NEWSee Promoosh in action, no login requiredSee demo

MCP library

Ask Claude about your own store numbers

Promoosh gives Claude read access to your store, ads and accounts. Here you can see every tool, every field and how each number is calculated.

Works with Claude

Connect ClaudeSee demo

31

tools

102

fields in the metric library

25

ready-made prompts

What and why

MCP (Model Context Protocol) is the standard AI assistants use to fetch data from other services. Promoosh's MCP connector uses the same tools as “Ask Promoosh” inside the app, so Claude gets the same numbers you see in the dashboard.

You can ask questions across sources in one conversation – for example which products make money after ad spend and where competitors are cheaper – without exporting spreadsheets.

Metric library

Every field the list_metrics tool describes, with unit, VAT and refund basis and formula. The list is generated from the same source as the tool.

102 of 102 fields

  • totalSalesInclVatquery_profitability
    amountVAT: incl. VATRefunds: before refunds

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads, Promoosh

  • returnsInclVatquery_profitability
    amountVAT: incl. VATRefunds: –

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads, Promoosh

  • netSalesExclVatquery_profitability
    amountVAT: excl. VATRefunds: after refunds

    order subtotal after discounts − VAT − returns excl. VAT

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads, Promoosh

  • cogsquery_profitability
    amountVAT: excl. VATRefunds: after refundsmay be an estimate

    unit cost × units sold net of returns (fallback: price incl. VAT × fallback rate)

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads, Promoosh

  • grossProfitquery_profitability
    amountVAT: excl. VATRefunds: after refunds

    netSalesExclVat − cogs

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads, Promoosh

  • grossMarginPctquery_profitability
    %VAT: –Refunds: –

    grossProfit ÷ netSalesExclVat × 100

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads, Promoosh

  • transactionFeesquery_profitability
    amountVAT: –Refunds: –

    Σ gateway (fee % × gateway net sales + fixed fee × orders)

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads, Promoosh

  • adSpendquery_profitability
    amountVAT: excl. VATRefunds: –

    Google Ads + Meta Ads account spend in the window

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads, Promoosh

  • operatingCostsquery_profitability
    amountVAT: unknownRefunds: –

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads, Promoosh

  • operatingResultquery_profitability
    amountVAT: unknownRefunds: after refunds

    grossProfit − transactionFees − adSpend − operatingCosts

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads, Promoosh

  • summary.totalOrdersquery_orders
    countVAT: –Refunds: –

    Sources: Shopify orders, Shopify catalogue, Promoosh pixel, Google Ads, Meta Ads

  • summary.totalReturnsquery_orders
    countVAT: –Refunds: –

    Sources: Shopify orders, Shopify catalogue, Promoosh pixel, Google Ads, Meta Ads

  • summary.negativeGrossProfitOrdersquery_orders
    countVAT: –Refunds: –

    Sources: Shopify orders, Shopify catalogue, Promoosh pixel, Google Ads, Meta Ads

  • recentOrders[].amountquery_orders
    amountVAT: excl. VATRefunds: after refunds

    Sources: Shopify orders, Shopify catalogue, Promoosh pixel, Google Ads, Meta Ads

  • recentOrders[].cogsquery_orders
    amountVAT: excl. VATRefunds: after refundsmay be an estimate

    Sources: Shopify orders, Shopify catalogue, Promoosh pixel, Google Ads, Meta Ads

  • recentOrders[].grossProfitquery_orders
    amountVAT: excl. VATRefunds: after refunds

    amount − cogs

    Sources: Shopify orders, Shopify catalogue, Promoosh pixel, Google Ads, Meta Ads

  • recentOrders[].netProfitquery_orders
    amountVAT: excl. VATRefunds: after refundsmay be an estimate

    grossProfit − (channel spend ÷ attributed orders of that channel)

    Sources: Shopify orders, Shopify catalogue, Promoosh pixel, Google Ads, Meta Ads

  • points[].netSalesquery_daily_profit
    amountVAT: excl. VATRefunds: after refunds

