Privacy and data use

Updated October 4, 2026.

Optional Google sign-in and account saves

When you choose Google sign-in, Google supplies a verified email address and a stable account identifier. We store your email, a hashed identifier, and account creation and update times to identify your account and provide access to your saved work. We do not request your contacts, Drive files or profile photo, and do not retain Google access or refresh tokens. Google handles its sign-in screen under its privacy policy.

Signing in does not upload existing work from your browser. Save to account sends the item you choose to our server so you can open it on another signed-in device. Saved items can include homes, property searches, references to public buying briefs, and private comparisons with addresses, prices, financial assumptions, priorities and notes. Each save is your choice; later edits need another save. We use this information to provide your account library, not to enroll you in the newsletter or request realtor contact.

A secure session cookie keeps you signed in for up to 30 days. It cannot be read by page scripts; the server stores a hashed session token. Signing out invalidates that session. We retain your account and its saved items until you delete them. On the account page, you can export your saved items, delete individual items, or delete the account. Account deletion removes its identity, saved items and sessions from active storage and makes its sessions unusable. Recovery backups can retain deleted data for up to 35 days; application logs have 30-day retention. Deleting an account does not erase device-only copies, downloaded files, or public buying briefs.

Private home comparisons

On Compare homes, your manually entered addresses, prices, financial assumptions, priorities and notes stay in this browser tab’s memory while you navigate this site, unless you choose a save action. Save on this device stores one comparison in this browser’s local storage, including its financial figures and evidence notes. Save to account stores your chosen comparison on our server as described above. Later edits are not saved automatically. Refreshing, closing the tab or leaving the site loses unsaved edits. Internal navigation keeps the active draft; another tab does not receive it. Delete saved copy removes the local copy; clearing browser data also removes it. Device-only entries are not sent to our server. We do not include private comparison entries in URLs, analytics or public city briefs.

Pasted home links and notes are parsed locally in this tab. We do not fetch the linked page or send the paste to a language-model service. Review and confirmation keep only the chosen fields, references and field-source information; the raw paste is discarded. Inline address search sends only your intentional query and search area to our server. Opening a selected or linked public-record requests the existing city and parcel identifier from our server; these requests can appear in normal access logs. We do not send your private notes, prices or financing with them. External listing URLs you enter are stored as references; we do not fetch their content. Looking up an address in another tool follows that tool’s policy below.

Printing or downloading private content from Compare homes requires the export checkbox. Private backups and account exports can include financial figures and notes and can be read by anyone with the file. Restoring a comparison backup reads it locally without uploading it; saving the restored comparison to an account is a separate action. Private figures and notes are excluded from comparison printouts unless you enable them. There is no email monitoring of these homes.

Interactive city map

The map background loads directly from OpenFreeMap. Its servers receive normal network information such as your IP address and the map tiles your browser requests. We do not send it your selected cities, property searches, saved homes, personal notes or financial assumptions. We do not request your location. The city comparison remains usable if the background cannot load. OpenFreeMap privacy policy.

Nearby project lookup

When you submit a street address on the nearby-project map, our server sends it to the U.S. Census Bureau to find an approximate street location. This feature does not create an address book or store your search as a saved home. Address lookup requests can appear in service access logs. After you confirm a location or choose a map point, its coordinates and filters appear in the page URL and browser history; anyone you share that link with can see the selected location. Map tile requests also indicate the area being viewed. Property-to-map links use the approximate center of the county parcel.

Private ownership-cost scenarios

Your monthly housing target, purchase cash and financial assumptions stay in memory in this browser tab while you navigate the map, current buying brief and ownership-cost page. Refreshing or closing the page clears them; Clear private budget resets them immediately. We do not send these figures to the server, save them in browser storage, put them in URLs, include them in analytics, or add them to saved public briefs or printouts. A saved brief does not restore a private budget. Open a current brief to make a new cost comparison.

My City Shortlist

A saved shortlist is a public, dated record of your selected cities, home type, optional priorities and the published evidence shown. Anyone with its link can read it. Identical briefs share the same link. It contains no name, contact information or personal notes. Saved records remain available; the first release does not provide a reader edit or delete function.

Your chosen investigation questions and completion state stay in this browser's local storage, separate from the public record. They are excluded from links, analytics and printouts. Use Clear checklist to remove them. If storage is blocked, the checklist works for that visit.

