vendor-performance-dashboard
Skill8 starsChanged 45 days agoMITView on GitHub
Build a live Amazon Vendor Central performance dashboard - KPI scorecards for vendor sales, ad spend, ad sales, ROAS and orders for the last 30 days vs the previous 30 days, a daily trend of vendor sales and ad spend, and an ASIN table with sales, ordered units, glance views and conversion rate - from DataDoe's Vendor Central and Amazon Ads data. In Claude Cowork it is a live artifact with a Refresh button; elsewhere a self-contained HTML file; both can be saved as PDF. Use for "vendor dashboard", "vendor central dashboard", "vendor performance", "1P performance", "vendor sales and ads", "glance views and conversion", or "live vendor report". Then offers an updated version with ordered revenue, TACoS, and per-ASIN lost Buy Box and out-of-stock rates.
In this repository26 other files it gives its agents.
- amazon-asin-search-auditorSkill
- amazon-listing-optimizerSkill
- buy-box-loss-root-causeSkill
- create-amazon-reconciliation-dashboardSkill
- create-orders-managerSkill
- daily-account-health-checkSkill
- daily-performance-health-checkSkill
- fba-inventory-ledger-reportSkill
- fba-stock-restock-plannerSkill
- keyword-rank-sqp-trackerSkill
- net-profit-pl-analyzerSkill
- ppc-asin-acos-alertSkill
- ppc-bid-optimizer-applySkill
- ppc-campaign-acos-alertSkill
- ppc-negative-keyword-applierSkill
- ppc-profit-bid-applySkill
- ppc-wasted-spend-watchdogSkill
- promotions-managerSkill
- restock-priority-alertSkill
- return-refund-analyzerSkill
- sales-movers-scannerSkill
- suppressed-inactive-listings-checkSkill
- vendor-inventory-reportSkill
- vendor-pnl-dashboardSkill
- weekly-business-reviewSkill
- weekly-sales-briefingSkill
The file
We split this skill into the rules it gives an agent
5 rules. About 0 are common, 4 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 56
If those bands put nearly every ASIN in one colour, colour against the account's median CVR instead (>= 1.25x median high, <= 0.75x low) …
- Very specific: line 70
Treat null numbers as 0 unless a step says otherwise.
- Distinctive: line 83
No Amazon Ads connection, or no ad rows for a window: show ad KPIs as "-" with a note, never as 0.
- Distinctive: line 121
Never embed a DataDoe key, MCP URL key or Anthropic key.
- Distinctive: line 204
Read-only. Vendor Central + Amazon Ads.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Vendor Performance Dashboard
A simple vendor performance dashboard that shows the main factors in one screen and takes a few minutes to build: KPI scorecards (vendor sales, ad spend, ad sales, ROAS, orders) for the last 30 days against the 30 before, a daily trendline of vendor sales and ad spend, and an ASIN breakdown with sales, ordered units, glance views and conversion. It is interactive - one Refresh pulls the newest data in seconds. Live from DataDoe.
When to use this
- A weekly or monthly look at a Vendor Central account, or before a vendor / client meeting.
- Spotting the ASIN that is not converting, or ad spend rising while sales stay flat.
- Trigger phrases: "vendor dashboard", "vendor central dashboard", "vendor performance", "1P performance", "vendor sales and ads", "glance views and conversion".
The framework. Same days on both sides
- Two equal windows, anchored on the newest vendor data. Vendor data arrives ~4 days
later than ads data. Find the newest
datewith vendor shipped revenue (D); current window =D-29 toD, previous window = the 30 days before. Use the same windows for ads, so a ROAS or a sales change is never computed against days that have ad spend but no vendor sales yet. - KPIs for each window, with the change vs previous (green good / red bad; for Ad Spend a
fall is green, a rise under 10% neutral grey, 10% or more red; ratio changes such as ROAS
are shown as the ratio change, percentages such as TACoS in points):
- Total Vendor Sales = sum
manufacturing_retail_shipped_revenue. - Ad Spend = sum
ad_spend; Ad Sales = sumad_sales. - ROAS = Ad Sales / Ad Spend (computed from the sums, never averaged).