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads

  • points[].grossProfitquery_daily_profit
    amountVAT: excl. VATRefunds: after refunds

    Σ order amount − Σ order cogs

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads

  • points[].adSpendquery_daily_profit
    amountVAT: excl. VATRefunds: –

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads

  • points[].netProfitquery_daily_profit
    amountVAT: excl. VATRefunds: after refunds

    grossProfit − adSpend

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads

  • points[].ordersquery_daily_profit
    countVAT: –Refunds: –

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads

  • topProductsByNetSales[].unitsSoldquery_product_advisor
    countVAT: –Refunds: –

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads

  • topProductsByNetSales[].netSalesquery_product_advisor
    amountVAT: excl. VATRefunds: after refunds

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads

  • topProductsByNetSales[].grossProfitquery_product_advisor
    amountVAT: excl. VATRefunds: after refunds

    netSales − cogs

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads

  • topProductsByNetSales[].adSpendquery_product_advisor
    amountVAT: excl. VATRefunds: –

    Google product spend + real Meta product spend

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads

  • topProductsByNetSales[].netProfitquery_product_advisor
    amountVAT: excl. VATRefunds: after refunds

    grossProfit − adSpend

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads

  • products[].spendquery_product_ad_spend
    amountVAT: excl. VATRefunds: –

    Sources: Google Ads, Meta Ads, Shopify catalogue

  • products[].revenuequery_product_ad_spend
    amountVAT: unknownRefunds: before refunds

    Sources: Google Ads, Meta Ads, Shopify catalogue

  • products[].roasquery_product_ad_spend
    ratioVAT: –Refunds: –

    revenue ÷ spend

    Sources: Google Ads, Meta Ads, Shopify catalogue

  • campaigns[].spendquery_campaigns
    amountVAT: excl. VATRefunds: –

    Sources: Google Ads, Meta Ads

  • campaigns[].clicksquery_campaigns
    countVAT: –Refunds: –

    Sources: Google Ads, Meta Ads

  • campaigns[].conversionsquery_campaigns
    countVAT: –Refunds: –

    Sources: Google Ads, Meta Ads

  • campaigns[].revenuequery_campaigns
    amountVAT: unknownRefunds: before refundsmay be an estimate

    Sources: Google Ads, Meta Ads

  • campaigns[].roasquery_campaigns
    ratioVAT: –Refunds: –

    revenue ÷ spend (only when roasIsReal)

    Sources: Google Ads, Meta Ads

  • campaigns[].dailyBudgetquery_campaigns
    amountVAT: excl. VATRefunds: –

    Sources: Google Ads, Meta Ads

  • ads[].spendquery_ads
    amountVAT: excl. VATRefunds: –

    Sources: Google Ads, Meta Ads

  • ads[].clicksquery_ads
    countVAT: –Refunds: –

    Sources: Google Ads, Meta Ads

  • ads[].impressionsquery_ads
    countVAT: –Refunds: –

    Sources: Google Ads, Meta Ads

  • ads[].conversionsquery_ads
    countVAT: –Refunds: –

    Sources: Google Ads, Meta Ads

  • ads[].revenuequery_ads
    amountVAT: unknownRefunds: before refundsmay be an estimate

    Sources: Google Ads, Meta Ads

  • scorequery_health_score
    scoreVAT: –Refunds: –

    Sources: Promoosh

  • rows[].revenuequery_fiken
    amountVAT: excl. VATRefunds: –

    Sources: Fiken

  • rows[].cogsquery_fiken
    amountVAT: excl. VATRefunds: –

    Sources: Fiken

  • rows[].payrollquery_fiken
    amountVAT: –Refunds: –

    Sources: Fiken

  • rows[].opexquery_fiken
    amountVAT: excl. VATRefunds: –

    Sources: Fiken

  • rows[].financequery_fiken
    amountVAT: –Refunds: –

    Sources: Fiken

  • rows[].resultquery_fiken
    amountVAT: excl. VATRefunds: –

    revenue − cogs − payroll − opex − finance

    Sources: Fiken

  • bankBalanceTotalquery_fiken
    amountVAT: –Refunds: –

    Sources: Fiken

  • vatPositionLast12Months.netOwedquery_fiken
    amountVAT: –Refunds: –may be an estimate

    output − input

    Sources: Fiken

  • purchaseOrders[].freightTotalquery_procurement
    amountVAT: unknownRefunds: –

