Privacy Policy
- Who we are
- Scope and our role
- Information we collect
- How we use information
- What we do not do
- AI interviews, voice and recordings
- Sharing and sub-processors
- Retention and deletion
- Privacy requests and your rights
- Security
- International transfers
- Children
- Changes to this policy
- Contact us
1. Who we are
Trooly is operated by PIKAIA LIMITED, a Delaware company, with its registered address at 8 The Green, Ste A, Dover, Kent County, Delaware 19901, United States ("Pikaia", "Trooly", "we", "us").
For any privacy question or request, contact us at customer_service@trooly.ai.
2. Scope and our role
This policy covers the Trooly app for Shopify: the app inside the merchant's Shopify admin, the interview invitation shown on the merchant's storefront (popup, on-site chat and thank-you page block), the storefront app pixel, email interview invitations, and the interview pages served at merchant.trooly.ai. It does not cover the merchant's own store, Shopify itself, or other Trooly products, which have their own terms.
- Merchant data. For information about the merchant and its staff account (for example the store domain and the email used to connect a Trooly account), Pikaia decides how the data is used to provide the app.
- Shopper data. For information about the merchant's customers and store visitors ("shoppers"), the merchant is the controller and Pikaia acts as the merchant's processor / service provider. We process shopper data only to provide the app's features to that merchant, on the merchant's instructions. Shoppers should also read the merchant's own privacy policy, and may direct requests to the merchant (see section 9).
3. Information we collect
3.1 From merchants and the Shopify installation
| Data | Why we use it |
|---|---|
Store domain (*.myshopify.com), Shopify access and refresh tokens, granted permissions, store timezone and admin language | To install and run the app, call the Shopify Admin API on the merchant's behalf, and show dates and AI output in the right language. |
| Product catalog (titles, descriptions, images, product and collection IDs) and the store's published legal policies (e.g. refund and shipping policies) | To let the merchant choose products for interviews and to build the interview outline and the interviewer's knowledge, so it can answer shoppers' questions accurately. |
| Recent order line items (product IDs, titles, quantities) | To suggest which products to interview about. |
| Knowledge the merchant enters or imports (FAQ, selling points, uploaded catalog documents) | Interviewer knowledge. |
| App settings: invitation rules, audiences, popup copy, incentives, email templates, notification email address, and optional integration settings (Klaviyo, Google Analytics 4 measurement ID and API secret) | To run the features the merchant configures. |
| Trooly account connection: the email address a staff member enters to connect or create a Trooly account, a one-time verification step (protected by a Cloudflare Turnstile human check), and the resulting Trooly API credentials | To link the store to a Trooly organization that hosts its interview projects. |
| Subscription and usage records (plan, interview counts per billing period) | Billing through Shopify and plan limits. |
| In-app usage events (event names and a few structured properties, no personal data) | To understand where merchants get stuck during setup and improve the app. |
3.2 About the merchant's shoppers
| Data | Why we use it |
|---|---|
Anonymous storefront identifier and on-device state. A random ID and invitation frequency state stored in the shopper's browser (localStorage / sessionStorage); the app does not set third-party cookies. | To avoid showing the same invitation too often and to link an interview to the visit that led to it. |
| Invitation events. Popup shown / clicked / dismissed / completed, with page type, scroll depth, time on page and the products viewed in the current session; the invitation click (visit) record. | To decide when an invitation matches the merchant's rules, to enforce frequency limits, and to report invitation performance to the merchant. |
| App pixel browsing events (where the merchant uses the Trooly app pixel): page viewed, product and collection viewed, search submitted, add to / remove from cart, cart viewed, checkout started and completed. We store only product and collection IDs, quantities, amounts, currency and order ID, keyed to the pixel's client ID. Search terms, page URLs, names, emails and addresses are discarded before storage. | To build the merchant's shopper profiles and audiences (for example "viewed this product twice in 7 days") used to target invitations. |
| Shopper profile. Computed traits (such as order count, amount spent and last order date) and identifiers that link visits together: the anonymous ID, the pixel client ID, the Shopify customer ID of a signed-in shopper, a Klaviyo profile ID, and a keyed hash of an email address (never the email itself). | Audience matching and to tell new from returning customers. |
| Identity lookup. If a shopper is signed in, Shopify tells us their customer ID and we read their order count and amount spent. If not, and the store uses Klaviyo, the storefront script may send the Klaviyo identifier or email address from the store's Klaviyo cookie to our server, which looks it up in the store's customer list and returns only "new" or "returning". That email is not stored and is never put in a URL. | To tailor which invitation is shown and what the interviewer asks about. |
Interview context passed into the interview: the store, the product being viewed, cart contents and totals, pages visited in the session, customer type, Shopify customer ID (if signed in), Google Analytics client ID (if present in the store's _ga cookie) and the invitation rule. | So the interviewer can ask relevant questions and so results can be attributed to the right product and invitation. |
| Interview content. Audio recordings of voice interviews, text chat messages, transcripts, and for video (camera) interviews the video recording; plus AI-generated analysis of each interview (themes, satisfaction, concerns, short quotes, product recommendations made by the interviewer and whether they were later clicked, added to cart or purchased). | To conduct the interview and give the merchant insights, reports and follow-up recommendations. |
