Plug your shop
straight into artificial intelligence.

MCP is a direct line between your shop's data and Claude, ChatGPT and Gemini. Connect it once, and your assistant answers questions about sales, ads and products using real numbers.

Your assistant
Claude ChatGPT Gemini
MCP
DataOrganizer
Your shop
sales ads products customers
Create an account and connect your data

This is what it looks like when AI has your numbers

You write in plain language. The assistant pulls the right data and gives a specific answer, live, below.

claude · connected to DataOrganizer live
›

Questions you never had time for

This is how it looks in your assistant: on the left, a question in everyday language; on the right, a specific answer from your data.

See what is behind the numbers

Instead of guessing the cause, simply ask.

Did I really earn less than the month before?
Revenue verified

Contrary to first impressions, no. April was the strongest month this year.

March
€74,111.50
743 orders
April
€81,953.25
839 orders · +10.6%
Where am I actually losing money on sales?
Identified SKUs dragging margin down

Five items are pulling margin down, and one is selling below cost.

ProductProfitMargin
Toscana olive oil−€84−3.0%
Brazil coffee 1 kg€902.3%
Assam tea€93.253.8%
Linden honey€622.509.9%
Where is advertising not returning what I put in?
Calculated ROAS per channel

ROAS by channel (30 days). One has dropped below the break-even threshold.

Google
5.2×
Newsletter
4.4×
TikTok
2.6×
Meta
1.6×
Meta is below the 2.1× threshold. Shift budget to Google to recover ~€1,800 / month.

Spot what no single report shows

Patterns that only emerge when your data is connected.

What do customers add to the basket most often?
Analysed product pairs and cross-sell

Most frequent pairs among returning customers (90 days).

Fruit granolaOat milk11×
Brazil coffee 1 kgV60 filters10×
Assam teaLinden honey9×
Which source brings in the most valuable customers?
Ranked channels by LTV

Average LTV by acquisition source (all time).

Newsletter
€310
Direct
€265.75
Instagram
€245
Google
€190
How much do I really keep from every hundred?
Combined COGS, shipping, fees and advertising

Of every €100 in revenue, €21 is left as net profit.

COGS
Ship.
Fees
Ad
Profit
COGS 42% Shipping 14% Fees 13% Advertising 10% Profit 21%

Test a decision before you make it

Model the scenario on your own numbers, not on gut feeling.

Is it worth raising the free shipping threshold to €49.75?
Modelled the impact on basket and conversion

Yes, with the same traffic you come out ahead.

+€1,600additional profit per month
Now · threshold €37.25 €49.50 basket
After the change · threshold €49.75 €55 basket
Basket +11%, conversion −2 pts, net result still positive.
Where will I end up at the end of the quarter if nothing changes?
Built a forecast from daily sales

Q2 should close at around €265k, in line with the trend.

April
€82k
May
€92.75k
June
€90k
How does this quarter compare with the previous four?
Aggregated quarterly sales

Q1 2026: €219k, the fourth consecutive quarter of growth.

€99.25k
Q2'25
€166.5k
Q3'25
€190.25k
Q4'25
€219k
Q1'26
Growth of +15.1% quarter on quarter, the best in the shop's history.

You talk instead of clicking through reports

A ready-made report ends where the next question begins. Here you simply keep asking: follow up, dig deeper, change the period, without starting from scratch.

Why did my margin drop in May?
Question

Start with a question

Just as you would put it to a colleague. No filters, no hunting for the right report.

Net margin fell from 24% in April to 19% in May, almost entirely due to higher advertising costs in the Dining room category.
Show me which campaigns?
Follow-up

Follow up

"Now show me that by campaign." The answer adapts within the same conversation.

Increase in advertising costs by campaign · May
Dining room · Google
+41%
Dining room · Meta
+33%
Bedroom · Google
+18%
Office · TikTok
+9%
Result

See the result

A clear picture straight away, instead of waiting for an analyst and yet another version of the spreadsheet.

What happens when you ask a question

Your data is already with us, organised. MCP is a secure window through which the assistant asks for it. Raw tables never leave.

Your sources
IdoSell Marketplaces Meta Ads GA4
DataOrganizer
your shop's data, organised
MCP
secure window
Your assistant
Claude · ChatGPT · Gemini
receives a ready answer

Why ask DataOrganizer rather than the shop itself?

IdoSell, WooCommerce or marketplaces only know their own patch: the orders and products on their platform.

They do not know how much went on Google Ads, what shipping really costs or what your margin is after everything. DataOrganizer brings it all together (ads, fees, shipping, costs, returns) and organises it, so whatever you ask, the assistant answers from the full picture, not a fragment.

Sales
Ad spend
Shipping costs
Costs and fees
One organised picture of the business
This is where the assistant draws its answers to your questions from.

46 analytical tools inside

You do not need to know any of them. The assistant picks the right one for your question. Six areas, one engine.