    Sources: Promoosh

  • purchaseOrders[].lines[].unitCostquery_procurement
    amountVAT: unknownRefunds: –

    Sources: Promoosh

  • keywords[].latestPositionquery_seo
    positionVAT: –Refunds: –

    Sources: Promoosh

  • topByAdSpend[].ownPricequery_price_position
    amountVAT: incl. VATRefunds: –

    Sources: Google Merchant Center, Google Ads, Shopify orders

  • topByAdSpend[].benchmarkPricequery_price_position
    amountVAT: incl. VATRefunds: –

    Sources: Google Merchant Center, Google Ads, Shopify orders

  • topByAdSpend[].suggestedPricequery_price_position
    amountVAT: incl. VATRefunds: –may be an estimate

    Sources: Google Merchant Center, Google Ads, Shopify orders

  • topByAdSpend[].marginNowPctquery_price_position
    %VAT: –Refunds: –

    (price excl. VAT − unit cost) ÷ price excl. VAT × 100

    Sources: Google Merchant Center, Google Ads, Shopify orders

  • topByAdSpend[].shoppingSpend30dquery_price_position
    amountVAT: excl. VATRefunds: –

    Sources: Google Merchant Center, Google Ads, Shopify orders

  • topByAdSpend[].poasquery_price_position
    ratioVAT: –Refunds: –

    gross profit ÷ ad spend

    Sources: Google Merchant Center, Google Ads, Shopify orders

  • topByAdSpend[].poasAtSuggestedquery_price_position
    ratioVAT: –Refunds: –may be an estimate

    gross profit at suggested price ÷ ad spend

    Sources: Google Merchant Center, Google Ads, Shopify orders

  • topMatches[].ownPricequery_competitor_prices
    amountVAT: incl. VATRefunds: –

    Sources: Competitor pages, Promoosh, Shopify orders, Google Ads

  • topMatches[].competitorPricequery_competitor_prices
    amountVAT: incl. VATRefunds: –

    Sources: Competitor pages, Promoosh, Shopify orders, Google Ads

  • topMatches[].gapKrquery_competitor_prices
    amountVAT: incl. VATRefunds: –

    ownPrice − competitorPrice

    Sources: Competitor pages, Promoosh, Shopify orders, Google Ads

  • topMatches[].gapPctquery_competitor_prices
    %VAT: –Refunds: –

    gapKr ÷ competitorPrice × 100

    Sources: Competitor pages, Promoosh, Shopify orders, Google Ads

  • topMatches[].poasquery_competitor_prices
    ratioVAT: –Refunds: –

    gross profit ÷ ad spend

    Sources: Competitor pages, Promoosh, Shopify orders, Google Ads

  • topMatches[].poasAtCompetitorPricequery_competitor_prices
    ratioVAT: –Refunds: –may be an estimate

    gross profit at competitor price ÷ ad spend

    Sources: Competitor pages, Promoosh, Shopify orders, Google Ads

  • counts.totalVariantsquery_feed_labels
    countVAT: –Refunds: –

    Sources: Promoosh

  • products[].last150Days.netSalesAfterReturnsExclVatquery_feed_label_explain
    amountVAT: excl. VATRefunds: after refunds