Shortlist measurement counts rendered previews, successful saves, link-copy results, update-check results and explicit helpfulness feedback. It includes only city count, home type, preview/saved surface, event category and date, with a scoped hash of a random browser-tab session identifier. It excludes city identities, brief IDs, full URLs and checklists. Events expire after 100 days. New passive measurement requires the optional analytics choice below and respects Do Not Track and Global Privacy Control; deliberate feedback can still be submitted without enabling analytics. Without consent, feedback uses a fresh one-off identifier rather than a stored tab session. These are session signals, not unique people or proof that a copied link was shared.

Free research emails

When you explicitly request the buyer brief, we store your email, confirmation status, request and consent dates, the consent wording/version, chosen cities, and any optional budget, buying timeline, or topic preferences you supply. We use these to manage the monthly newsletter and plan coverage. We also retain the public page associated with your first signup to understand which content helps readers. Newsletter signup is a separate choice from creating an account or requesting an introduction.

While email delivery is being prepared, we save new requests as pending and send nothing. We will ask you to confirm when delivery is ready; pending addresses cannot receive newsletter campaigns. Repeat submissions do not replace existing preferences, undo an unsubscribe or create extra signup counts. We retain pending requests until confirmation or a removal request. To remove an unconfirmed request, contact research@bayareahousingalpha.com. A request can contain a typo or an address entered by someone else, so it does not establish a verified subscriber.

Restricted administrative reports show aggregate saved requests separately from email-confirmed and active subscriptions. Counts are derived from subscription records, with one retained record per normalized email address; they are not unique-reader counts. We do not send subscriber emails, email hashes, city choices or profile preferences to Google Analytics or buyer-workflow analytics. Signing up works without analytics consent.

Confirming the newsletter does not authorize realtor contact. Each edition includes a preference and unsubscribe link. When confirmation delivery is available, you can request a fresh management link from the subscription page. Unsubscribed and suppressed email records are retained to avoid unwanted delivery.

Introduction requests and possible compensation

Requesting an introduction is separate from subscribing or saving to an account. We use the email, cities, optional help-question categories and broad buying timeline you submit to discuss a possible introduction with you. New requests do not collect budgets, home addresses or freeform private notes. Only email-confirmed requests appear in our follow-up queue. This request permits our team to contact you; it does not authorize sharing with a realtor. We ask for your approval of a named realtor and the details to be shared before any introduction. Approval is a separate step: you see the agent, brokerage, license, contact email and unpaid-pilot disclosure, then choose the exact fields to share. Your email is required for a reply; cities, help questions and timing are optional and start unselected. Our team can send only that approved snapshot. There is no automatic transfer to a partner CRM.

We retain proposal and approval times, the named recipient, the exact approved fields, sending status and separately submitted buyer and agent feedback with your introduction request. Feedback asks only whether a conversation happened and whether it helped. Missing feedback does not count as a completed or useful conversation. These records are separate from your Google account and private saved work.

We are exploring compensated introductions and clearly labeled sponsorships to support Housing Alpha. No compensated introduction program or partner data transfer is active. If we offer an introduction involving compensation, we will explain that relationship alongside the named recipient and proposed details before asking for your separate approval. Creating an account, saving a comparison or subscribing does not grant this permission. Your private comparisons and account library are not included in an introduction request.

Use the request email link to withdraw. Withdrawal cancels proposals and approvals whose sending has not started. Messages already being sent or delivered cannot be recalled; ask an introduced agent directly to stop their follow-up. Feedback links stop working after withdrawal or closure. Pending requests expire after seven days; verified requests are retained for up to 180 days after confirmation. Withdrawn requests are retained for 30 days before scheduled deletion. Expired requests are hidden immediately; deletion from storage and backups can take longer. After a record expires, a fresh request starts a new retention period. Re-entering an email before expiry does not reopen a closed or withdrawn request or erase its approval history.

Reading signals

On published articles and reports, we count page views using a random identifier stored for your browser tab session. We store a hash of that identifier with the page path, event type, and date. We do not attach your email, full referral URL, URL query parameters, or subscription tokens to these events. These first-party events expire after 100 days. Earlier retained signals predate the updated opt-in requirement.

New passive view measurement requires the optional analytics choice below and respects Do Not Track and Global Privacy Control browser settings. Explicit feedback buttons still submit the response you choose; without consent they use a fresh one-off identifier without storing a tab session. Blocking browser storage or analytics does not restrict reading or subscription access. Counts can include bots or repeated sessions and are used as directional feedback.

