Introduction
WitCloud brings your ads, shop, marketplace, CRM and analytics data together in your own BigQuery, keeps it up to date and turns it into answers - reports, alerts, customer segments and an AI assistant that knows your business.
Updated
Your marketing data already exists - in Google Ads, Meta, TikTok, GA4, your shop and on marketplaces. The problem is that every system keeps its own piece and speaks its own language. To answer a simple question - what did we really earn on marketing last month, per market? - someone has to export six reports, match the names, convert the currencies and add it all up. And next month, do it again.
WitCloud does that work for you, every day, automatically. It collects the data from all your platforms into your own Google BigQuery, unifies it into one consistent model, and builds on top of it what your team actually needs: ready-made tables and reports, KPI trees with alerts, customer segments, audience lists for your ad platforms and an AI assistant that answers questions about your business in plain language.
What you get
All your data in one place - and it stays yours
Integrations connect your ad platforms, shop and marketplace systems, CRM, affiliate networks and Google Analytics 4 - more than 30 sources, from Google Ads and Meta to Shopify, Allegro and Amazon. Each one loads its history and then keeps it up to date; you see and schedule the runs in Workflow.
The data lands in the BigQuery dataset of your own Google Cloud project. Nothing is locked in WitCloud: open it in Looker Studio, Power BI, Google Sheets, your own SQL or any AI tool.
One language for every source
Meta calls ad spend spend, Google calls it cost; every shop names its order statuses differently. All in One turns all of it into one model: ad costs, orders, products, GA4 sessions and attribution paths in ready-made tables, in one currency and timezone. Your channel grouping decides how raw sources and campaigns become the channels and markets your team talks about.
Answers, not just tables
On top of the unified data, Solutions answer the questions you would otherwise ask an analyst:
- Customer Analysis - who your customers are: their value, how often they come back, who is about to leave and which channels bring the best ones.
- Custom Audience - those segments as audience lists on Facebook, TikTok, Google Ads and Klaviyo, refreshed automatically.
- KPI Tree - your KPIs as a tree that shows where a change came from, with alerts when something breaks and goals that tell you mid-month that you are behind.
An AI assistant that understands your data
WitCloud MCP connects Claude, ChatGPT or Gemini to your project. Ask "why did sales drop last week?" and get an answer with real numbers - built on the prepared tables, your documentation and your business context, not on guesses from raw exports.
How it all fits together
Each part of WitCloud solves one step on the way from raw data to a decision - and each step stands on the one before it.
1. One source of truth
Collecting the data is only half the work. The other, more important half is putting it in order, so that answers can be read straight from it. Until then, every report is a project and every number is open to debate - and connecting AI directly to raw systems only moves the problem: the model queries six sources, pulls in thousands of raw rows, mixes up spend and cost, and answers slowly, expensively and inconsistently.
All in One does this work once, before anyone asks: it joins, cleans and aggregates the data in your warehouse, so that reports, alerts, segments and the AI all read the same prepared tables. Having everything side by side also makes problems visible - a source that stopped delivering, revenue that no longer matches your shop, a channel that suddenly disappeared. Data monitoring shows how each of your sources is doing.
2. Structure: a tree, not a list
A dashboard with twenty numbers side by side is a list. When revenue drops, it shows everything at once and leaves you with "but why?". A KPI tree is a map instead: revenue is sessions × conversion rate × average order value, the conversion rate is the steps of your funnel, ROAS is revenue against cost, and cost is clicks × price per click. A movement at the top can always be explained by a branch below it.
3. Watch: let the data tell you
Nobody checks dozens of KPIs across several markets every morning - and the subtle changes are the first to get lost. Alerts and goals watch them for you: once a day they compare every KPI with its own period and speak up only when a threshold is crossed or a goal falls behind its pace. A good set of alerts stays silent most days - so when it speaks, it is worth reading.
4. Diagnose: from the alert to the cause
An alert says what and where. The tree says why: go down from market to channel to campaign, then take the KPI apart - did cost go up or revenue go down? More clicks, or more expensive ones? You can do it yourself in the tree, or ask the AI assistant, which walks the same path - only deeper and faster.
5. Act: turn insight into campaigns
Knowing which customers are about to leave is only useful if they hear from you before they do. Customer Analysis finds them; Custom Audience turns them into audience lists on your ad and e-mail platforms that refresh by themselves - no CSV files, no lists that went stale two weeks ago.
6. Ask: the AI on top of everything
The assistant does not add new data - it opens what you already have to your whole team. Every module you turn on gives it new questions it can answer: All in One brings sales, costs and ROI, the KPI tree brings a path to diagnose along, Customer Analysis brings questions about customers and cohorts. No SQL, no knowing table names - just knowing what to ask.
Define it once, use it everywhere. A segment, a KPI or a channel is defined in one place - and it is the same in the report you look at, the list sent to Meta, the alert in your inbox and the answer in the chat. No three definitions of the same customer in three tools that stop agreeing after six months.
A process, not a project
Data that was analysed once goes stale quickly: customers move between segments every day, stock runs out, campaigns change. The value comes from looking at the changes regularly and reacting to them. The work splits into three layers:
- Data, lists and alerts - WitCloud. Sources load every day, the model is rebuilt, segments are recalculated, audience lists refresh, alerts and goals are checked.
- Campaigns - your marketing team. A segment sells nothing by itself: someone builds the remarketing, the e-mail sequence and the creative on top of it.
- Decisions - you. Which products to push, where to move the budget, what offer to give customers at risk. WitCloud prepares the numbers and the reasoning; the decision stays yours.
What keeps it going is one fixed meeting - weekly or every two weeks - with the people from paid campaigns, e-mail and CRM, and whoever decides on range and promotions. No materials to prepare: the reports are ready and the questions go to the assistant during the meeting. Five questions are enough: What changed since last time? What is off? Where are we losing? What did our last changes deliver? What one thing do we change next?
Why teams choose WitCloud
- No more manual reporting. The data is collected, cleaned and joined every day - your team spends its time on decisions, not on spreadsheets.
- One version of the truth. Marketing, e-commerce and management look at the same numbers, defined the same way.
- Built for many markets. Several countries, shops, ad accounts and currencies end up in one model you can compare side by side.
- Your data, your cloud. Everything is stored in your own Google Cloud project, with access you control.
- Ready for AI. Clean, documented tables are what makes AI answers reliable - WitCloud gives your assistant exactly that.
Who it is for
- E-commerce teams that sell in several markets, shops or marketplaces and want one view of sales and marketing.
- Marketing specialists who need reliable cost, revenue and ROAS figures across platforms.
- Analysts who would rather analyse than build and maintain data pipelines.
- Agencies that manage many accounts and clients and need consolidated data for each.
- Managers who want to ask questions about the business and get answers, without waiting for a report.
Where to start
You do not need perfect data to start - you need to see it in one place, because only then do you know what to fix first. You also do not need every module at once: each one you add widens the range of questions you get answers to.
- Onboarding walks you through the setup step by step: your Google Cloud project, data location, currency and timezone, and your first integrations.
- Connect your integrations - the more sources, the fuller the picture.
- All in One starts with the currency and timezone and builds your unified tables.
- Pick the Solutions you need and connect your AI assistant through WitCloud MCP.
Stuck at any point? Ask the assistant - it reads this documentation and can guide you through the setup.
