Data to AI™ / Gemini/Google Ads
Pull Google Ads report data into scripts, notebooks and pull requests from the terminal, so tracking and feed problems surface before a release instead of in next month's report.
Better answers, fewer tokens.
Gemini CLI
Which search campaigns lost the most impression share to budget last month?
Competitor terms lost 44% of its eligible search impressions to budget and Generic - Running shoes lost 31%. Brand - Exact lost only 2%.
| Campaign - Name | Cost Micros (Spend) | Search Impression Share | Search Budget Lost Impression Share |
|---|---|---|---|
| Competitor terms Campaign Report | 5,410 | 38% | 44% |
| Generic - Running shoes Campaign Report | 12,880 | 54% | 31% |
| Brand - Exact Campaign Report | 3,240 | 91% | 2% |
Cost Micros (Spend) was divided by 1,000,000 to match the amounts in the Google Ads interface.
Google Ads data comes as Reports with Metrics, Reports without Metrics and Keyword Planner Historical Metrics. You pick a report type such as campaign, keyword, search term or shopping, then the attributes, segments and metrics to extract.
Using the dataddo server, list each Campaign - Name from the Campaign Report for the last 7 days where Conversions is above 0 but Conversions Value is 0.
A developer who shipped a checkout change runs this the next morning. An empty list means conversion values still arrive; any row goes into the bug ticket.
From Shopping Performance View, find every Item ID with Conversions Value this month but an empty Brand or Product Type Level 1, and write them to feeds/missing-attributes.csv.
The engineer who maintains the product feed generator opens the CSV in Gemini Code Assist in VS Code, fixes the mapping and links the file in the pull request.
Query Keyword Planner Historical Metrics for our seed list and save Keyword, Avg Monthly Searches, Competition and both top of page bids, divided by 1,000,000, to data/keyword-ideas.parquet.
A technical marketer clusters the ideas in a notebook next to the repo and proposes new ad groups, with no manual Keyword Planner export.
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 | |
|---|---|---|---|
| Ad hoc pulls from code | Every one-off check needs its own Google Ads API client and query, so most checks never get written. | Gemini CLI queries the extracted reports by entity and field name through one MCP server, and no Google Ads API code lives in the repo. | Checks without API plumbing |
| Micros in generated files | A script copies Cost Micros (Spend) into a CSV as is, and a stakeholder later reads spend a million times too high. | The field carries its description from the semantic layer, so Gemini can divide micros by 1,000,000 before writing the file. | Files in real currency |
| Control over each call | A script with stored API credentials can pull any report at any time, and nobody reviews what it requested. | Gemini CLI asks you to confirm each Dataddo tool call, and you decide per tool whether to choose Always allow. | Every lookup visible |
| Project-scoped setup | Each teammate wires up their own Google Ads access and the scripts in the repo assume different accounts. | A .gemini/settings.json in the project holds the Dataddo server, and each person signs in with their own Dataddo account. | Same setup for the team |
| Token consumption | Gemini spends tokens exploring raw Google Ads 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 Google Ads 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
|
|---|---|---|
| Campaign Report |
|
Which Campaign - Name values had Conversions last week but a Conversions Value of zero? |
| Campaign Budget Report |
|
What is the total Campaign Budget - Amount Micros across all active budgets, converted to currency? |
| Keyword View Reports |
|
Which Ad Group Criterion - Keyword - Text values run in more than one Ad Group Criterion - Keyword - Match Type? |
| Search Term View Report |
|
Which Search Term View - Search Term values contain our brand name but matched under a non-exact Search Term Match Type? |
| Shopping Performance View |
|
Which Item ID values have Conversions Value this month but an empty Brand or Product Type Level 1? |
| Ad Group Ad Reports |
|
How many Ad Group Ad - Ad - Responsive Search Ad - Headlines does each ad carry, and does the count track Ad Group Ad - Ad Strength? |
| Keyword Planner Historical Metrics |
|
Which Keyword ideas have more than 5,000 Avg Monthly Searches and a High Top Of Page Bid Micros below 2 in currency? |
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 →Extract Google Ads reports such as the Campaign Report in Dataddo and attach them 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.
No. Dataddo extracts the Google Ads reports into a flow, and Gemini CLI reads that flow through the Dataddo MCP server. The setup is the same as for any other source in Dataddo.
Use the httpUrl key, not url, because url falls back to SSE, which Dataddo does not support. Leave out headers and oauth blocks, restart Gemini CLI and check /mcp.
Yes. Agent mode in VS Code reads the same settings.json as Gemini CLI. In IntelliJ, add the Dataddo server to the mcp.json of Gemini Code Assist.
Every flow or Context attached to an AI destination or AI Model in your Dataddo account, not only the Google Gemini ones. Keep unrelated Google Ads accounts out of AI destinations if Gemini should not see them.
Google reports Performance Max only at campaign level. Add a Campaign Report source to the flow so Gemini can include those campaigns.
No. Gemini queries extracted data through the semantic layer by entity and field name. Dataddo never accepts SQL from the model and nothing is written back to Google Ads.
