Privacy
Who we are
TenantWard is run by TenantWard LLC. It decides how the data described here is used. For anything about your data, email hello@tenantward.com.
What we store
Your account, your company details, the clients, connections and domains you add, and the results of each check: scores, issues, their history and the notes and decisions your team records. This is what makes trends, verification of fixes and a shared queue possible. Only members of your account can read it.
The free domain check
When anyone checks a domain on our site, we record the domain, its score and the issues found, the page you came from, and a one-way hash of your IP address so we can count repeat visits. We don’t keep the IP address itself. We use this to improve the check and to contact businesses whose domains have problems we can help with. If you ask us to email you the report, we keep your address and what you agreed to, send the report, recheck the domain weekly for 12 weeks (emailing only when something changes), and may send occasional follow-up about that domain; every email has a one-click unsubscribe, and after that we don’t email the address again.
Your sign-in
Passwords are handled by Supabase Auth and stored only as a hash. When you choose one, the first five characters of its SHA-1 hash are sent to Have I Been Pwned’s range service to check it hasn’t leaked; the password itself never leaves the page. Two-step sign-in uses an authenticator app; if TenantWard support resets it for you, that’s recorded in your account’s Activity and emailed to you. If it’s switched on, sign-up uses Cloudflare Turnstile to tell people from bots.
Microsoft 365 and Google Workspace connections
An administrator connects Microsoft 365 or Google Workspace by approving TenantWard’s own app, with the permissions listed on the connect page and under Settings › Connections. A read-only connection can only read. Our server then reads settings, role assignments, app consents, license counts, sharing settings, device compliance and each user’s name, email, sign-in and MFA status, on a schedule and when you press Recheck. Exchange checks send a fixed list of commands through our server to Microsoft. Checks never read the contents of email, files, calendars or chats, and never see passwords.
To reach your Microsoft 365 or Google Workspace without anyone signed in, TenantWard keeps what Microsoft or Google issued for that connection (for example a sign-in for a partner account, or a Google refresh token) encrypted in Supabase Vault, readable only by our server. Removing the connection overwrites it at once. The approval itself lives in your directory: delete TenantWard’s app there to end it on your side too.
The user list, issues and results of each check are saved to your account.
The TenantWard agent, on a computer where your administrator installs it, sends that computer’s name, model, serial number, signed-in user, security settings (antivirus, encryption, firewall, updates, administrators) every 4 hours, and once a day the names, versions and publishers of installed programs, SQL Server and .NET. It never reads files, mail or browsing.
TenantWard’s use and transfer of information received from Google APIs follows the Google API Services User Data Policy, including the Limited Use requirements. Google Workspace data is used only to show your company’s checks, alerts and changes to the people in your account. It is never sold, never used for advertising, never used to train AI models, and no person at TenantWard reads it unless you ask us to for support or the law requires it.
Pooled, anonymized data
TenantWard pools settings and events across customers: which checks pass, how long fixes take, which senders and outside apps other companies marked dangerous, which outside Microsoft 365 apps companies have and remove (shown only once at least five other companies have an app), and score distributions. Never the contents of email, files, calendars or chats, and never names, addresses or anything that identifies a company or a person. Pooled data powers shared threat warnings, pass rates and benchmarks inside TenantWard, and reports we may publish or license. Google Workspace data is used only inside TenantWard and is never part of anything we publish, license or sell. An owner or admin can opt out in Settings › AI and data; from then on nothing from the account is added, and the account still gets warnings from what others share.
Changes, onboarding and offboarding
A change runs on our server through the connection, when someone on your team presses Fix it or approves a request, or on its own in two cases you switch on: autopilot, which puts settings back the way your baseline says (only changes that can be undone, inside your change windows, after a learning period), and automatic incident response, which can block a sign-in or end sessions for an account that looks taken over. Your change policies decide who approves what.
Every change is planned, backed up, applied, checked and recorded: what changed, who or what started it, and when, so it can be reviewed and, where possible, undone. Passwords set for people are shown once and never stored.
Monitoring
On Protect and above, with monitoring on, our server reads directory audit events, sign-ins and inbox rule definitions every 15 minutes. It stores the alerts it raises, settings changed outside TenantWard, and the countries each person signs in from. It never reads message contents.
Mailbox investigations
An investigation reads sign-in history, sign-in methods, app consents, mailbox rules and forwarding, message trace and the audit log for the account you choose. It lists the IDs of messages accessed from suspicious locations but never opens their contents. What it finds is shown to you and kept with the incident if you open one.
Suppliers
On paid plans, once a week, TenantWard reads the last 10 days of Microsoft 365 message trace (who sent mail to whom, never subjects or contents) to find the outside domains your company emails most. It keeps only those domain names and how many messages went each way, then checks each domain’s public DNS records. No email address or message is stored.
Domain checks
Our server reads the domain's public DNS records, its website certificate and its public registry entry. Nothing private is involved.