| Email invitations. Email address, first and last name of customers read from the merchant's Shopify customer list only if their email marketing status is "subscribed", together with order count, amount spent, last order date and product names, and customer tags used to filter the audience. We store only the recipient email address, to prevent duplicate sends, enforce a 30-day contact limit and keep an unsubscribe list. If the merchant connects Klaviyo, recipients may instead come from the merchant's Klaviyo lists or segments (email and name). | To send interview invitations the merchant requests, and to honor unsubscribes. |
| Rewards. One-time discount codes created in the merchant's Shopify store for completed interviews, linked to the interview and anonymous ID; orders that later use a reward code. | To deliver the thank-you reward and measure its effect. |
| Follow-up interviews and browser notifications (only where the merchant enables them): a shopper's explicit consent, given during an interview, to be invited again; and, if a shopper opts in, their browser push subscription (stored encrypted). | To invite that shopper to a follow-up interview. |
3.3 Consent on the storefront
By default, the storefront invitation script only starts, and only stores its anonymous ID, after Shopify's customer privacy API reports that the shopper allowed analytics processing. The app pixel declares the analytics category, so Shopify does not load it at all without that consent. Merchants can switch the invitation script's consent check off in the app settings; a merchant who does so is responsible, as controller, for having a lawful basis for that processing.
4. How we use information
- To provide the app: show invitations that match the merchant's rules, conduct AI interviews, create and deliver rewards, and send the email invitations the merchant sets up.
- To generate interview outlines from the merchant's products and policies, and to analyze completed interviews into insights, weekly summaries and alerts for the merchant.
- To attribute interviews, rewards and orders to invitations so the merchant can see results.
- To operate, secure, debug and bill for the service, and to respond to support and privacy requests.
- Where the merchant connects Klaviyo or Google Analytics 4, to send the events described in section 7 to the merchant's own account.
5. What we do not do
- We do not sell shopper data, and we do not share it with advertisers or data brokers.
- We do not use shopper data for advertising or ad targeting, or to build profiles for anyone other than the merchant who collected it.
- We do not use shopper data or interview content to train AI models.
- We do not make decisions that produce legal or similarly significant effects on shoppers. The only automated decision is whether to show or send an interview invitation, which a shopper can dismiss or unsubscribe from.
- The app does not read shoppers' postal addresses or phone numbers.
6. AI interviews, voice and recordings
- Interviews are conducted by an AI interviewer, not a person, by voice or by text chat. Shoppers take part voluntarily after choosing to open an invitation.
- Before a voice interview starts, the interview page tells the shopper that they will be talking with an AI interviewer and that the conversation is recorded and transcribed so the store can learn from their feedback; the browser then asks for microphone permission, and nothing is recorded until the shopper starts the interview and allows the microphone. The on-site text chat shows the notice "This chat is recorded and analyzed by Trooly to improve the store."
- For video (camera) interviews, the shopper must tick a box agreeing to the Consent to Participate and Data Use Agreement before the interview can start. A separate, optional box authorizes marketing use of the interview content under its own terms; it is not required to take part.
- Audio is streamed in real time through our media provider and processed by speech-recognition, speech-synthesis and language-model providers (section 7) to run the conversation and produce transcripts. Completed transcripts are analyzed by a language model to produce the merchant's insights.
- Recordings and transcripts are made available to the merchant who invited the shopper.
7. Sharing and sub-processors
We share data only with the merchant it belongs to, with the service providers below who process it for us under contract, with integrations the merchant chooses to connect, and where required by law. Interviews run on the Trooly interview platform, which is operated by Pikaia and hosted on the same infrastructure.
| Category | Provider | Purpose |
|---|---|---|
| Hosting and storage | Amazon Web Services (us-east-1: EKS, RDS for PostgreSQL, S3) | Application hosting, databases, recordings and files |
| Email delivery | Resend | Interview invitation emails and merchant notifications (recipient email, name, email content) |
| AI language models | Anthropic Claude, accessed through AWS Bedrock; Google Gemini | Interview outlines, the AI interviewer's conversation, and interview analysis |
| Real-time audio / video | LiveKit | Streaming interview audio, video and chat |
| Speech recognition and synthesis | May include Deepgram, Google, Amazon Web Services, AssemblyAI, Cartesia and ElevenLabs, depending on the interview language | Turning speech into text and the interviewer's replies into speech |
| Human verification | Cloudflare Turnstile | Protecting the merchant account-connection step from bots |
| Platform | Shopify | App installation, billing, customer privacy signals, discount codes |
Integrations the merchant connects. These are the merchant's own accounts and are used only after the merchant connects them:
- Klaviyo: invitation emails are then sent by the merchant's Klaviyo flow. We send Klaviyo the recipient's email address, the personal interview link, the reward and product names, and after an interview we write the interview result (satisfaction, concerns, churn risk, follow-up consent and a short summary) to that Klaviyo profile. The merchant may also sync Trooly audiences to Klaviyo lists. Disconnecting Klaviyo or uninstalling the app revokes our access.