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads, Promoosh

  • products[].last150Days.grossProfitquery_feed_label_explain
    amountVAT: excl. VATRefunds: after refunds

    netSalesAfterReturnsExclVat − cogs

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads, Promoosh

  • products[].last150Days.netProfitAfterAdsquery_feed_label_explain
    amountVAT: excl. VATRefunds: after refunds

    grossProfit − googleAdSpend − real metaAdSpend

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads, Promoosh

  • products[].last150Days.poasquery_feed_label_explain
    ratioVAT: –Refunds: –

    grossProfit ÷ (googleAdSpend + real metaAdSpend)

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads, Promoosh

  • products[].priceInclVatquery_products
    amountVAT: incl. VATRefunds: –

    Sources: Shopify catalogue, Promoosh

  • products[].unitCostExclVatquery_products
    amountVAT: excl. VATRefunds: –

    Sources: Shopify catalogue, Promoosh

  • products[].potentialMarginPctquery_products
    %VAT: –Refunds: –

    (price excl. VAT − unit cost) ÷ price excl. VAT × 100

    Sources: Shopify catalogue, Promoosh

  • products[].unitsSoldLast30Daysquery_products
    countVAT: –Refunds: –

    Sources: Shopify catalogue, Promoosh

  • products[].series[].unitsSoldquery_product_timeseries
    countVAT: –Refunds: –

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads

  • products[].series[].netSalesAfterReturnsExclVatquery_product_timeseries
    amountVAT: excl. VATRefunds: after refunds

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads

  • products[].series[].grossProfitquery_product_timeseries
    amountVAT: excl. VATRefunds: after refunds

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads

  • products[].series[].googleAdSpendquery_product_timeseries
    amountVAT: excl. VATRefunds: –

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads

  • products[].series[].metaAdSpendquery_product_timeseries
    amountVAT: excl. VATRefunds: –

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads

  • products[].series[].netProfitAfterAdsquery_product_timeseries
    amountVAT: excl. VATRefunds: after refunds

    grossProfit − googleAdSpend − real metaAdSpend

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads

  • products[].series[].poasquery_product_timeseries
    ratioVAT: –Refunds: –

    grossProfit ÷ (googleAdSpend + real metaAdSpend)

    Sources: Shopify orders, Shopify catalogue, Google Ads, Meta Ads

  • lines[].netSalesExclVatquery_order
    amountVAT: excl. VATRefunds: after refunds

    Sources: Shopify orders, Shopify catalogue, Promoosh pixel

  • lines[].unitCostquery_order
    amountVAT: excl. VATRefunds: –may be an estimate

    Sources: Shopify orders, Shopify catalogue, Promoosh pixel

  • lines[].grossProfitquery_order
    amountVAT: excl. VATRefunds: after refunds

    netSalesExclVat − unitCost × kept units

    Sources: Shopify orders, Shopify catalogue, Promoosh pixel

  • rows[].spendquery_ads_breakdown
    amountVAT: excl. VATRefunds: –

    Sources: Google Ads, Meta Ads

  • rows[].clicksquery_ads_breakdown
    countVAT: –Refunds: –

    Sources: Google Ads, Meta Ads

  • rows[].impressionsquery_ads_breakdown
    countVAT: –Refunds: –

    Sources: Google Ads, Meta Ads

  • rows[].conversionsquery_ads_breakdown
    countVAT: –Refunds: –

    Sources: Google Ads, Meta Ads

  • rows[].conversionValuequery_ads_breakdown
    amountVAT: unknownRefunds: before refunds

    Sources: Google Ads, Meta Ads

  • rows[].roasquery_ads_breakdown
    ratioVAT: –Refunds: –

    conversionValue ÷ spend

    Sources: Google Ads, Meta Ads

  • totals.roasquery_ads_breakdown
    ratioVAT: –Refunds: –

    Σ conversionValue ÷ Σ spend

    Sources: Google Ads, Meta Ads

  • discountCodes[].usesquery_influencers
    countVAT: –Refunds: –

    Sources: Promoosh

  • discountCodes[].revenueAttributedquery_influencers
    amountVAT: unknownRefunds: unknown

