Data to AI™ / Copilot/Google Analytics 4
Let content, regional and product teams ask a Copilot agent in Teams about website traffic, page views and new users instead of waiting for the analytics team.
Less tokens, better answers.
Microsoft 365 Copilot
Which channel brought the most key events for its traffic last month?
Organic Search brought the most Key Events, 1,446. Email converted best, with 7 key events per 100 sessions against 5 for Paid Search and 3 for Organic Search.
| Channel | Sessions | Key Events | Engagement Rate |
|---|---|---|---|
| Organic Search google / organic | 48,200 | 1,446 | 61% |
| Paid Search google / cpc | 21,500 | 1,075 | 58% |
| Email newsletter / email | 6,400 | 448 | 72% |
Sessions are taken by channel without a Page Path breakdown, so no session is counted twice.
Google Analytics 4 measures how people reach and use your website and apps. Dataddo extracts one flexible GA4 dataset: pick from 184 standard metrics and 374 dimensions, plus custom ones defined in your property.
@Web Traffic Which ten Page Path values gained the most Screen Page Views this week, and which of them were not in last week's top ten?
An editorial team mentions the agent in its Teams channel, so the planning meeting starts from GA4 figures rather than a screenshot someone took on Friday.
@Web Traffic Show New Users and Sessions for my Country by week this month, and tell me how they compare with last month.
Country managers ask the agent in Microsoft 365 Copilot while writing their monthly update and paste the answer into the document they already have open.
Use the dataddo tools to compare Event Count by Event Name before and after today's deploy, and list any event that stopped firing.
A developer runs GitHub Copilot in VS Code in Agent mode and catches a renamed or missing event before marketing finds a hole in its reports.
Data to AI™ is free while in private pre-release. We onboard teams in waves and shape the product with their feedback. At launch, it will be priced like the AI assistants your team already uses. Need tailored contracts or volume pricing? Talk to us about Enterprise.
Serve governed business data to Claude, ChatGPT and other AI tools over MCP.
Free during pre-release
Flexibly deploy Dataddo in any cloud or hybrid environment.
We have a payment model that works for you
| Without Dataddo |
With Dataddo
|
Outcome for you | |
|---|---|---|---|
| Where people ask | Editors and country managers email the analytics team for page views or new users and wait days for a report to come back. | A Copilot Studio agent in Teams and Microsoft 365 Copilot answers from the GA4 flow whenever someone mentions it with @. | Answers inside Teams |
| Breakdowns that look short | A country or gender split in an export adds up to less than all users, and someone patches the gap by hand. | The agent reads rows as GA4 returns them, where users Google cannot classify are left out of the breakdown, so the gap can be explained. | Gaps explained, not patched |
| Several websites or apps | Each GA4 property is exported on its own, and merged rows lose track of which site they belong to. | Multi-account extraction, enabled by the Dataddo Solutions team, pulls the same fields from every property, and Include property ID labels each row. | Every row tied to a property |
| Developers checking tracking | After a release, developers ask someone with GA4 access to confirm that events still fire, often a day later. | GitHub Copilot in VS Code queries Event Count by Event Name through the server in .vscode/mcp.json, in Agent mode. | Tracking checked from the editor |
| Token consumption | Copilot spends tokens exploring raw Google Analytics 4 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. | Less 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 Analytics 4 datasets marketing teams use most, with their real field names. Copilot queries them by name through the Dataddo semantic layer.
Runs on your schedule with a relative date range. Google needs 24 to 72 hours to process data, so the last 3 days can still change, and the GA4 API allows up to 1 year of history.
Google Analytics 4 connector →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 →Set up a Google Analytics 4 source in Dataddo with fields such as Page Path, Screen Page Views and New Users, attach it to a Copilot destination or AI Model, then add https://headless.dataddo.com/mcp-data as a Model Context Protocol tool in a Copilot Studio agent with OAuth and publish it to Microsoft 365 Copilot and Teams.
The server description you enter in Copilot Studio tells the agent when to call Dataddo. Describe it as the source for website traffic and conversions.
Yes. After publishing, open Channels > Teams and Microsoft Copilot and add the channel. People in Teams then mention the agent and get answers from the same GA4 data.
Yes. Add the Dataddo server to .vscode/mcp.json, start it, sign in and use Copilot Chat in Agent mode. On Business and Enterprise an org admin must enable the MCP servers in Copilot policy.
Google leaves out users it cannot classify by gender or country, without adding an unknown row. Breakdown totals can therefore be lower than the overall figure Copilot shows.
Yes, through multi-account extraction set up with the Dataddo Solutions team. Turn on Include property ID so the agent can tell which property each row belongs to.
Google takes 24 to 72 hours to process GA4 data. Data appears after the flow runs, and the most recent three days may still change once Google finishes processing.