Optional buyer-workflow measurement

The updated analytics choice below requires fresh permission; earlier analytics choices do not enable the expanded measurement, and older refusals stay off. When allowed, our first-party endpoint receives only fixed event labels for an eligible public landing, a visible city chart with comparable data for at least two cities, a visible home-cost result, a successful device save, a successful device reopening and a newsletter request accepted by the site. The newsletter label measures interest; a response can acknowledge a repeat request and does not establish a new or email-confirmed subscription. New and reopened results have separate labels; neither proves completion or usefulness. No home details, financial figures, notes, account identifiers, URLs or full referrers are included. Requests omit browser credentials and the referral URL. Our hosting provider still processes ordinary network information needed to handle the request.

After fresh consent, a random tab-session identifier and one finite source label may be held in session storage. The buyer-pilot campaign allows personal-network, buyer-community, search-content or partner-referral. An eligible public visit can instead carry a ChatGPT, Perplexity, Claude, Gemini or Copilot hint from an exact recognized referral host or source marker. The browser reduces that hint to a fixed provider label; the full referrer and query are never stored in analytics. Referrers and markers can be missing or misleading, so these are attribution hints, not proof that an AI service recommended us. Unknown campaign values and extra private query fields prevent measurement. No acquisition data is stored before consent; visits without a recognized source are unattributed. The server stores a separate event-scoped hash for deduplication, fixed labels and a date, with a 30-day expiry; it does not store the raw identifier or join these events to an account. Events are excluded after 30 days and scheduled for automatic deletion; physical deletion and recovery backups can retain them longer.

The home-comparison helpfulness buttons are a separate, deliberate feedback submission and work without analytics consent. Without consent they send only the selected fixed response with a fresh one-off deduplication identifier, not a stored session or campaign label. They do not enable Google Analytics. These responses are self-reports, not verified outcomes. With consent, the response may be counted with the approved source label. Withdrawal stops new optional measurement and clears the pilot tab session; it cannot recall already received events.

Optional Google Analytics

Google Analytics is off unless you allow it using the control below and our reviewed measurement configuration is active. We do not load Google’s analytics script or send it consent pings before that choice. Do Not Track and Global Privacy Control keep it off even if you previously allowed it.

When allowed, Google receives fixed public-page categories and general visit and session activity, together with normal network information such as your IP address and browser information. Analytics may create first-party cookies with a maximum lifetime of 30 days; we do not extend that lifetime on each visit. We use fixed page labels and locations, not your actual page title, query string or referral URL. We exclude account pages, saved comparisons, subscription and introduction links, individual property records and other private routes. We do not send addresses, parcel identifiers, budgets, notes, email addresses or account identifiers. We do not connect these measurements to Google sign-in or saved account work.

Enhanced Measurement, automatic user-provided data collection, Google Signals, granular location/device collection and advertising personalization are disabled. We do not use advertising pixels, session replay or automatic form capture. GA4 can still produce normal session and engagement metrics from permitted visits; these counts describe consenting browser activity, not all visitors, unique buyers or proven usefulness. Event and user retention are set to two months without resetting on new activity; Google’s aggregate reports are not subject to that setting.

You can turn optional analytics off below. This stops further measurement in this page and removes its analytics cookies; it does not remove prior aggregate reports. After Google Analytics stops on a private page or following withdrawal, it stays stopped until a fresh page load. First-party workflow measurement can run on the private comparison route only with the separate updated scope described above. Google’s privacy policy and how Google uses information from sites that use its services explain its processing.

Hosting and administration

Our hosting and email providers process information needed to serve pages and deliver requested messages. Restricted administrative access protects subscriber and request records. Infrastructure providers can process network information such as IP addresses when serving requests. Application logs are configured for 30-day retention.

Use the account page to export or delete account data. For privacy questions, support or a deletion request, email research@bayareahousingalpha.com. This contact does not mean newsletter or introduction delivery is open. Never send sensitive financial documents in these forms or support messages.

Changes to this policy

We publish changes here with an updated date. Before using previously collected information for a new purpose that requires your permission, we will explain that use where you make the choice and ask for that permission. A policy update alone does not authorize realtor contact or transfer of your saved work.

Partnership and editorial policy