MCP App Store
by KeenEthics
IdealData.io: Real-time signals for eCommerce Stores icon

IdealData.io: Real-time signals for eCommerce Stores MCP Server for Claude and ChatGPT

By IdealData.io

Overview

The IdealData connector brings an online store's live operating picture directly to Claude. It surfaces the store's active signals along with the data behind them: orders, products and per-warehouse inventory, customers and their lifetime value, active and abandoned carts, traffic sources, search performance, and headline business figures for any date range. Every tool is read-only, so nothing in a store is created, changed, or deleted. IdealData monitors a connected store around the clock and raises a signal when something needs a person. Six signal types are live: a repeat buyer who passed their expected reorder date without ordering, a customer whose behavior indicates they are in a buying window now, a product heading for a stockout with supplier lead time and a decide-by date attached, a pricing error such as a product listed at zero or far outside its margin target, a high-value cart left behind, and a cart abandoned by a known high-value customer. Each signal arrives with what triggered it, the value at stake, and a recommended next step. The connector reads across every system a merchant has connected: their store platform (Adobe Commerce, Shopify, BigCommerce, WooCommerce), inventory and fulfillment, customer records, call activity, and site and search performance. A sales rep preparing for a call can see an account's full purchase history alongside what they have browsed and who they have spoken to since. Weekly planning gets the reorders worth chasing and the products that need a purchase order now. Month-end revenue traces back to the traffic and search queries that produced it. None of it requires opening four dashboards, exporting reports, and stitching the results together in a spreadsheet. Requires an IdealData account with at least one connected store. A 15-day free trial is available with no credit card.

How to connect

  1. Claude: Open the IdealData.io: Real-time signals for eCommerce Stores listing in the Claude connectors directory and choose Connect. Claude listing IdealData.io: Real-time signals for eCommerce Stores exposes 15 tools here.
  2. ChatGPT: Open the IdealData.io: Real-time signals for eCommerce Stores listing in the ChatGPT app directory and choose Connect. ChatGPT listing IdealData.io: Real-time signals for eCommerce Stores exposes 26 tools here.

Tools

get_alert

ChatGPT
Full detail for a single alert by id, including what triggered it and its current status.

get_business_health

ChatGPT
A complete health report for one period in a single call: revenue, order count, average order value and abandoned-cart value, each compared against the equivalent earlier period, plus every open alert broken down by type. Prefer this over several get_business_kpi calls when asked how the business is doing. Figures cover orders placed in the period; open alerts are every alert unresolved right now, whenever it was raised. The result states today's date — if it shows the period is not the one the user meant, call again with the right range rather than describing a different period than the card shows.

get_business_kpi

ChatGPT
One headline business figure for a date range: revenue, average order value, missed cart opportunity, or one of the open-signal impact figures. The signals-* keys describe OPEN (unresolved) signal events, not historical totals.

get_call_transcript

ChatGPT
The full transcript of one of a customer's calls, line by line, with the reps named. Takes the customer id and the callId printed on a call row of get_customer_journey (only calls marked with a transcript have one).

get_cart

ChatGPT
Full detail for a single shopping cart: the shopper and how to reach them, the cart's status and value, and its line items — flagging any that are low on stock or carry a high return rate. On a cart that was closed or converted it also says, per line, what the shopper actually bought.

get_company

ChatGPT
One B2B company (business account) in full: identity, which platforms it came from, live order metrics (lifetime and trailing-12-month revenue, average order value, ordering cadence, predicted next order), the products it usually orders, and its most recent orders. Companies are buying organizations the merchant sells to — for a person, use get_customer.

get_customer

ChatGPT
Full detail for a single customer: contact details and postal addresses, revenue and ordering metrics, open alerts, and recent/top orders. Includes the businessId needed by get_customer_journey.

get_customer_journey

ChatGPT
The customer's activity timeline since their last order: site visits, cart activity, calls (who handled them, the summary, the voicemail), emails (who wrote, the subject, the opening lines) and texts (the words), plus their predicted next order date. Requires the customer's businessId (from get_customer). Read a whole email with get_email and a call transcript with get_call_transcript, using the ids on the rows.

get_email

ChatGPT
One email with a customer, in full (what was written that time, earlier quoted mail cut), or with wholeConversation the whole thread, oldest first. Takes the customer id and the emailId printed on an email row of get_customer_journey.

get_my_day

ChatGPT
The asking person's own to-do list for the next 24 hours: who to call, text or email, in order, with the time, the reason, what to ask and the minutes it takes. Use for 'who should I call today', 'what's on my plate', 'my day', 'what should I do next'. Only the customers this person owns — never a colleague's.

get_my_day_progress

ChatGPT
Whether a call, text or email with one customer has been recorded since a time — used by the My Day card to move on after the rep calls. Answers only what was seen; "not seen yet" is not "it did not happen".

