Data to AI™ / Gemini/Similarweb
Pull competitor traffic and engagement from Similarweb into the terminal, where analysts and marketing engineers turn them into scripts, notebooks and checked-in reports.
Better answers, fewer tokens.
Gemini CLI
How did our traffic and engagement compare with our two closest competitors last month?
Last month ourshop.com had 1.42M visits, behind rivalstore.com at 2.05M but ahead of shopnext.com. Our Bounce Rate was the lowest of the three.
| Domain | Visits | Bounce Rate | Pages per Visit |
|---|---|---|---|
| rivalstore.com Monthly Visit Data | 2,050,000 | 48.1% | 5.2 |
| ourshop.com Monthly Visit Data | 1,420,000 | 41.7% | 5.9 |
| shopnext.com Monthly Visit Data | 980,000 | 52.3% | 4.4 |
Monthly Visit Data has one row per Country, so Visits are summed across countries while Bounce Rate and Pages per Visit are weighted by Visits.
Similarweb estimates website traffic and engagement for your own domain and for competitors. Dataddo extracts visits, bounce rate, pages per visit, visit duration, country split and traffic sources per domain, so marketing can compare sites in plain language.
Read weekly Visits - all traffic and Bounce rate - all traffic for every tracked Domain over the last eight weeks and write data/similarweb/benchmark.csv with one row per Domain and week.
A marketing analyst runs Gemini CLI each Monday in the reporting repo, reviews the diff and commits it, so the BI job picks up a file with a clear history.
Create notebooks/competitor-mobile.ipynb that charts mobile web Visits by Source Type for our Domain and two competitors, using data from the dataddo server.
A growth engineer gets a working notebook in a few minutes and adds the commentary before the strategy meeting, instead of starting from an empty file.
Before I deploy the market report site, compare Pages Per Visit and Average Visit Duration in report/data.json with the latest Similarweb values and list any mismatches.
A technical marketer catches stale numbers in the static report before it goes live, since Gemini asks to confirm each Dataddo tool call along the way.
Data to AI™ is priced per seat, like the AI assistants your team already uses, with ~25% off on annual billing. Need tailored contracts or volume pricing? Talk to us about Enterprise.
Serve governed business data to Claude, ChatGPT and other AI tools over MCP.
$10.49/month
per seat
3 seats minimum
Flexibly deploy Dataddo in any cloud or hybrid environment.
Custom
We have a payment model that works for you
| Without Dataddo |
With Dataddo
|
Outcome for you | |
|---|---|---|---|
| Data in the repository | Benchmarks arrive as spreadsheets in email, get copied into the repo by hand and drift from the source. | Gemini CLI queries the Similarweb flow and writes CSV or JSON straight into the working tree, ready to review and commit. | Versioned benchmark files |
| Field names in code | Scripts guess at column names from an old export and break when someone renames a header. | Gemini reads the real fields, such as Source Type, Organic Visits and Paid Visits, before it writes the code. | Code that matches the data |
| Editor and terminal | The analyst switches between the browser, Similarweb and the editor to move numbers into a notebook. | Gemini Code Assist in VS Code agent mode reads the same settings.json, so the server is available in the editor too. | One setup, two places |
| Token consumption | Gemini spends tokens exploring raw Similarweb columns, sampling rows and retrying until it understands the data. | Field definitions and relations are supplied up front, so the model skips the exploration and goes straight to the answer. | Fewer tokens per answer |
| Answer quality | Prone to wrong joins, invented fields and numbers that are hard to check. | Answers grounded in defined fields. In the Dataddo benchmark, 78.5% of questions were answered correctly, against 65.5% with plain CSV files. | Better answers |
The Similarweb datasets marketing teams use most, with their real field names. Gemini queries them by name through the Dataddo semantic layer.
| Dataset | Metrics and attributes |
Ask Gemini
|
|---|---|---|
| Visits - all traffic |
|
Which Date shows the sharpest drop in Visits for each Domain, so we can check it against releases? |
| Monthly Visit Data |
|
Which countries combine high Visits with a high Bounce Rate on our Domain? |
| Average visit duration - all traffic |
|
Which Domain keeps visitors longest by Average Visit Duration in the selected Country? |
| Bounce rate - all traffic |
|
What is the gap in Bounce Rate between our Domain and the closest competitor? |
| Pages per visit - all traffic |
|
How stable is daily Pages Per Visit for each Domain we track? |
| Traffic Sources Overview - Desktop |
|
Which Source Type grew Paid Visits fastest for a competitor Domain last month? |
| Traffic Sources Overview - Mobile Web |
|
How does the mobile web Source Type mix of Visits differ between us and a competitor Domain? |
We asked 43 pre-registered questions of three real datasets - Google Search Console, HubSpot deals and Google Ads - and delivered the same rows to the same model in four ways. Only the data layer changed.
of questions answered correctly - the best of the four delivery methods tested
fewer confidently wrong answers - 14.0% of runs against 24.0% and 26.3%
more correct answers on messy CRM data - better on 8 questions, worse on none
Same model, prompts and rows in every condition; 759 graded runs, ground truth frozen before any run. On Google Search Console alone, documented CSV files did as well as Data to AI™ - the advantage comes from messier data such as HubSpot deals.
Read the benchmark →Create a Similarweb source in Dataddo with your API key and attach the flow to a Google Gemini destination or AI Model. Add dataddo under mcpServers in ~/.gemini/settings.json with httpUrl https://headless.dataddo.com/mcp-data, restart Gemini CLI, run /mcp auth dataddo and sign in.
In Gemini CLI, url falls back to SSE, which Dataddo does not support. Add the server under mcpServers with the httpUrl key and no headers or oauth block, then restart Gemini CLI.
Run /mcp auth dataddo. Gemini CLI opens a sign-in page in a browser on the same machine, where you use your Dataddo email and password, with MFA if prompted. Then check /mcp.
No. Agent mode in VS Code reads the same settings.json as Gemini CLI, while Gemini Code Assist in IntelliJ uses its own mcp.json, where you add the Dataddo server separately.
Once signed in, Gemini reaches every flow or Context attached to an AI destination or AI Model in that Dataddo account, not only the Google Gemini ones. Attach only the Similarweb flows you want it to query.
Yes. Create one Similarweb source per domain, each with a single domain such as example.com. Gemini can then compare them in one query, because every row carries the Domain field.
No. Gemini asks for data by entity and field name through the semantic layer. Dataddo never accepts SQL from the model, and nothing is written back to Similarweb.