    Sources: Promoosh

  • discountCodes[].aovquery_influencers
    amountVAT: unknownRefunds: unknown

    revenueAttributed ÷ uses

    Sources: Promoosh

  • totalAssetsReturnedquery_assets
    countVAT: –Refunds: –

    Sources: Promoosh

  • baselines[].currentRevenuequery_goal
    amountVAT: unknownRefunds: unknown

    Sources: Promoosh

  • baselines[].targetRevenuequery_goal
    amountVAT: unknownRefunds: unknown

    Sources: Promoosh

  • pixel.eventsLast7Daysquery_tracking_health
    countVAT: –Refunds: –

    Sources: Promoosh pixel, Promoosh

  • orders30d.trackedAttributionPctquery_tracking_health
    %VAT: –Refunds: –

    orders30d.withTrackedChannel ÷ orders30d.paidOrders × 100

    Sources: Promoosh pixel, Promoosh

  • forwarding24h.forwardedPctquery_tracking_health
    %VAT: –Refunds: –

    forwarding24h.forwardedOrders ÷ forwarding24h.paidOrders × 100

    Sources: Promoosh pixel, Promoosh

  • pixelCoverage.pixelSeenPctquery_tracking_health
    %VAT: –Refunds: –

    pixelCoverage.pixelSeenOrders ÷ pixelCoverage.paidOrders × 100

    Sources: Promoosh pixel, Promoosh

Tools

Every tool the connector gives Claude. All are read-only, and all take an optional organization_id when you have several stores.

Account

list_organizations

Lists the stores you have access to, with the modules each one has.

Parameters: No parameters

Profitability

query_profitability

Profit and loss for a period: net sales, cost of goods, gross profit, fees, ad spend and operating result.

Parameters: periodfromtocompare

Periods: today, yesterday, this_week, last_week, this_month, last_month, this_quarter, last_quarter, this_year, last_year, last_7_days, last_14_days, last_30_days, last_60_days, last_90_days, last_180_days, last_12_months, custom

Comparison: prev, mom, qoq, yoy

query_daily_profit

Net sales, gross profit, ad spend and net profit day by day.

Parameters: daysperiodfromtocompare

Periods: today, yesterday, this_week, last_week, this_month, last_month, this_quarter, last_quarter, this_year, last_year, last_7_days, last_14_days, last_30_days, last_60_days, last_90_days, last_180_days, last_12_months, custom

Comparison: prev, mom, qoq, yoy

query_reports

The tiles in the report builder.

Parameters: No parameters

Ads (Google and Meta)

query_product_ad_spend

Ad spend and revenue per product from Google Shopping/PMax and Meta.

Parameters: daysperiodfromtocompare

Periods: today, yesterday, this_week, last_week, this_month, last_month, this_quarter, last_quarter, this_year, last_year, last_7_days, last_14_days, last_30_days, last_60_days, last_90_days, last_180_days, last_12_months, custom

Comparison: prev, mom, qoq, yoy

query_campaigns

Campaigns with status, spend, revenue, ROAS, daily budget, clicks and conversions.

Parameters: daysperiodfromtocompare

Periods: today, yesterday, this_week, last_week, this_month, last_month, this_quarter, last_quarter, this_year, last_year, last_7_days, last_14_days, last_30_days, last_60_days, last_90_days, last_180_days, last_12_months, custom

Comparison: prev, mom, qoq, yoy

query_ads

Performance per ad: spend, revenue, impressions, clicks and conversions.

Parameters: daysperiodfromtocompare

Periods: today, yesterday, this_week, last_week, this_month, last_month, this_quarter, last_quarter, this_year, last_year, last_7_days, last_14_days, last_30_days, last_60_days, last_90_days, last_180_days, last_12_months, custom

Comparison: prev, mom, qoq, yoy

query_ads_breakdown

Flexible ad breakdown from Google or Meta per campaign, ad group, ad, product, search term or keyword.