get_net_spend_by_account

ChatGPT
What each ACCOUNT spent in a date range, with shipping, tax, refunds and chargebacks taken out: one row per B2B company (every buyer under it added up), or per customer when they have no company, with the company name, the account email and the net spend. Use it for "total spend per customer between two dates minus shipping and tax and returns", for working out credits or rebates, or for ranking accounts by what they really spent. Cancelled, voided/failed/expired and test orders are left out. IMPORTANT: no OTHER fee is taken out (an anodizing, finishing or handling fee that is not shipping or tax is still inside the figure) — say so to the person asking, and never claim it was removed. Both dates are required and inclusive, in the organization's own time zone; ask for them if the person did not give both. One account in two currencies is two rows and must never be added together. Results are biggest-first and paged: keep passing the returned cursor until none is returned so no account is missed. For one order use get_order; for lifetime or trailing-12-month revenue use search_customers or search_companies.

get_order

ChatGPT
Full detail for a single order: the customer, where it ships and who pays, line items, fulfillment, refunds and financials, the customer's history with the store, and how this order compares to their average.

get_product

ChatGPT
Full detail for a single product: pricing, per-warehouse inventory and bins, its sales over the last 12 months (revenue, units, trend, segment), and its stock outlook — forecast sell rate, the day it runs out, the last day to order and what is on order — computed exactly as the Low Stock alert computes it. Use the outlook for any "will it run out / how long will stock last" question; do not work out days of stock yourself.

get_purchase_order

ChatGPT
One purchase order in full: vendor, status, dates, and every line — SKU, quantity ordered, unit cost, expected delivery and what has already been received.

get_seo_summary

ChatGPT
Search-performance summary: impressions, clicks, CTR, average position and sessions with month-over-month and year-over-year movement, what is working, and CTR opportunities. Impressions/clicks/position describe organic search; the channel filter applies to sessions only.

get_traffic_sources

ChatGPT
Where site traffic comes from, per source: sessions, share, orders and revenue attributed to that source, and (for organic search sources) impressions and clicks. Matches the Traffic Sources page.

list_alert_types

ChatGPT
The alert types that actually have events in this organization. Use this to discover valid signalCode values before filtering with search_alerts.

list_brands

ChatGPT
The organization's brands (businesses), with their connected integrations. Use a brand's businessId to scope other tools to that brand.

search_alerts

ChatGPT
Search this organization's alerts (signal events) — abandoned carts, low stock, late purchase orders, pricing issues, re-engagement and second-order opportunities, follow-ups and shoppers on the site now. Returns the most recent matches. Narrow with signalCode when the question is about one kind — for "will anything run out?" ask for low_stock_warning and late_po. Use get_alert for the full detail of one.

search_carts

ChatGPT
Search shopping carts — active, abandoned, recovered or converted — with the shopper and cart value. "needsAttention" returns the abandoned/active carts most worth acting on. Returns compact rows — use get_cart with the cartId for one cart's line items and full detail.

search_companies

ChatGPT
B2B COMPANIES — buying organizations (business accounts) the merchant sells to, unified across HubSpot and storefront B2B. NOT individual customers (use search_customers for people) and NOT the merchant's own brands (that is businessId / list_brands). Rows carry lifetime and trailing-12-month revenue, order count, linked people and last order date. Use get_company for one account in full. Note: a customer whose record shows a free-text company NAME may have no company record here — only real B2B accounts from a connected source become companies.

search_customers

ChatGPT
Search customers by name or email, rank them by a metric (lifetime value, order count, recency, churn risk, …), or both at once — e.g. top customers by lifetime value matching a name. Filters by customer type, lifecycle state and last-order date range combine freely with the ranking.

search_orders

ChatGPT
Search orders: by order number, customer email or last name; filter by payment/fulfillment status, attention flags, tags, date range and LINE-ITEM COUNT. Reports how many orders match in total, so it answers "how many orders …" questions without paging through them. Returns compact rows — use get_order for one order's full detail.

search_products

ChatGPT
Search the product catalog by name or SKU; filter by performance segment or product type. Rows carry revenue and velocity metrics. Use get_product for one product's full detail.

search_purchase_orders

ChatGPT
Purchase orders the merchant has placed WITH VENDORS to restock inventory — not customer orders. Answers "what is on order", "what is late", "what is stuck waiting for approval". Rows carry the vendor, status, expected delivery date and outstanding units. Use get_purchase_order for the line items.

get_alert

Claude

get_business_kpi

Claude

get_customer

Claude

get_customer_journey

Claude

get_order

Claude

get_product

Claude

get_seo_summary

Claude

get_traffic_sources

Claude

list_alert_types

Claude

list_brands

Claude

search_alerts

Claude

search_carts

Claude

search_customers

Claude

search_orders

Claude

search_products

Claude

Similar Apps

Listed on MCP App Store

Listed on MCP App Store

Add this badge to your site or README. It links to this page.

HTML

<a href="https://mcp-app-store.keenethics-labs.com/app/idealdata-io"><img src="https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg" alt="Listed on MCP App Store" width="220" height="44"></a>

Markdown

[![Listed on MCP App Store](https://mcp-app-store.keenethics-labs.com/badges/listed-on-mcp-app-store.svg)](https://mcp-app-store.keenethics-labs.com/app/idealdata-io)