Your business data is not marketing decoration.

TaskChad may process information you submit through a diagnostic, contact form, booking flow, purchase, support request, or approved customer workflow. This can include business details, operating answers, website URLs, optional contact information, attribution tags, appointment details, and payment status. Payment card data is handled by the payment provider and is not placed in the Revenue Leak Score result.

General analytics may record page, campaign, partner, industry, score version, fit tier, and recommended path. TaskChad does not intentionally send raw diagnostic answers, names, email addresses, phone numbers, FAQs, recordings, transcripts, credentials, or provider resource identifiers to general analytics.

Answers are sent to the TaskChad server to calculate the deterministic score. Contact details are optional. A contact email alone does not request a private review; that requires a separate affirmative choice. The public result excludes contact details, raw answers, attribution, and operator notes.

If the page says the submission record could not be stored, the score was calculated but no durable Revenue Leak Score record was created by TaskChad. Infrastructure providers may still keep limited security or request logs under their own operational retention rules.

The TaskChad Revenue Diagnostic is an anonymous, read-only ChatGPT plugin. If you choose to use it, ChatGPT sends exactly nine categorical operating answers to the TaskChad MCP endpoint. TaskChad validates those answers, calculates a deterministic score in memory, returns the score and bounded recommendations to ChatGPT, and discards the answers and result when the request ends. The tool does not request your name, contact details, precise location, website, full chat, customer record, or ChatGPT account identifier.

TaskChad does not store diagnostic inputs or results, create a lead, join the result to another TaskChad record, contact you, sell the data, use it for advertising, or use it to train a model. Application success logs are disabled. If an internal failure occurs, the application logs only a coarse failure class and an ephemeral per-process count, without the answers, result, IP address, user agent, conversation data, or a stable user identifier.

OpenAI and ChatGPT transmit the selected answers and display the result under OpenAI's terms and privacy controls. Vercel hosts the TaskChad endpoint and may process short-lived network and request metadata to route, secure, and troubleshoot it. The production configuration uses a Vercel Pro project without Observability Plus or a third-party log drain; Vercel runtime logs are retained for one day. TaskChad will update this notice before enabling longer retention or another recipient for this endpoint.

You can choose not to invoke the diagnostic, remove or disconnect the plugin in ChatGPT, and manage the conversation using OpenAI's controls. TaskChad has no durable V1 diagnostic record to access, correct, or delete. Contact TaskChad through the support page about TaskChad-held security metadata. Deleting a ChatGPT conversation is governed by OpenAI's controls and policy, not by TaskChad.

The anonymous local composition step runs in the browser and does not create a provider workflow, place a call, book an appointment, or write to a CRM. A Revenue Leak Score handoff carries only validated industry, workflow, and language settings in single-use browser session storage. It excludes business and contact details and is removed when consumed.

Any future browser voice test, recording, transcript, production agent, or connected system requires a separate disclosed and authorized workflow. The current public local builder does not silently activate those features.

When you ask TaskChad to scan a website, TaskChad reads up to twelve public pages on that same HTTPS domain. The scan may extract ordinary business facts, public contact paths, logo or favicon references, and visual brand signals. Pricing, guarantees, testimonials, licenses, awards, regulated advice, unsupported superlatives, login pages, private hosts, and cross-domain redirects are excluded from the instant-production knowledge path. Unpaid scan results expire after twenty-four hours.

An automated language model may normalize the wording of already extracted facts for the approval screen. It is not allowed to invent a fact or source. Public contact paths are excluded from that normalization request. A paid Instant Chat specification is compiled again on the server and every production business fact must match the stored, same-domain scan evidence.

The production chatbot does not store chat transcripts. A short-lived signed session authorizes messages only from the approved website origin. If a visitor chooses the separate lead form, the form requires affirmative disclosure and consent. The visitor's name, contact information, and request bypass the language model, are encrypted for delivery, and are deleted seven days after successful alert delivery or no later than thirty days after a delivery failure.

To evaluate the initial Instant Chat price, TaskChad keeps a private operational record of whether a Stripe checkout was opened, paid, expired, or failed, plus the pricing version, installation choice, amount due, and timestamps. This record excludes the customer's email, business name, website, scan result, and approved knowledge. It may remain after an abandoned builder session is deleted because it is a de-identified commercial measurement rather than a customer workspace.

TaskChad uses information to provide the requested result or service, secure and troubleshoot the site, prevent abuse, maintain customer records, honor an explicit review request, communicate about an active relationship, and improve the diagnostic or delivery system. TaskChad does not sell personal information.

TaskChad keeps a record only while it is reasonably needed for the requested service, an active customer or review workflow, security, accounting, dispute resolution, or a legal obligation. When those purposes end, the record should be deleted or de-identified. Different records can have different retention periods because a diagnostic, booking, payment, and production workflow serve different purposes.

To ask what TaskChad has about you, correct it, withdraw a review request, or request deletion, email pedro.mendoza@taskchad.com with the subject “Privacy request.” TaskChad may need to verify that the request belongs to you before changing or disclosing a record. Some information may be retained when required for security, accounting, dispute resolution, or law.

TaskChad uses infrastructure, database, analytics, email, payment, and approved automation providers to operate the service. They receive only the information needed for their role and operate under their own security and privacy terms. No internet service can promise perfect security, so TaskChad uses access controls, server-only credentials, input validation, and fail-closed product gates to reduce exposure.

Material changes will be reflected on this page with a new update date. Questions about a paid service are also governed by the applicable agreement and the TaskChad terms.