Parameters: platform (required)level (required)segmentfromtodaysperiodcomparenamelimit

Periods: today, yesterday, this_week, last_week, this_month, last_month, this_quarter, last_quarter, this_year, last_year, last_7_days, last_14_days, last_30_days, last_60_days, last_90_days, last_180_days, last_12_months, custom

Comparison: prev, mom, qoq, yoy

Products

query_product_advisor

Sales, profit, ad spend and a recommendation per product.

Parameters: periodfromtocompare

Periods: today, yesterday, this_week, last_week, this_month, last_month, this_quarter, last_quarter, this_year, last_year, last_7_days, last_14_days, last_30_days, last_60_days, last_90_days, last_180_days, last_12_months, custom

Comparison: prev, mom, qoq, yoy

query_feed_labels

Configuration and product counts per smart label (Champion, Assistant, Sleeper, Enemy).

Parameters: No parameters

query_feed_label_explain

Why one product has its smart label, with the numbers the rule used.

Parameters: product (required)

query_products

Search the whole product catalogue with price, unit cost, stock and sales over the last 30 days.

Parameters: queryfilterlimit

query_product_timeseries

One product month by month or week by week: sales, profit, ad spend and POAS.

Parameters: product (required)granularitymonthsweeks

Orders and sources

query_orders

Order overview: number of orders, returns, loss-making orders and the latest orders with profit and tracked channel.

Parameters: daysperiodfromtocompare

Periods: today, yesterday, this_week, last_week, this_month, last_month, this_quarter, last_quarter, this_year, last_year, last_7_days, last_14_days, last_30_days, last_60_days, last_90_days, last_180_days, last_12_months, custom

Comparison: prev, mom, qoq, yoy

query_order_sources

Where orders came from: first and last tracked visit, channel and counts per channel.

Parameters: daysproduct

query_order

One order: amounts, discounts, refunds, profit per line and where the buyer came from.

Parameters: order (required)

Tracking

query_tracking_health

Tracking status: pixel, share of orders with a tracked channel, forwarding to Google and Meta and open alerts.

Parameters: No parameters

SEO and competitors

query_health_score

Latest technical SEO score (0–100) with its breakdown.

Parameters: No parameters

query_seo

Keywords with their latest Google position, and changes to competitors' product pages.

Parameters: limit

query_price_position

Price position from Google Merchant Center: own price, market benchmark and Google's price suggestions.

Parameters: No parameters

query_competitor_prices

Competitor prices for products you have linked, with price gap, margin and latest price changes.

Parameters: No parameters

Accounting and purchasing

query_fiken

Booked Fiken figures per month: revenue, purchases, payroll, operating costs and result.

Parameters: months

query_procurement

Supplier invoices and goods receipt with supplier, freight and line items.

Parameters: limit

Setup, content and goals

query_content

Content in the queue (blog and social) with status and scheduled date.

Parameters: limit

query_setup_tasks

The setup checklist with what is done and what is left.

Parameters: No parameters

query_proposed_actions

Proposals in the approval queue with their status.

Parameters: limit

query_influencers

Discount codes with uses and revenue, and influencer contacts.

Parameters: limit

query_assets

The image library with rights status.

Parameters: limit

query_goal

Stored revenue goals over time.

Parameters: limit

Reference and definitions

query_promoosh_docs

How Promoosh calculates: definitions, formulas, thresholds and time windows.

Parameters: topic (required)

list_metrics

Machine-readable list of every field with unit, VAT and refund basis and formula.

Parameters: toolarea

Prompt library

Copy a question and paste it into Claude. Below each question you can see which tools it uses.

  • Give me a short weekly summary: profit last week compared with the week before, the three campaigns that spent the most, and whether tracking works. State the period and whether figures exclude VAT.

    Uses: query_profitabilityquery_campaignsquery_tracking_health

  • What should I look at today? Check unfinished setup steps, proposals waiting for approval and open tracking alerts.