8
Sales and orders
how much you earned and whether that is good for your scale, what pushed sales up or down, returns, deliveries, marketplace profitability after fees
12
Products and range
what is taking off and what is slowing down, the 80/20 rule, what customers buy together, brands and categories, ratings and gaps in descriptions
8
Marketing and advertising
ROAS and profit after ads, campaign results by platform, Google keywords burning budget, coupon profitability
9
Google
your prices vs the market median, feed health, which bestsellers you are missing, search positions and the traffic funnel from entry to purchase
5
Customers and loyalty
what a customer is really worth (LTV), whether they return and how quickly, whether their basket stays the same, order value distribution
4
Data quality
what deviates from your norm, whether the numbers can be trusted before you base a decision on them, and a clear glossary of every metric
All 46 tools
The assistant picks the right one for your question. You just ask.
Sales and orders
Sales summary
Net revenue, orders and discounts, with an assessment of whether that is a lot for your scale.
Quick KPI overview
Key business metrics in one place.
Breakdown of "why it fell / rose"
Whether the change came from volume, price, mix, or new and lost products.
Order health
Fulfilled, cancelled and returned orders: return and cancellation rates.
Delivery and payments
Which methods are used, shipping costs, the impact of the free delivery threshold.
Marketplace profitability
Fees and commissions per listing: where they eat into margin.
Margins
How much of your sales is high-, medium- and low-margin.
Additional costs
Manual costs (influencers, events) by your own categories.
Products and range
Bestsellers
Top products by revenue, quantity or orders, with their share of turnover.
What is growing and what is slowing
Products with significant growth or decline, without the noise.
The 80/20 rule
Which products generate most of the revenue and which form the long tail.
What they buy together
Product and category pairs that often end up in the same basket.
Brands
Which brands drive sales.
Categories
Results by category, including comparisons between markets.
Attributes
Which colours, sizes and collections sell best.
Words in names
Hidden patterns in product names.
Gaps in descriptions
Profitable products with missing data (colour, size, EAN…).
Product search
Find by name, SKU, EAN, brand or collection.
Allowed filter values
Suggests what you can filter by (brands, categories…).
Marketplace ratings
Average rating and star distribution of listings.
Marketing and advertising
Ad effectiveness
Overall ROAS, cost of sale and profit after ads.
Campaign results
Drill down by platform, campaign and ad set.
Coupon profitability
Whether a discount attracts new customers or eats into margin.
Google Ads keywords
Which keywords burn budget and which deserve a higher bid.
Google Shopping (paid)
Product ad results per product from the feed.
Landing pages
Which page visitors start on and how they move through the funnel.
Session funnel
Where customers drop off overall: basket, checkout, payment.
Product funnel
Where a specific product loses customers: listing or checkout.
Google: Shopping, Merchant, Search
Feed health
How many products are approved, how many rejected, and why.
Google Shopping
Product ad results as measured by Google.
Free Google listings
Traffic from organic Shopping, an often overlooked channel.
Prices vs the market
Your price against the Google median: where you are more expensive.
Shopping competitors
Who competes with you and how often they rank higher.
Bestseller gaps
Proven hits in your category that you do not stock.
Visibility trend
Whether a drop in visibility is down to you or the whole industry.
Google Search (organic)
Clicks, impressions and keyword positions.
Traffic sources
Where visits and sessions come from.
Customers and loyalty
Customer segments
New vs returning, VIPs, return cadence.
Customer value (LTV)
How much a customer really spends over their lifetime.
Cohort returns
Whether customers come back and whether retention is improving.
Basket evolution
Whether returning customers buy the same things or their purchases change.
Order value distribution
How baskets are split and where high revenue comes from.
Data quality
Anomalies
Days that deviate from your norm for that day of the week.
Trust gate
Whether the data is complete before you base a decision on it.
Metrics glossary
What each number means exactly (e.g. net vs ROAS).
What is connected
List of sources and integrations, plus data freshness.

Get started in three steps

No installation and no code. A bit like adding an app to the assistant you already use.

01

Copy the address

In DataOrganizer, under Data sharing, there is a single server address.

02

Paste it into your assistant

Claude, ChatGPT or Gemini. Add the address as a new connection.

03

Log in and ask

Confirm with your account and choose your company. Done, the engine starts.

Server address https://mcp.dataorganizer.io/mcp

The assistant does not learn from it

It receives only the data needed to answer. It does not build its own memory from it or train a model on it.

Only the finished answer

The assistant receives the result of a specific question, never access to raw tables.

Disconnect whenever you like

You can see what is connected and when it last connected. One click and it is over.

How shops use it

Briefly, in their own words: what DataOrganizer MCP is useful for day to day.

"I no longer wait for the Monday report. Over my morning coffee I ask Claude and instantly know which campaign is underperforming."

MK
Marta K.
Clothing shop owner · IdoSell

"I run three shops across marketplaces and WooCommerce. At last I ask one question about all of them and get one consistent answer."

TŻ
Tomasz Ż.
E-commerce manager · Marketplaces + Woo

"I never had net margin after all costs to hand. Now it is my first question every morning."

AN
Anna N.
DTC brand founder · PrestaShop

Frequently asked questions about MCP

Start the engine on your own data

One connection and your assistant knows your shop. Works with Claude, ChatGPT and Gemini.

Create an account and connect your data
mcp.dataorganizer.io/mcp