Flowpoint
Flowpoint is on-premise website analytics with an AI agent that builds interactive dashboards from plain-English requests. A built-in MCP server connects Claude, Cursor or any client, and your analytics data never leaves your own infrastructure.
What is Flowpoint?
Flowpoint is a web analytics platform that you run yourself. Rather than shipping visitor data to a vendor's cloud, you licence the software and deploy it on-premise — your servers, your cloud account, your network — then add a lightweight JavaScript tag to the pages you want to measure. From that point on, the analytics are collected, stored and processed entirely inside your own walls.
What sits on top of that pipeline is an AI agent. Instead of hunting through a fixed report library, you describe the dashboard you want in ordinary English — traffic overview for the last seven days, conversion funnel from homepage to signup, top sources by country — and the agent assembles it: charts, tables, funnels and metrics, refined by carrying the conversation further. No SQL and no code are involved.
The second distinguishing piece is a built-in MCP server. It exposes traffic, sessions, events, sources and funnels as callable tools, so Claude, Claude Code, Cursor or any MCP-compatible client your team already uses can read the analytics and assemble dashboards directly. Because that server is part of your deployment, requests never leave your network, and you decide whether inference runs on a local model or a cloud one such as Claude.
Flowpoint draws a deliberate line between the two layers: the agent writes the React code defining the interface, and only afterwards does your browser fetch the actual figures and load them in, so the model produces a view without ever reading the values behind it.
The measurement itself is conventional and thorough — page views, click tracking, sessions, device and browser details, approximate location from an anonymised IP, UTM parameters, custom events fired through a flowpoint() JavaScript call, and web performance metrics such as load time and layout shift. Installation guides cover HTML, React, Next.js, WordPress, Shopify, Framer and Angular, and the tag is claimed to take under two minutes.
The publisher is Flowpoint Analytics Ltd, registered in England and Wales under company number 14068900. Access is by commercial licence only: scope, term and fees are negotiated per customer, starting from a request form and a scoping call.
What it does
- Build an interactive dashboard by describing it in plain English
- Query traffic, sessions, events, sources and funnels from Claude, Cursor or any MCP client
- Watch live visitors, sessions, events and conversions in real time
- Map conversion funnels and pinpoint where users drop off
- Slice the data by device, browser, country and UTM source
- Track custom events and reuse them as funnel steps or conversion goals
- Export funnels and reports to PDF for stakeholders
When to use Flowpoint / When not to
A quick filter to help you decide if Flowpoint is the right fit.
When to use Flowpoint
- Regulated or security-conscious organisations that cannot send visitor data to a third-party analytics cloud
- Growth and performance marketers who live in UTM reports, traffic sources and conversion funnels
- Product and UX teams trying to see exactly where users drop out of a journey
- E-commerce teams working on conversion, in the vein of the Rentoza and ActionVFX case studies
- Platform and DevOps engineers who already run software in-house and want analytics to sit beside it
When not to use Flowpoint
- Anyone hoping to sign up and start today: access begins with a licence request and a scoping call
- Small teams and solo site owners who need to see a price before they talk to anyone
- Buyers who want a free plan or a trial before committing, since neither is documented
- Organisations with no servers and no ops function, because you host and run the software yourself
- Mobile-first teams: there is no iOS or Android app, and no interface language other than English is documented
How to use Flowpoint
A typical end-to-end flow, from setup to results.
