofCollect or validate
Proof gaps
Customer evidence, real reviews, original data, examples and consistent product claims.
This questionNo public reviews mention sales tax filing
Table view of this map
| Question | Named in answers (of 12) | Recurring citations (of 12) | Gaps |
|---|---|---|---|
| best accounting software for Shopify sellers with multi-state sales tax | Competitor A 11; Competitor B 10; Competitor C 6; Your brand 3 | Review platform category page 9 (includes you); Industry listicle: ecommerce accounting tools 7 (omits you); Competitor A comparison page 5 (omits you); Tax compliance publication guide 4 (omits you) | Owned: No page answers multi-state sales tax for Shopify sellers; Earned: Absent from the listicle cited in 7 of 12 runs; Proof: No public reviews mention sales tax filing |
| how to reconcile Shopify payouts with bank deposits | Competitor B 8; Your brand 4; Competitor A 3 | Shopify sellers forum thread 8 (omits you); YouTube review: sales tax apps compared 5 (omits you); App marketplace listing 3 (includes you) | Owned: Help article exists but is not indexable; Earned: No founder or expert presence in the recurring forum thread; Proof: No walkthrough video of the payout reconciliation flow |
| alternatives to spreadsheet bookkeeping for a growing Shopify store | Competitor C 9; Your brand 7; Competitor A 7 | Industry listicle: ecommerce accounting tools 6 (omits you); Review platform category page 6 (includes you); App marketplace listing 4 (includes you) | Owned: No alternatives page for spreadsheet users; Earned: Missing from the same listicle as q1; Proof: Migration time from spreadsheets is not documented |
| is automated sales tax filing accurate for multi-state sellers | Competitor A 10; Competitor B 6; Your brand 2 | Tax compliance publication guide 9 (omits you); Competitor A comparison page 6 (omits you); Review platform category page 4 (includes you) | Owned: Filing accuracy is claimed but never explained; Earned: Not referenced by the compliance guide cited in 9 of 12 runs; Proof: No original data on filing error rates |
What changes in months one, two and three.
Month 1
Days 1-30
Baseline and first fixes
- Demand set from approved sales, support and search data
- Baseline across agreed engines, then re-measured on an agreed schedule
- Observable Source Map and technical blockers documented
- Agreed priority fixes started, not just planned
- First priority pages and materials published
What success means
A reproducible baseline, an agreed priority cluster, and the first hands-on changes already shipped in the paid month.
Month 2
Days 31-60
Implement and expand
- Crawlability, entity, schema and internal-linking fixes shipped
- Priority category, comparison and use-case pages live
- Useful content expanded across the priority cluster
- Work on selected external sources and reviews underway
What success means
The priority owned narrative is live, entity contradictions are removed, and relevant third-party sources are in progress.
Month 3
Days 61-90
Compare and decide
- Same-panel retest under documented conditions
- Analytics, AI referrals and form attribution reviewed
- Accumulated results compared with the baseline
- Decision: amplify what works, change it, or stop
What success means
A transparent measured delta and a defensible next decision. Positive movement is the target; a specific ranking or citation count is not promised.
What we commit to and what we do not claim
What we commit to
- The agreed implementation and outreach work.
- A fixed and inspectable measurement method.
- Raw answer and citation evidence where we can lawfully retain it.
- Honest reporting of wins, unchanged results, regressions and uncertainty.
- A documented decision at every cycle: scale, revise or stop.
What we do not claim
- Complete visibility into training data or hidden retrieval sources.
- Causal proof that one mention produced one model answer or one sale.
- Control over a model's wording, ordering or update schedule.
- A guaranteed Top-1 or Top-3 placement, or a fixed citation count.
See whether your brand makes the shortlist.
On a strategy call we walk through where your brand shows up across Google and AI answers, how you compare with a competitor, and the clearest gaps to close.