- Google Analytics 4: if the merchant enters a GA4 measurement ID and API secret, we send interview-completion events with the shopper's Google Analytics client ID to the merchant's GA4 property.
We may update this list as providers change; see section 13.
8. Retention and deletion
- While the app is installed, we keep merchant and shopper data for as long as it is needed to provide the app to that merchant, except as stated below.
- Invitation events (shown / clicked / dismissed) are deleted after 45 days; completion events are kept for reporting while the app is installed.
- App pixel browsing events are kept for 13 months. Anonymous profiles that were never linked to a known customer are deleted, with their events, 30 days after they were last seen.
- Interview recordings, transcripts and insights are kept until the merchant deletes them, the shopper's data is erased (below), or the store's data is erased after uninstall.
- Unsubscribe list. Email addresses that unsubscribed are kept while the app is installed so that the person is not contacted again.
- On uninstall, we immediately delete the store's Shopify sessions and Trooly API credentials, revoke and delete any Klaviyo connection, and stop showing invitations. About 48 hours after uninstall Shopify sends us a
shop/redactrequest. We then delete this app's data for the store from our database, and delete the interview projects and knowledge documents that this installation created on the Trooly interview platform (projects that the merchant created elsewhere and merely connected to the app are not touched). If an automatic deletion step cannot be completed, the request is flagged to our team and completed manually. - Customer erasure. When Shopify sends a
customers/redactrequest, we delete that shopper's interviews (including transcripts and recordings) on the Trooly interview platform, then their invitation events, visits, rewards, recommendations, behavior profile and events, AI insight records, email send and unsubscribe records, and remove their identifiers from the remaining interview statistics. Billing usage records are kept with the visitor reference replaced by a random value. - Known limits. Deleting a knowledge document removes it from the interviewer's knowledge, but the originally uploaded file may remain in object storage. Deleted data may persist in encrypted database backups until those backups expire on their automatic rotation cycle. A minimal log of privacy requests (encrypted), a record of which platform resources this installation created, and in-app usage events that contain no personal data are kept as evidence that requests were fulfilled and to operate the service.
9. Privacy requests and your rights
- Merchants can see the status of privacy requests for their store and download customer data exports in the app under Privacy (
/app/privacy), and can contact us at customer_service@trooly.ai for access, correction or deletion of their own data. - Shoppers can ask the merchant, or us at customer_service@trooly.ai, to access or delete their data. Because the merchant is the controller, we may forward your request to the merchant and act on its instructions. Depending on where you live (for example under the GDPR, UK GDPR or California law) you may also have rights to correct, restrict or object to processing, to data portability, and to complain to a data protection authority.
- Shopify privacy requests are handled automatically through Shopify's mandatory webhooks:
customers/data_request(the merchant is notified and can download the export from/app/privacy),customers/redactandshop/redact(section 8). - Email invitations include an unsubscribe link; unsubscribing stops further invitations from that store.
10. Security
- All data in transit is encrypted with TLS. Our databases run on AWS RDS with encryption at rest.
- Credentials and sensitive fields (such as Trooly API keys, Klaviyo tokens, invitation link recipient data, push subscriptions and privacy request payloads) are additionally encrypted with AES-256-GCM.
- Access to production systems is restricted to authorized personnel over private networks. Each store's data is scoped to that store, and application logs record email recipients only as a truncated hash.
- We follow a written security incident response policy. If an incident affects personal data, we will notify affected merchants and Shopify, and authorities where the law requires it, without undue delay.
11. International transfers
Data processed by the app is hosted in the United States (AWS us-east-1). Pikaia is a United States (Delaware) company. Some service providers listed in section 7, and our personnel who support the service, may process data in other countries. Where data protection law requires it, we put appropriate safeguards in place for these transfers.
12. Children
The app is intended for businesses and is not directed to children. Interviews are intended for adult shoppers; if we learn that we have collected personal data from a child without appropriate consent, we will delete it.
13. Changes to this policy
We may update this policy when the app or our providers change. We will post the new version on this page with a new effective date, and take reasonable steps to let merchants know about material changes, for example in the app.
Effective date: September 27, 2026.
14. Contact us
PIKAIA LIMITED (a Delaware company)
8 The Green, Ste A, Dover, Kent County, Delaware 19901, United States
Email: customer_service@trooly.ai