    Uses: query_setup_tasksquery_proposed_actionsquery_tracking_health

  • Compare this month with the same period last year: net sales, gross profit, ad spend and operating result.

    Uses: query_profitability

  • Show last month day by day. Which days had negative net profit after ads?

    Uses: query_daily_profit

  • Why is my gross margin what it is this quarter? Explain how Promoosh calculates cost of goods and whether anything is estimated.

    Uses: query_profitabilityquery_promoosh_docs

  • Which Google campaigns spent the most last month, and how did total spend and conversions change from the month before?

    Uses: query_campaigns

  • Which search terms cost the most in the last 30 days without converting?

    Uses: query_ads_breakdown

  • Split last quarter's Google Ads spend by device, with ROAS per device.

    Uses: query_ads_breakdown

  • Show Meta spend and conversions per month this year.

    Uses: query_ads_breakdown

  • Which Meta ads had the highest cost and fewest conversions in the last 14 days?

    Uses: query_ads

  • Split Meta spend in the last 30 days by country and campaign.

    Uses: query_ads_breakdown

  • Which products lose money after ad spend this month, and what does Promoosh recommend for each?

    Uses: query_product_advisor

  • How many products are Champion, Assistant, Sleeper and Enemy? Then explain why a product I name has its label.

    Uses: query_feed_labelsquery_feed_label_explain

  • List the products that are missing a cost price and have sold something in the last 30 days.

    Uses: query_products

  • Find the product with the highest net sales in the last 30 days and show how sales, ad spend and POAS have developed month by month.

    Uses: query_product_advisorquery_product_timeseries

  • How many orders lost money last week, and how many are missing a cost price?

    Uses: query_orders

  • Did sales in the last 30 days come from ads, organic search, AI assistants or direct? Give the number of orders per channel.

    Uses: query_order_sources

  • Is my tracking working? What share of orders has a tracked channel, and were the last 24 hours' orders forwarded to Google and Meta?

    Uses: query_tracking_health

  • What did I make on order #1001, and where did the buyer come from?

    Uses: query_order

  • Where are competitors cheaper than me, and what would my margin be if I matched their price?

    Uses: query_competitor_prices

  • How is my SEO score, and where do my keywords rank in Google right now?

    Uses: query_health_scorequery_seo

  • Which price suggestions does Google have for my products, and which don't pay off?

    Uses: query_price_position

  • What is the booked result per month for the last six months in Fiken?

    Uses: query_fiken

  • Why does the result in Fiken differ from Promoosh's calculated result last month?

    Uses: query_fikenquery_profitabilityquery_promoosh_docs

  • Which ad figures can Promoosh give me, and how is each of them defined?

    Uses: list_metrics

How it works

  1. 1

    Connect

    Add Promoosh as a connector in Claude and sign in with your Promoosh account.

  2. 2

    Ask

    Ask your question in plain language. Claude picks the tools it needs.

  3. 3

    Get numbers

    The answer is based on your store's numbers, and every tool result has a meta object that says what the numbers mean.

The meta object in every answer

Every tool result carries a meta object, so a number never stands without context:

  • period: from and to date, whether the period is still running, and which convention it follows
  • timezone: the store's timezone, in which periods are calculated
  • VAT and refunds: which fields are excluding VAT, including VAT or unknown, and whether numbers are before or after refunds
  • attribution: whether a number is reported by the ad platform or tracked by Promoosh, and the window
  • low data: lowData is set when the period has fewer than 15 orders or conversions
  • estimates: isEstimate and estimateFields show which fields may be estimates, for example fallback cost of goods
  • warnings: codes that say when something is missing, limited or follows an older period definition

Periods and comparison

Periods are calendar days in the store's timezone. “Last 30 days” is the 30 days ending today, and today is included. “This month” runs to now, “last month” is the whole month. Weeks start on Monday.

Comparison: previous period, month, quarter or year (prev, mom, qoq, yoy). A period that is not finished is compared with the same number of days of the period before. If something cannot be compared, the answer says so with the reason.

