"Third on Google" sounds fine — until you find out you're third outside your own office and invisible two streets away. Rankfinity scans Google Maps from 49 points around a business, tracks organic positions by coordinates, and measures whether ChatGPT recommends you. One tool instead of three subscriptions — one login, one bill, and the credits already included.
Local Falcon draws the map. Nightwatch or AccuRanker track the organic positions. Peec or Profound measure AI. Three logins, three invoices, the same keyword list typed three times — and then you stitch it all into one client report by hand.
Every one of them meters you on points or keyword slots. That pricing quietly trains you to check less often: you postpone the scan after a profile edit, you skip the "did the competitor move?" check, because you're not sure it's worth the spend.
And most of them set location with a location_code — a region that can span a hundred miles. That's where the 10–20% discrepancy between tools comes from, the one everybody complains about and nobody explains.
Rankfinity folds all three into one tool, pins location to exact coordinates, and puts the data on the same bill instead of a third account to open.
Figures are the publicly listed entry and top tiers for each tool as of July 2026, shown for orientation, not as a quote. The stack total is the sum of all three categories in a realistic agency setup.
Every module reads the same project — the same business, the same keywords, the same history. Add a keyword in research and it shows up in the tracker, on the map and in the report immediately.
A grid of points around the business. Each point is its own map request, anchored to its own coordinates.
Classic search positions — set by coordinates, not by region. That's where the difference with other tools comes from.
Whether ChatGPT, Claude, Gemini, Google AI Overviews and Perplexity mention you when someone asks "who's good at…".
One seed keyword in, ideas out — with volume, difficulty, CPC and intent. Tick the ones you want and they join the project.
Who actually owns the map pack around you — measured as a share of grid points, not as a hunch.
Ahrefs-style reports per domain: organic keywords, top pages, backlinks, referring domains.
A block builder: arrange the map, the table and the chart, then send a link. Or print it to PDF.
Ask your own data in plain language from Claude or any assistant — from your phone, without opening the dashboard.
A crawl of the site behind the rankings, so a position problem and a technical problem stop being the same conversation.
The keywords competitors rank for and you do not, taken from the same project rather than a separate tool.
Ask questions about your own project data in plain language and get answers built from the numbers already in your account.
Search Console and GA4 in the same dashboard — real impressions and traffic next to the positions you measured.
Four steps, each one a real screen. No onboarding call, no sales demo, no "book a slot with our specialist".
No provider account to open and no key to paste anywhere. The plan you pick already carries its monthly credit allowance, and you set a hard cap right here, before a single scan runs.
Business name, place_id for exact matching, the coordinates of your centre point,
the grid size and your keywords. Matching is on the profile ID, so "Acme Plumbing" never gets
confused with "Acme Plumbing & Heating".
Manually when there is a reason to, or on an automatic schedule you set per project. Every point is its own request, the result is written to permanent history, and it is only deleted if you delete it.
Arrange the blocks, freeze the data and send a link carrying your logo. They see the map and the difference against last month — with no login, no seat, and no access to anything else.
When the assistant names three businesses and yours isn't one of them, the customer never learns you exist. That loss doesn't show up anywhere in a classic rank tracker — your positions can look perfect while the traffic quietly moves somewhere you don't measure.
Rankfinity runs a real prompt set through ChatGPT, Claude, Gemini, Google AI Overviews and Perplexity, reads the answers and pulls out four things: were you mentioned, in which position, in what tone, and which pages got cited instead of yours.
Requests go to the platforms themselves, not through a generic router. Different assistants run different retrieval stacks, so blending them through one proxy would quietly corrupt the numbers — and the citation list is the whole point: it's next quarter's outreach plan, written for you.
Two ways to run AI checks. Through Rankfinity, using the access the platform provides, which draws down your Rankfinity credits. Or with your own API key for a supported provider, in which case those AI requests cost no Rankfinity credits — though the provider still bills your own account for what you use. Bring-your-own-key is available from the second paid plan upwards; the entry plan runs AI visibility on Rankfinity credits. Coverage is not identical for every source, so the key options are listed per provider.
