Data Use Policy

This policy explains how SEOmise uses the data in your workspace and the data we get from other providers. It adds detail to the Privacy Policy and the Terms of Service.

1. Inside your workspace

Your workspace's data (your sites, keywords, scans, connected Google data, reports and approvals) is used to provide the features you use. Only people in your workspace, people you share a project with, and our support staff with a recorded reason can see it. Agencies see the projects of the clients they manage.

2. Reusing answers we already paid for

Many features buy data from outside providers (for example search results, keyword figures and backlink data). To keep costs down, an answer bought for your workspace is kept and reused for your workspace for a period set per source, within what the provider's licence allows. For example, an AI answer to exactly the same task and input in your workspace is reused for up to 7 days.

• Answers are reused only within the workspace that bought them.
• Sharing an answer between workspaces stays off for every source until its provider's licence allows it in writing and we switch it on. Your own Google data is never shared between workspaces, whatever the settings.
• Ask Mise conversations, AI visibility questions, review replies and similar personal tasks are never reused.

3. How we use data to improve SEOmise

We use the data the service handles (scans, audits, searches and queries, results, Ask Mise conversations, and which suggestions people accept or reject) to:

• operate, secure and support SEOmise;
• improve it, including developing, training, testing and evaluating our own models, rules, rankings and recommendations (the Intelligence Core); and
• build aggregated, de-identified statistics and benchmarks.

Customers grant us a licence for this in the Terms of Service. We keep this data for the periods in the Data Retention Policy, including in the Data Vault. Where personal data is involved, our lawful basis is legitimate interests, and anyone can object by e-mail at [email protected].

The Intelligence Core

The Intelligence Core is SEOmise's own engine for advice. It has three parts that use data:

• The rulebook: our versioned library of search engine guidance and best practice. It is built from official sources (such as the search engines' own documentation); staff approve every change, and each version is published with what changed.
• Signals: figures about your own sites that SEOmise already holds (for example scan results and rank positions), read once a day, and what people do in SEOmise (for example which suggestions are accepted), counted per workspace.
• Learning from results: comparing what changed with what happened afterwards, for your site privately and, anonymously, across groups of similar sites. Results are correlation, not proof.

Anonymous group figures

• A group figure is published only when at least 10 separate businesses are in the group. An agency and its client workspaces count as one business.
• No single business may make up more than 30% of a group's sites.
• Groups are fixed (platform, kind of site, region, industry and pairs of these). Nobody can ask the pools a free question, and a small group that could be worked out by subtracting one figure from another is held back.
• A little random noise is added to every published figure, so no single business's value can be read from it.
• Pooled data never contains customer names, domains or website addresses.
• Text people type is never pooled as written; at most it is reduced to topics and counts, with personal data removed first.
• The industry your workspace chose is used only to form these groups and to tailor advice.

Never pooled: Google data (Search Console, Analytics, Business Profile and any other Google API user data) is never used to develop, improve or train AI or machine-learning models, and is never pooled. Google's API Services User Data Policy restricts this use. Bing AI reports you upload are not pooled. Data licensed from search and backlink providers is pooled only where its licence allows.

Opting out: a workspace owner can switch off Help improve SEOmise's advice (Account). From the next daily update, the workspace's figures are left out of every group figure. It does not stop the use of your data inside your own workspace (your reports, history and advice), because the service needs it to work for you.

4. The Data Vault

The Data Vault keeps an encrypted record of every outside call SEOmise makes and its answer, every AI request and answer, every Ask Mise question and answer, every page our crawlers fetch and every audit, by channel. We use it to check exactly what was asked and answered, to reuse answers instead of paying again, to investigate problems and complaints, and to improve the service.

• Records are encrypted at rest. Passwords, keys and tokens are removed before anything is written.
• Each record is deleted at the end of its period (at most 1 year), sooner where its source requires it: Google data after 30 days, Ask Mise records with the leads after 12 months, third-party data as its licence allows.
• Only staff whose role includes the Data Vault can open it, with a fresh code from an authenticator app every time. It locks after 5 minutes without activity. Every unlock, search, record opened and export is in the audit log. Staff viewing as a customer cannot open it.

5. AI

• AI runs through our own gateway, only for a task someone started (or Ask Mise on our website). Only the text the task needs is sent: for example a page's title and content to suggest a better description.
• Under our agreements with them, we do not allow the AI providers we use to train their models on customer data sent through SEOmise.
• AI output is a suggestion. Nothing reaches your website, Google Business Profile or customers until a person approves it.
• Our security reviews also use AI to review SEOmise's own code. Secrets are removed before anything is sent; no customer data is involved.

6. Third-party data

• Each source of outside data has a rights record (what we may keep, for how long, whether it may be shared between workspaces, shown publicly or exported). The code follows it.
• When a licence ends or a provider asks, staff can delete every stored copy of that provider's data at once ("Delete stored third-party data"), including its Data Vault records.
• Data from providers can be incomplete or late. We show where each figure came from.

7. Our crawler and the page catalogue

Our crawler, SEOmiseBot, reads public web pages for scans and audits. It respects robots.txt and identifies itself (see SEOmiseBot). We never keep the raw HTML of pages it reads.

Adding the public addresses of pages from customers' scans to our own page catalogue is a separate switch. It is off: customer scans do not add anything to the catalogue.

8. Your website, connectors and the tracker

• The WordPress, Laravel and universal connectors read what SEOmise needs about your site and apply only the changes you approve (or an automatic option you switched on).
• The Live view tracker on your site records visits for your reports; it is your data, processed on your instructions. Earlier page captures used as a backdrop are kept for 180 days.
• ReFame (for approved PR agencies) and the CDN (Edge) module use your data only for the work you set up; CDN changes need approval like any other change.

9. Changes

We will publish a new version of this policy with what changed. A material change is announced by e-mail and in the app at least 30 days before it applies.