Data to AI™ / Claude/Microsoft Ads
Ask Claude which search terms waste budget, where impression share is lost and which ad copy earns clicks, using Microsoft Advertising reports that refresh on schedule.
Better answers, fewer tokens.
Claude
Which campaigns lost the most impressions to budget last month, and were they profitable?
Brand Search lost 38% of eligible impressions to budget while returning 6.2 per unit of spend, so it is the clearest case for more budget. Generic Search lost less and returned far less.
| Campaign Name | Spend | Impression Lost To Budget Percent | Return On Ad Spend |
|---|---|---|---|
| Brand Search Campaign Performance Report | $4,100 | 38% | 6.2 |
| Generic Search Campaign Performance Report | $9,800 | 21% | 1.9 |
| Performance Max Campaign Performance Report | $6,300 | 12% | 3.4 |
All three columns come from one Campaign Performance Report with Monthly time aggregation, so the shares are not averaged from daily rows.
Microsoft Advertising (formerly Bing Ads) runs paid search campaigns across Bing and Microsoft's partner network. Dataddo extracts its reports, from account and campaign down to keyword, search term and ad, with the attributes and metrics you pick for each report type.
In the Search Query Performance Report for the last 30 days, list every Search Query with Spend above $50 and zero Conversions, grouped by the Keyword that triggered it.
Claude groups wasted search terms under the keywords that matched them, so the search specialist can add negatives at the right level instead of scrolling through thousands of rows in the Microsoft Advertising UI.
From the Campaign Performance Report, compare Impression Lost To Budget Percent and Impression Lost To Rank Agg Percent for each Campaign Name, with Return On Ad Spend next to them.
Claude separates campaigns that are starved of budget from those that lose auctions on rank, so the budget request goes to profitable campaigns and the rest get bid or Quality Score work first.
Using the Ad Performance Report, which Title Part 1 and Headline phrases appear in ads with the best Ctr and Conversion Rate? Draft five new headline variants in that style.
Claude reads ad text next to delivery results and writes new variants grounded in what already works, so the next round of responsive search ads starts from evidence rather than a blank page.
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 | |
|---|---|---|---|
| Search term cleanup | The specialist downloads the search terms report, filters it in a spreadsheet and copies a sample into Claude, losing the link to the triggering keyword. | Claude queries Search Query and Keyword together from the scheduled Search Query Performance Report and ranks wasted Spend by keyword. | Negatives added where they belong |
| Impression share diagnosis | Lost-to-budget and lost-to-rank columns sit in different UI views, so the cause of missing impressions is guessed rather than measured. | Both shares come from one Campaign Performance Report, so Claude can compare them per campaign alongside Spend and Return On Ad Spend. | Budget asks backed by data |
| Ad text in context | Ad copy lives in the editor and results in a report, so writers rarely see which phrases actually earned clicks. | The Ad Performance Report holds Ad Title, Title Part 1, Headline and Ad Description next to Ctr and Conversion Rate, ready for Claude to compare. | Copy briefs from evidence |
| Analysis in a repository | An analyst rebuilds Bing Ads export cleanup in a notebook whenever the selected columns change. | Claude Code queries the same Microsoft Ads data through the dataddo server, so scripts use field names from the semantic layer. | Repeatable search analysis |
| Token consumption | Claude spends tokens exploring raw Microsoft 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 Microsoft Ads datasets marketing teams use most, with their real field names. Claude queries them by name through the Dataddo semantic layer.
| Dataset | Metrics and attributes |
Ask Claude
|
|---|---|---|
| Campaign Performance Report |
|
Which campaigns had the highest Return On Ad Spend last month, and how much Spend did each take? |
| Keyword Performance Report |
|
Which keywords with a Quality Score below 5 spent the most last month? |
| Search Query Performance Report |
|
Which Search Query terms cost the most last month without a single conversion? |
| Ad Performance Report |
|
Which Title Part 1 variants have the best Ctr in our search ads? |
| Budget Summary Report |
|
Which campaigns have used the largest share of their Monthly Budget so far this month? |
| Professional Demographics Audience Report |
|
Which Industry Name groups bring the most Conversions for our B2B campaigns? |
| Geographic Performance Report |
|
Which Country markets return the best Return On Ad Spend? |
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 Microsoft Ads source in Dataddo, pick a report type such as Search Query Performance Report with its attributes and metrics, and attach it to a Claude destination or AI Model. In Claude, open the connectors directory, add Dataddo Data Access and sign in; Claude Code signed in with your Claude account gets it too.
No. Claude reads Microsoft Ads data that Dataddo has already extracted, through the semantic layer. Dataddo never accepts SQL from the model and nothing is written back, so keywords, bids and budgets change only in Microsoft Advertising.
Performance Max data is only available in the Campaign Performance Report. Create that report type with Base Campaign Id, Campaign Id, Campaign Labels, Campaign Name, Campaign Status and Campaign Type selected, then attach it to Claude.
The type set for your login in Microsoft Advertising, either Work or school or Personal. If you pick the wrong one, the setup test fails with the AADSTS650052 error. Claude never sees these credentials.
Yes. Each report type is its own Microsoft Ads source with its own attributes and metrics. Attach the campaign, keyword and search query sources to your Claude destination or AI Model and Claude can query each of them by name.
Yes. Signed in with your Claude account, Claude Code loads the Dataddo connector automatically; check it with /mcp. With an API key, add the server with claude mcp add and sign in with claude mcp login dataddo.
Yes. Claude lists the connector's tool permissions, and for each Dataddo tool you choose whether Claude may use it freely or must ask you first. On Team and Enterprise each member signs in with their own Dataddo account.
