One account, eight scans, eleven modules. Watch each screen do its job: how often AI names you, who it names instead, where its answers come from, and what to change first.
The real navigation · 11 modules · one account, 8 scans · 3 Aug to 21 Sep
01 · How AI sees youAI Visibility
How often does AI actually name you?
Every engine’s own number, side by side, plus the visits and revenue AI already sends you.
Scan 8 · named in 31% of answers, up from 17% at the first scan
app.findrix.ai/sites/shifta/citations live screen
Real visits from AI · 1,342, 7.5% of all trafficChatGPT · named in 35.4% of answersPerplexity · 41.5%, up 2.4 pointsTrend · 8 scans, 3 Aug to 21 Sep
What the clip doesSwitching engine tabs: the mention rate changes from Claude to ChatGPT to Perplexity, then the trend by scan.
Brand mentions, URL citations, share of voice and a preference score, one engine at a time. Never averaged.
AI visits and revenue from AI, attributed by your Google Analytics, right above the mention rate.
A trend by scan, every earlier run in the Runs switcher, and an Export button for what you see.
Ghost citations · Growth and Authority
03 · How AI sees youCompetitors
Who does AI recommend instead of you?
The five brands AI names in your market, their share on every engine, and a head-to-head with any rival.
Scan 8 · 3rd of 5 brands with 17.5% of mentions. Rosterlane leads with 36.6%
app.findrix.ai/sites/shifta/competitors live screen
Shifta · 17.5%, 3rd of 5 across 7 enginesOn Perplexity · 22.4%, still 3rdOn ChatGPT · the same five, a different orderYou vs Shiftanchor · engine by engine
What the clip doesThe share-of-voice race on all engines, then on Perplexity and ChatGPT, then the duel switches to Shiftanchor.
The share-of-voice race across all 7 engines, then per engine. The order changes when you switch.
Pick a rival, see the duel: you against them on every engine, with “They’re ahead” or “Probable gap” spelled out.
Where you collide: the 30 prompts where both of you are named, and who each engine picks. Up to 20 brands tracked.
04 · How AI sees youAI Source Map
Where does AI get its answers?
The 78 sites behind your market’s AI answers, ranked by influence, with the ones you are missing from.
Last 8 runs · 78 sources, shifta.com ranks 5th. 31 of them you can get on
app.findrix.ai/sites/shifta/sources live screen
78 sources behind your market’s AI answers17 communities where buyers ask81 like this one
What the clip doesThe influence map, then the Type filter narrows the list to the 17 communities, then reddit.com opened to the 81 buyer questions it answers.
Six source types with their share of influence: owned content 35%, community 22%, editorial 18%, review platforms 13%, directories 9%, reference 3%.
Filters “You appear” and “You’re absent”, and a “Can you get on?” verdict per source: 31 you can get on, 34 worth pitching, 8 not a target.
Open any source for the buyer questions it answers, which engines cite it, and which rivals already sit there.
07 · Get mentioned moreFix Your Website
What do I change on my site first?
Every fix for your site, ordered by what moves AI answers, with the text written for you.
5 of 16 fixes done · answers naming Shifta at 31.0%, up 2.1 points since the previous check
app.findrix.ai/sites/shifta/fix live screen
5 of 16 doneUp 2.1 pointsThe fix, in plain words
What the clip doesThe fix plan, then the Priority #1 card opens: what the scan found and what to do, written in plain words.
A fix plan with priorities, from crawlability and search presence to page-by-page cards: /pricing, /product, /compare.
Open a card and the fix is written: what we found, what to do, the llms.txt or schema to publish. Copy it, then “How to apply”.
“What fixes earn you”: the share of answers where AI names you. The next check is dated on the screen, so you know when a fix counts.
08 · Get mentioned moreGet Featured
Which sources can I actually get onto?
The places AI reads in your market, sorted into what can be won and what can’t, each with the next step drafted.
64% of what AI reads here can be won · 2 placements done, in the Work Log
app.findrix.ai/sites/shifta/sources-win live screen
64% of what AI reads here can be wonCited 309× · 81 of 92 questionsShoot these firstInto your Work Log
What the clip doesThe worklist, then the YouTube card opens: 309 citations, 81 of 92 questions, the five questions competitors win there and how to shoot a video that gets cited.