Prompts are not limited to English. Rankfinity supports a broad range of languages, Bulgarian included. Depth of coverage and result quality still vary by language and by provider, so treat cross-language comparisons with care.
If two rank trackers report different positions for the same keyword, it's almost always because they define location differently. Here are the four decisions that close that gap — and why each one was made this way.
Most tools send a location_code for a city — which in the underlying database
often means the whole metro region, hundreds of square miles. The result is an average that matches
nobody standing anywhere in particular.
Rankfinity sends location_coordinate — an actual point.
Compared keyword by keyword against a dedicated rank tracker, this produces agreement
instead of a 10–20% spread.
rank_absolute counts everything on the page, ads and SERP features included.
rank_group counts organic results only — which is what people mean
when they say "position".
And when a featured snippet is present it takes position 1, so organic results shift by +1. Skip that and you'll report a number-one that's actually number two.
Substring name matching is the quiet killer of accuracy in local tools: "Delta Plumbing" happily matches "Delta Plumbing & Heating", "Delta Drain" and "New Delta Plumbers".
Rankfinity compares on place_id / cid — the unique
profile identifier. Either it's them or it isn't. There is no "probably".
Requests come in two modes: live and standard queue.
The difference is a few minutes of waiting — and roughly three times the credits.
So every batch and scheduled scan goes through the standard queue, and live is reserved for the "scan now" button. Same data, a third of the credits.
On the legal side, briefly: Rankfinity doesn't scrape Google. The data is bought through an API from a provider that runs its own collection infrastructure — you receive a structured result. No requests leaving your address, no proxies to maintain, no prohibited use of a Google API.
A client who can't see what you did churns. A client who gets a link every month — where last month was orange and this month is green — doesn't ask "what am I paying you for".
That's why reporting is a first-class module, not an export button bolted on the end: frozen data snapshots, your logo, a share link with no login, and a before/after view drawn on the same grid.
Start freeAbout €23 per client. If that adds €100 to one retainer, the tool pays for itself on the first client and the other four are margin. Software and data arrive on one invoice — no second account, no provider bill to reconcile at month end.
You pay one monthly fee and the data credits come with it. No provider account to open, no second invoice, no card on file somewhere else. Top up any time — $29 for 1,000 credits, scan nothing this month and burn nothing. Every account has a hard monthly cap, on by default.
Credits come with the plan. For orientation on a weekly cadence: a weekly 7×7 geo-grid on 5 keywords is about 500 credits a month; an organic tracker with 100 keywords is 200. Every point is logged, so you can see exactly where the allowance went. Set a cap in settings and the system stops itself before it reaches it.
Every tool below is good at its own job. The difference is that seeing the whole picture takes three of them, three logins and three invoices.
| Capability | Rankfinity | Local Falcon | BrightLocal | Nightwatch | Peec / Profound |
|---|---|---|---|---|---|
| Geo-grid map of local rankings | yes | yes | limited | no | no |
| Organic rank tracker | yes | no | partial | yes | no |
| AI visibility (ChatGPT, Gemini, AIO) | yes, 9 tabs | add-on | no | basic | yes |
| Keyword research and volumes | yes | no | partial | partial | no |
| Backlinks and domain explorer | yes | no | no | no | no |
| Location by exact coordinates | yes | yes | by city | yes | n/a |
| Any market, any language | yes | partial | partial | yes | varies |
| Pricing model | flat fee, credits included | credits per point | plan + add-ons | keyword slots | prompt slots |
| Runs on your own server | yes | no | no | no | no |
| You own the database | yes | no | no | no | no |
| Access from an AI assistant with agent (MCP) | soon | no | no | no | no |
| Entry price | $29 / mo | $25 / mo | $39 / mo | $32 / mo | $99 / €89 / mo |
Reflects publicly listed capabilities and pricing as of July 2026, shown for orientation, not as an offer. Products change — check the current details with each vendor before you decide.
Provider credentials and access tokens are stored encrypted in the database, never in a plain text file.
The whole site sits behind HTTPS and authentication — including the API console. Nothing that touches your data is left open.
Set a monthly limit. Every request is counted and the system stops itself before it crosses the line. No runaway bills.
Self-hosted, the database is one file you own. Nobody can lock it, price you out of it, or shut it down.