- Request a licence through the contact form, giving your name, company, work email and a note about your setup
- Expect a reply from the Flowpoint team within 48 hours at most
- Take an intro call so the team can scope your environment, requirements and timeline
- Receive a licensed build sized to your deployment and your choice of model
- Deploy Flowpoint inside your own environment, with the team's help integrating it into your stack
- Add the tracking script to the head section of your pages, replacing YOUR-SITE-ID with your own site identifier
- Follow the guide for your platform: HTML, React, Next.js, WordPress, Shopify, Framer or Angular
- Instrument custom events with the flowpoint() call and reuse them as funnel steps or conversion goals
- Use the built-in dashboard UI, or add the MCP server to Claude, Claude Code or Cursor with its URL and a bearer token
- Point the agent at a local or cloud model, describe the dashboard you want, refine it by chatting, then export to PDF
Pros & Cons
Pros
- Analytics data stays inside your own infrastructure, which is the claim the entire product is built around
- GDPR, UK DPA 2018 and data-residency compliance follow from the architecture rather than from a contractual promise
- Model sovereignty: inference can run locally, and raw analytics are never handed to the model
- The native MCP server turns the AI client your team already uses into an analytics client, with no integration work
- Dashboard building is open to non-technical people, since the interface is plain English rather than SQL
- Tag installation is genuinely quick, with guides for seven platforms and a single call for custom events
- Unusual contractual transparency: published DPA, ten named subprocessors, policies versioned and dated
Cons
- No public pricing whatsoever, so a budget cannot be sized without a sales conversation
- No free plan and no trial are documented, leaving no way to evaluate the product on your own
- You supply the infrastructure and the people to run it, which rules out teams with no ops function
- The promise that nothing ever leaves your infrastructure needs qualifying: chat messages do reach Anthropic's API when a cloud model is used
- The legal pages still speak the language of a SaaS subscription, with accounts, recurring Stripe billing and references to the vendor's own servers
- All ten disclosed subprocessors sit in the United States, including OpenAI for report data queries
- No security certification is published, there is no mobile app, and no interface language other than English is documented
Pricing & Plans
No price is published. Flowpoint is sold exclusively as a commercial on-premise licence whose scope, term and fees are, in the wording of the terms of service, agreed on a per-customer basis according to the deployment and the requirements; the structured data on the homepage says simply that the software is licensed and that pricing is obtained by contact. No entry-level amount and no currency can therefore be quoted. Neither a permanent free plan nor a free trial is mentioned anywhere on the site, although neither is expressly ruled out either. Payments are processed by Stripe, and a pro-rata refund may be claimed within seven days where the service has been unavailable for more than 24 consecutive hours.
Data, GDPR & hosting
A consolidated view of how Flowpoint handles your data.
GDPR overview
Implementation is concrete rather than declarative. A dedicated GDPR statement, version 1.1 revised March 2026, sits alongside a published Data Processing Agreement that, in the vendor's words, applies automatically to every client. The privacy policy names its legal basis, Article 6(1)(b) for account data, lists the six data subject rights, promises a reply within 30 days and points to the ICO as supervisory authority. Transfers outside the EEA and the UK are covered by Standard Contractual Clauses, adequacy decisions or the UK IDTA, and security is tied explicitly to Article 32. Flowpoint also offers help with impact assessments and client audits. Two gaps remain: no Article 27 representative and no DPO are named, and no security certification is published.
Who owns the data?
Section 7.2 of the terms leaves ownership with you: reports, dashboards built by hand or generated by the AI, and exported data all remain yours, and Flowpoint expressly claims none of it. It takes only a limited, non-exclusive licence to store and process that content in order to run and improve the service. What stays the vendor's is the software itself, its code and its design, which are the exclusive property of Flowpoint Analytics Ltd. The GDPR statement is blunter still: you are the data controller for anything collected through the platform and Flowpoint acts as your processor. The Data Processing Agreement carries that through, confirming that the controller retains full control and may require deletion or return of the data at any point.
Reuse rights
Yes, and without asking. Dashboards, reports and exported data are yours to reuse, circulate or publish as you see fit; nothing in the terms makes that conditional on the vendor's permission, and funnels and reports can be exported to PDF for stakeholders. The GDPR statement adds export and deletion tooling so you can manage the personal data you have collected. Two duties travel with that freedom. Section 5.2 makes you responsible for obtaining your visitors' consent, disclosing Flowpoint in your own privacy policy, running a cookie mechanism where the law demands one, and answering subject access requests. Section 8 forbids extracting data from the service by automated means outside the API provided for it.