DNS hosts
If you connect a DNS host (Cloudflare, GoDaddy, AWS Route 53, Azure DNS or IONOS), the token or key you paste is encrypted and stored in a vault that only our server-side DNS service can read; it’s never shown again or sent to a browser. It’s used to read your DNS records, to change email security records (SPF, DKIM, DMARC, MTA-STS) when you approve a fix, and, on Protect and above, to check daily for unexpected changes. Disconnecting overwrites it at once. One-click approval through Domain Connect needs no key: you approve the change on your DNS host’s own screen.
Leaked-password alerts
Microsoft and Google check known leaks for your users’ passwords themselves. TenantWard reads their results (Microsoft Entra ID Protection’s leaked-credentials detections, and Google Workspace’s Alert Center) and turns them into alerts.
Ticketing systems
If you connect ConnectWise PSA, Autotask or HaloPSA, the API keys are encrypted in the same vault as DNS keys and used only by our server to open a ticket for each alert you chose and add a note when it’s resolved. The ticket contains the alert’s title, explanation, client, connection and domain names, the details behind it (such as the settings that changed, or the person involved), and a link back. Disconnecting overwrites the keys at once.
Employee help desk and requests
When staff use your help desk, we store their work email, name, what they asked for and any phone number they give, plus the replies and notes your team adds. Only people in your Microsoft 365 or Google Workspace directory can use it. The one-time code we email them is stored only as a hash and deleted from the email record once sent; sign-in sessions last 12 hours.
Reported email
When someone reports a suspicious email, we keep its headers (sender, reply-to, subject, message ID and authentication results), the addresses its links point to, and attachment names and sizes. The message body and attachments are not kept. If AI is on, the sender, subject, link addresses and the first part of the text are sent to Anthropic for a second opinion. If an admin approves removing it, our server finds copies of that one message by its ID and moves them to Recoverable Items, recording where each came from so it can be put back.
HR systems
If you connect BambooHR or send events from another HR system, we read each new starter’s and leaver’s name, work email, job title, department, manager and start or end date to prepare an onboarding or offboarding request. API keys and webhook secrets are encrypted in the same vault as DNS keys. Disconnecting overwrites them at once.
MFA rollouts and insurer forms
An MFA rollout emails instructions to the people you choose, at their work address only, and records who has enrolled and how many reminders they’ve had. An insurer’s form you upload is sent to Anthropic to find its questions; we keep the question wording and your answers, not the file.
AI help
AI help is off for every company until an owner or admin turns it on for that company and signs for it. Until then nothing about the company is sent to an AI model. We keep the record of who turned it on, what they confirmed and when.
When you ask for an explanation, the issue and what the check found are sent to Anthropic’s API to write the answer. When you ask a question, the company’s open issues and its people list are sent too (for up to 250 people: name, email address, admin roles, MFA status, license and last sign-in), so it can answer questions like “who has no MFA”. Answers are stored in your account with any rating you give. Shared, general explanations of each check are written without any company’s data. Anthropic doesn’t train on API data. Switch it off again in Settings and nothing more is sent.
Visit counts
We count page visits with Vercel Web Analytics, which sets no cookies and doesn’t follow you to other sites. It sees the page address, the referring site, your country and your browser and device type. Before an address is sent we remove anything after the “?”, private-link codes, the domains you check and record ids, so it never sees your clients or your data. When you arrive from a link we tagged (an email signature or a post), we also count the tag and the page you landed on, per day, with nothing that identifies you.
Ads
We advertise on Google. To learn which ads lead to sign-ups, the public pages of this site (the home page, pricing, guides and free tools) load Google’s ad tag, which sets Google cookies and tells Google which of those pages you visited. It only runs when the page’s address has nothing private in it: a page showing a domain you checked, or an address carrying an email or a code, doesn’t load it, and Google isn’t told about a page like that as the one you came from. When you create an account it runs once more to count the sign-up, from the sign-in page or the first page of your new account, with the address cut down to the page’s name. Apart from that it never runs on signed-in pages, the sign-in page or private links, and it sees nothing about your company, your connections or your data.
If your browser sends Global Privacy Control, the tag doesn’t load at all. In the European Economic Area, the UK and Switzerland it doesn’t store ad cookies. You can also turn off personalized ads in your Google account’s ad settings.
Billing
Payments are handled by Stripe. We store your Stripe customer and subscription references, plan and billing dates, never card or bank details.
Keeping and erasing data
Activity, alerts, changes made outside TenantWard and check results are kept for 400 days. Emails we sent and background job records are kept for 90 days, records of calls to Microsoft and Google for 30 days, and sign-up attempts (with your network address scrambled) for 30 days.
Remove a connection and its stored sign-in is overwritten straight away; its data is erased 30 days later unless you connect it again. Remove a client and you can put it back for 90 days; after that it’s erased. An owner can also delete a client at once, with everything about it.
Owners and admins can download all of the account’s data under Settings › AI and data. An owner can close the account there too: the subscription is cancelled, every stored sign-in and key is overwritten, everything stops, and the account’s data is erased 30 days later, along with the sign-ins of people who belong to no other account. Until then, email hello@tenantward.com to reopen it. Erased data can stay in our database host’s backups until they roll over, and Stripe keeps its own billing records.
For anything else about your data, email hello@tenantward.com.
TenantWard is in early access. This page will be expanded into a full privacy policy before general availability.