“Can be won”: communities, news, review platforms and directories you can reach. “Can’t pitch”: competitors’ sites and reference giants.
Every source with its weight: YouTube is cited 309 times and feeds 81 of your 92 tracked questions.
For YouTube and Reddit, the questions competitors win there and a short brief on what to publish. For publishers, the contact to write to and the letter drafted. “Mark one done” moves it to your Work Log.
10 · Get mentioned moreFix Wrong Info
What does AI get wrong about you, and where?
The sources feeding AI a false fact about you, and the correction letter, already written.
2 sources still state a false fact · 1 corrected since the start · 22 facts tracked
app.findrix.ai/sites/shifta/fact-checks live screen
2 sources state a false fact · 22 facts trackedDrafted from your facts · nothing inventedCopy, send from your inbox · mark as sent
What the clip doesThe fact checker, then “Open the letter” on softwarepicks.io reveals the drafted correction letter, with the author and inbox already found.
Every claim is graded against your 22 confirmed facts, and the ones that make you look worse are flagged.
The letter is drafted from your facts, nothing invented. Edit it, copy it, send from your own inbox, then mark as sent.
“Find contact” finds the author and an inbox. Each source walks Letter ready → Sent → Corrected.
Source checks · Growth and Authority
Five more · at a glance
The rest of Findrix, one panel each.
What AI believes about you, the report you send, the questions we ask, your own pages, and the ledger of everything done.
02 · How AI sees youBrand Image in AI
What does AI believe about you?
app.findrix.ai/sites/shifta/perception
How AI defines you, word for word, and where you rank on each topic: leader, challenger, or wrong.
Red means AI has a fact wrong: open the topic and get its sources corrected. Re-scan up to 4 times a month.
05 · How AI sees youReports
What do I show my client or my boss?
app.findrix.ai/sites/shifta/reports
A report after every scan and a monthly one: what actually moved, things to do, things done, AI traffic.
Share with client, download the PDF, or open the print view. Written in plain words, no dashboard to explain.
06 · How AI sees youYour Prompts
Which questions do you ask on my behalf?
app.findrix.ai/sites/shifta/prompts
92 live prompts on Growth. 82 never name your brand, because that is how buyers ask before they know you.
Prompt research: a run suggests the topics buyers ask about in your market, you pick one or type your own, and Findrix drafts the prompts for it. Review them, tag them, add them to the set.
Edit, tag, import or add your own. Changes take effect on the next run. 60, 100 or 180 prompts by plan.
09 · Get mentioned moreOwned sources
Does my own content earn citations?
app.findrix.ai/sites/shifta/owned-sources
Your pages doing the work, the answers quoting them, and what gets quoted: product 30%, blog 17%, guides 12%.
Content gaps: 45 topics where AI answers without you, heaviest first. Write one, log it, watch it leave the list.
11 · Get mentioned moreWork Log
What has been done, and did it work?
app.findrix.ai/sites/shifta/actions/history
Every website fix, content piece, fact correction and placement in one ledger, with a verified mark when the next scan confirms it.
Log work in one click, filter by type, export. The report reads “what was done” straight from this log.
IN YOUR INBOX
Four emails that carry value, never noise.
Monday, after the scanYour report is readyWhat moved on each engine, the fixes the scan confirmed, the new cards for the week. One link to the full report.
Thursday, only when something didWhat worked this weekThe fix or placement that landed, the engine it moved, the number before and after. A quiet week sends nothing.
The moment it happensyoursite.com needs your attentionYour site started blocking an AI bot, or went down for the scan. Never more than one a day.
Monday, for agenciesYour sites this weekOne digest across every client site: who moved, who needs a hand, and the report link for each.
Nothing happened, nothing sent. Every email is one thing that changed, and what to do about it.
PLANS
Three plans. Here is exactly what changes.
Presence measures three engines. Growth and Authority add two more, the source fact-checks and Ghost citations. Claude and Grok are add-ons on any plan.
Presence
$49per site / mo
3 engines · 60 prompts
See where you stand.
Full dashboard: visibility, brand image, competitors, sources
Know how often AI recommends you, then move it: exact fixes for your site, the sources AI trusts that you're not in yet, and every fact it gets wrong about you. Across ChatGPT, Gemini, Perplexity and 4 more engines.
Deutsch
Français
Español
Português (Brasil)
Italiano