Data retention & training
Hosting summary
Two layers have to be told apart. Your analytics are hosted wherever you deploy Flowpoint: the product is on-premise, so collection, storage and processing happen inside your own infrastructure, and the vendor states that nothing is sent to it or to any external service. The vendor's own services are a different matter. Its Data Processing Agreement names ten subprocessors, all located in the United States: Google Cloud for hosting, Stripe for payments, Anthropic PBC for AI dashboard generation, OpenAI for report data queries, plus Intercom, Slack, Google Workspace, HubSpot, Mailchimp and Namecheap. The controller is Flowpoint Analytics Ltd, registered in England and Wales and based in London, and the terms, the DPA and any dispute fall under the law of England and Wales. Transfers outside the EEA and the UK rely on Standard Contractual Clauses, adequacy decisions or the UK IDTA. Security measures cited are TLS in transit, encryption at rest, role-based access control, and VPN with multi-factor authentication for internal infrastructure. No broad hosting region is named anywhere on the site.
Things to keep in mind
Risks and trade-offs to weigh before adopting Flowpoint.
- AI-generated dashboards come with no accuracy guarantee: the terms put verification squarely on you, and a mis-specified chart can look authoritative while measuring the wrong thing
- Deciding from a dashboard you did not build encourages a quiet loss of analytical skill, because knowing how to frame a question and check a calculation atrophies when an agent always does it for you
- The privacy promise is architectural, not absolute: chat text reaches Anthropic's API when a cloud model is used, and Flowpoint itself advises keeping personal data, trade secrets and confidential information out of those messages
- All ten disclosed subprocessors sit in the United States, including OpenAI for report data queries, which is worth settling early if data residency is your reason for buying
- The terms transfer responsibility for visitor consent, cookie mechanisms, privacy disclosures and subject access requests to you, so deploying the tag without that groundwork creates your exposure, not the vendor's
- Liability is capped hard, at the fees paid over the previous twelve months or one hundred US dollars where none were paid
- Running the software yourself means patching, hardening and availability become your responsibility: self-hosting moves the risk, it does not remove it
Setup & Integrations
Technical difficulty
Two levels, and they sit far apart. Instrumenting a site is easy: one script in the page head, claimed to take under two minutes, with step-by-step guides for HTML, React, Next.js, WordPress, Shopify, Framer and Angular, plus a single JavaScript call for custom events. Deploying the platform is not: you install and operate licensed software inside your own servers, cloud or network, and wiring up the MCP server means supplying an endpoint and a bearer token. Flowpoint helps with integration, but you need infrastructure and an engineering team able to run and maintain it.
Deployment
Integrations
Behind Flowpoint
Fundraising
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
Where does Flowpoint run, and where does my data live?
Which AI model does it use?
Does the AI see my raw analytics data?
So does anything leave my environment?
How much does it cost?
Is there a free plan or a free trial?
How difficult is it to install?
What does the built-in MCP server add?
How long is my data kept, and is there a DPA?
Who is behind Flowpoint, and which law applies?
Should you pick Flowpoint?
Flowpoint answers a question many teams have quietly given up on: can you have modern, conversational analytics without handing your visitor data to somebody else's cloud? Its answer is to stop being a SaaS. You licence the software, deploy it inside your own environment, and the analytics are collected, stored and read there. The AI agent that turns a sentence into a dashboard, and the MCP server that lets Claude or Cursor query your traffic directly, sit on top of that arrangement rather than around it.
For regulated industries, security-conscious enterprises and anyone bound by data-residency rules, that is a genuinely different offer, and the supporting paperwork is better than most: a published DPA, ten named subprocessors, policies versioned and dated to March 2026, and a promised 48-hour reply to licence requests.
What you accept in exchange deserves naming plainly. There is no price list, no free plan, no trial and no self-service, so evaluation runs through a sales cycle. You provide the infrastructure and the people to keep it running. And the headline promise carries one qualification the site itself supplies: point the agent at a cloud model and your chat messages do travel to Anthropic's API, even though your raw analytics do not. Only local inference closes that loop entirely.
The legal pages also still carry SaaS-era vocabulary — subscriptions, recurring billing, references to the vendor's own servers — that has not been rewritten for the on-premise story the product pages tell. It reads as a young company mid-repositioning rather than as a contradiction, but it is worth raising before signing. If privacy is why you are shopping, Flowpoint is a serious candidate. If speed and price transparency are, it is not.
- Choosing a selection results in a full page refresh.
- Opens in a new window.