It doesn't scrape. Rankfinity buys structured data through an API from a provider that maintains its own collection infrastructure. You receive ready JSON — no requests to Google from your address, no proxies to maintain, no violation of Google API terms.
It's the same model Local Falcon, BrightLocal and most rank trackers use. The difference is that here it arrives on one bill together with the software, instead of a second account you have to manage.
Yes, from the second paid plan upwards. Connect a key for a supported provider and those AI requests draw no Rankfinity credits. That does not make them free: the provider bills your own account for what you use.
The entry plan still gets AI visibility, run on Rankfinity credits. Not every AI source supports a customer key in the same way, so the options are listed per provider.
Yes. Prompts can be written in a broad range of languages, Bulgarian included.
Coverage is not uniform. Depth and answer quality differ by language and by provider, so I would not claim identical results everywhere.
No. On the hosted plans the data credits are part of your subscription — you open a Rankfinity account and nothing else. One login, one card, one invoice at the end of the month.
If you'd rather own the whole chain, the self-hosted licence still runs on your own provider key and your own server. Either way every scan point is logged, so you can see where it went, line by line.
A 7×7 geo-grid scan on five keywords is 245 scan points. An organic tracker with a hundred keywords is 100. AI visibility is priced per prompt-run across all three models.
Every plan carries a monthly allowance sized for a normal agency rhythm, and you can top up. Set a cap in settings and the system stops before it, so a bill can't surprise you.
Almost always location. Most tools define a city through a region code that can cover a hundred miles. The result is an average that doesn't match what a real person standing downtown sees.
Rankfinity sends exact coordinates. Compared keyword by keyword against a dedicated tracker, that produces agreement instead of a 10–20% spread.
Yes. Plumbers, cleaners, movers, locksmiths and electricians serve an area, not an address — and the geo-grid is most useful exactly for them. You set a centre and a radius and see how far your visibility actually reaches.
The usual surprise is that a business dominates one neighbourhood and vanishes two streets over — something a single averaged position will never show you.
A prompt set runs through ChatGPT, Claude, Gemini, Google AI Overviews and Perplexity. The answers are read and four things are extracted: whether you're mentioned, in what position, in what tone, and which pages were cited as sources.
Requests go to the platforms themselves rather than through a generic router, because different assistants run different retrieval and ranking stacks. Blending them would make the numbers wrong.
Three ways, and they are deliberately different. A share link opens one fixed report: no login, read-only, your branding, revocable at any time. A read-only client dashboard lets the client follow the live data without being able to change anything. A project-scoped user is a real account that can only see the one project you assign.
For most agencies the share link is still the easiest: the client gets one link a month rather than another system to log into.
The app runs in a Docker container on a server you choose — a €5/month VPS is plenty. Two commands and it's up. The database is a single file on that server.
The practical difference: nobody can raise your price, sunset the product, or lock your history. If you ever stop taking updates, the install keeps running.
Yes. Scan manually whenever there is a reason to: after a profile edit, before a report, when you suspect a competitor moved.
You can also set an automatic schedule per project. Scheduled runs go through the standard queue, every point is logged, and the monthly spend cap applies to them exactly as it does to manual scans.
It lets you query your own data in plain language through an AI assistant with agent — from your phone, on the way to a meeting: "how did the downtown location do this week versus last?" and you get real numbers back without opening the dashboard.
Read-only by default, behind its own token, revocable whenever you want.
Keywords, projects and competitor lists — yes, through a spreadsheet import. Historical positions from another tool can't be imported honestly, because they were measured with different location parameters. Merging them would mix two different units of measurement.
The cleaner approach is to run the first scan as a new baseline and go forward from there.
Not yet, and I would rather say so than show you a mockup. The modules are working, but access is not open, so there is no public demo and no link I can send you today.
What the list buys you is order. When access opens, the people on it go first, and they see the real product: a project with real keywords from a local service niche, not illustrated screenshots.
The modules are working and access is opening soon. Leave your details and you're on the early-access list — when it opens, you're in before it goes public. No calls, no credit card, no drip sequence.
Start freeOne person reads every request. No sequence, no newsletter.