Some older values keep their old meaning so no number changes: “last_30_days” in profitability and the product advisor is a rolling 30 × 24 hours, and “days” means rolling 24-hour days. The answer flags this.

Reporting rules

These are the same rules the connector gives Claude:

  1. Always state the period and basis when you report a number.
  2. Never sum or average ratios (ROAS, POAS, margins, percentages). Recalculate them from the sums.
  3. Never add Google's and Meta's conversions together. Both can claim the same order.
  4. Never mix numbers with a different VAT or refund basis in one calculation.
  5. Estimates are marked in the answer and must be described as estimates.

Security and privacy

  • Read-only: the connector only has read tools, and the database rejects writes from connected AI apps.
  • The tools do not show your customers' names, email addresses, phone numbers or addresses.
  • You only get data from stores you are a member of. This is checked on every call.
  • Data from modules that are not in your plan is not available.
  • Claude cannot change budgets, prices or content. Changes are created as proposals in Promoosh and must be approved there.
  • Every call is logged with tool and time, never the content, and there is a limit of 60 calls a minute.
  • The sign-in Claude receives does not give direct database access to raw tracking data, order payloads, chat or invitations.
  • The AI provider you connect receives the numbers you ask about.
  • You can revoke access under Settings → Account → Connected AI apps.

Setup in Claude

Applies to claude.ai and Claude Desktop. The connector is added to your Claude account and is then available in both.

Server URLhttps://www.promoosh.com/api/mcp
  1. In Claude, go to Customize → Connectors, click “+ Add” and choose “Add custom connector”. On Team and Enterprise plans an owner first adds it under Organization settings → Connectors.
  2. Name it Promoosh, paste the server URL above and click Continue.
  3. If Claude asks for an OAuth client, choose “Register automatically”.
  4. Click Connect, sign in with your Promoosh account and approve read access.
  5. Open a new conversation in claude.ai or Claude Desktop and ask a question from the prompt library.

Other MCP clients

Promoosh follows the MCP standard, but we have only tried Claude. Other MCP clients may work, but we have not confirmed it.

What we don't do yet

  • The AI cannot make changes. Anything that changes something goes through the approval queue in the app.
  • No customer-level numbers: new vs returning customers, LTV and RFM.
  • No AI tagging of ad creatives.
  • One attribution model per store (a setting), not selectable per question.
  • Ad numbers are fetched directly from Google and Meta, at most 400 days per query. We don't store ad history ourselves.
  • Comparison exists for profitability, orders, daily profit, products and the ad tools, not for every tool.
  • No combined report across stores with currency conversion.

Questions and answers

What is MCP?

MCP (Model Context Protocol) is an open standard that lets AI assistants like Claude fetch data from other services through tools. Promoosh's connector gives Claude read tools for your own store numbers.

Which data can Claude see?

Numbers from the modules in your plan: profitability, orders, products, ads from Google and Meta, tracking, SEO, competitor prices and Fiken. The metric library above lists every field.

What about privacy?

The tools do not show your customers' names, email addresses, phone numbers or addresses. Claude only gets the numbers you ask about, and only from stores you are a member of.

Can Claude change anything?

No. The connector only has read tools, and the database rejects writes from connected AI apps.

I have several stores. How do I choose?

Ask Claude to list your stores with the list_organizations tool, and say which one you mean. Every tool takes an optional organization_id.

Can Claude change my ads?

No. Changes such as pausing or budgets are created as proposals in Promoosh and must be approved in the app before anything happens in the ad account.

Why do the numbers differ from ROAS in Meta?

Meta's ROAS is the conversion value Meta attributes to itself divided by spend, with Meta's own attribution window (standard 7-day click and 1-day view). Promoosh also shows POAS – gross profit divided by ad spend – and its own pixel tracking. meta.attribution in every answer says which source a number comes from.

Try Promoosh with Claude

Install Promoosh from the Shopify App Store and connect Claude.

Try free for 10 daysSee demo