From question
to a clearer answer.

Start in a few minutes.#

1. Sign in and add your brand name and website.
2. Choose the market and language of the answers you want to monitor.
3. Choose suggested questions, write your own or import a CSV. Organize them with topics and tags.
4. Add up to ten competitors. Lastfind also suggests names found in comparable answer contexts; you choose which ones to track. Sentiment reads your brand and the first five competitors.
5. Compare channels and topics, then open any response to inspect its original text and sources.

The public Netflix study shows collected answers only after they arrive. Your workspace shows the responses actually collected for your projects.

Understand what the numbers mean.#

Visibility#

The percentage of valid AI answers mentioning your brand. A mention is not necessarily a recommendation. An absent AI Overview does not count as an answer.

Share of voice#

Your share of mentions among the brands you track. Each brand is counted once per response. This compares your tracked competitors, not the entire market.

Cited and consulted sources#

Cited sources are links referenced in the answer. Consulted sources are search results the provider reports retrieving, including pages it did not cite. Not every channel reports consulted sources; unavailable data is different from an empty list.

Drill into the evidence#

Reports start with all channels. Filter by channel, date, topic, prompt, brand or mention status. Compare with the preceding equal-length period. Click a channel, topic or visibility gap to inspect the matching responses. CSV exports the full filtered period, independently of the visible page.

Reports cover up to 90 days; older records remain stored. Editing tracked brands reinterprets up to the 10,000 newest saved answers in that period without requesting them again. The original answer and collection settings remain preserved.

Free plan and credits.#

Free monitors up to 10 prompts weekly in ChatGPT, in 1 topic. These are the Cloud options available for this launch.

Each saved question occupies one prompt slot. Free collects each prompt once per week, on Mondays at 4 AM Brasília. Pausing preserves history but keeps the prompt slot; archiving releases it.

Your private installation.#

For most people, Cloud is the simplest choice: no installation, API account or maintenance. If you prefer your own infrastructure, the open-source installer guides you through an independent Cloudflare installation.

git clone https://github.com/agencia-conversion/lastfind.git
cd lastfind
npm ci
cp .env.example .env

1. Open .env in a text editor and fill DATAFORSEO_LOGIN and DATAFORSEO_PASSWORD with your account credentials. Save the file and keep it private.
2. In your Cloudflare dashboard, activate R2 if it is not already enabled. Review the charges shown by Cloudflare; it may request a payment method. The installer never activates billing or changes subscriptions.
3. Run setup below, then select your Cloudflare account and follow the installation prompts.

npm run setup

Use Node.js 22.23 or later. Setup checks R2 before creating resources, then configures your control database, private SQLite per project, private archive, secrets and native schedule. Keep the generated access key in your password manager. API and hosting charges belong to your own accounts.

To update, run git pull, npm ci and npm run deploy:selfhost. Your database, access key and data are preserved. The installation guide and architecture are documented in the repository.

Scheduled collection, even with the browser closed.#

New Free prompts enter the next Monday window at 4 AM Brasília (07:00 UTC); independent installations collect daily. Standard batch processing keeps provider costs low. Results arrive as each channel finishes; most scraper tasks complete within 45 minutes, while Claude may take up to 72 hours. Perplexity only offers Live processing.

Cloud and independent installations use native Cloudflare scheduling. The browser only reads results. In your project settings you can pause collection; tasks already submitted still finish and remain in history. Gemini currently requires an English-language project.

Open source. Built by Conversion.#

The Lastfind core uses AGPL-3.0-only. You can study, modify, distribute and use it commercially. If you offer a modified version to others, including over a network, you must make its corresponding source available under AGPL-3.0. Preserve license and copyright notices. The license does not grant rights to Lastfind or Conversion trademarks.

Our production Cloud consumes the open-source core. The independent installer creates a private single-owner workspace; operating a commercial multi-account service requires your own engineering and operations. Read CONTRIBUTING.md and SECURITY.md in the repository to contribute or report a vulnerability.

Commercial license

Conversion exclusively offers a commercial license for use without the AGPL obligations, for example to offer a modified, closed-source version as a web service. Contact comercial@conversion.com.br.