- Total Orders (Units) = sum
manufacturing_retail_ordered_units(customer-ordered units).
- Total Vendor Sales = sum
- Trend: daily vendor sales and daily ad spend across the current window (two axes).
- ASIN breakdown (current window): sales, ordered units, glance views, and CVR = ordered units / glance views. Colour CVR: >= 15% high, >= 8% medium, below that low. If those bands put nearly every ASIN in one colour, colour against the account's median CVR instead (>= 1.25x median high, <= 0.75x low) and say so in a footnote.
Configuration
- MCP base:
https://mcp.datadoe.com/mcp/v1 - Export rules (every export in this skill): each
exports_createcall needslimitandoutputType. Aggregation aliases must differ from column names (sumofad_spend-> aliastotal_ad_spend; reusing the column name fails with ALIAS_COLLISION). Text columns can't be aggregated (maxon a name fails with INVALID_AGGREGATE) - put them ingroupByand merge rows in code. Load every page ofexports_source_get(a table's columns are spread over several pages). Ifexports_createalready returnsCOMPLETED, skip polling; otherwise pollexports_getevery ~5 s. For results you process in code, useexports_raw_url_getand save the file (the URL expires after ~15 minutes) instead of pulling it into the conversation. Treat null numbers as 0 unless a step says otherwise. Marketplace codes are ISO (the UK isGB). - Vendor data:
Sales, Traffic & Inventory by ASIN & Date(amazon_vendor_sales_traffic_and_inventory_by_child_and_date) -date,child_asin,manufacturing_retail_shipped_revenue(+manufacturing_retail_shipped_revenue_currency),manufacturing_retail_ordered_units,glance_views. Use the manufacturing view only (thesourcing_retail_*columns overlap it - never add them). Glance views are manufacturing traffic only. - Ads data:
Ad Performance by Campaign & Date(amazon_ads_performance_by_campaign_by_date)date,ad_spend,ad_sales,ad_orders(all campaign types; Sponsored Products attribute 7 days, Brands / Display 14 days). The old 14-day column names no longer exist - usead_salesandad_orders.
- Both sources require
from/to(YYYY-MM-DD). Row caps: 1,000 rows per JSON export, 5,000 per CSV - daily totals for 60 days fit easily in JSON. - No Amazon Ads connection, or no ad rows for a window: show ad KPIs as "-" with a note, never as 0.
- Currency: from the vendor
*_currencycolumn; check the ads budget currency (ad_campaign_budget_currency, null on Sponsored Brands rows) matches before computing ROAS. - Some marketplaces report no shipped revenue on Sundays (seen on a German account): draw those days as gaps in the trend line, not as 0, and footnote them.
Step-by-step workflow (MCP-native)
sellers_and_vendors_list-> pick the vendor account (accountType = VENDOR, with an Amazon Ads connection for the ad KPIs).exports_sources_get(queries "vendor sales", "campaign") -> resolve both sources;exports_source_getto confirm the columns above.- Newest vendor date
D:exports_createon the vendor table,from= 14 days ago,to= today,groupBy ["date"], summanufacturing_retail_shipped_revenue(aliasshipped_revenue_sum).D= the newest date whose sum is above 0 - rows can exist a day before their revenue does, and acountwould end the window on an empty day. - Three exports over
from=D-59,to=D(current + previous window):- Vendor daily:
groupBy ["date", "manufacturing_retail_shipped_revenue_currency"], summanufacturing_retail_shipped_revenue(total_revenue),manufacturing_retail_ordered_units(total_units),glance_views(total_glance_views), order bydate, JSON. Merge the rows of each date in code (rows without a currency belong to the same date). - Ads daily:
groupBy ["date", "ad_campaign_budget_currency"], sumad_spend(total_ad_spend),ad_sales(total_ad_sales),ad_orders(total_ad_orders), order bydate, JSON. Sponsored Brands rows have no budget currency - treat a null as the account's currency and merge per date. - ASIN breakdown:
from=D-29,to=D,groupBy ["child_asin"], the three vendor sums,orderByColumn total_revenue DESC,limit25 (top 25 by sales).
- Vendor daily:
- Download the results, split the daily rows at
D-29 into current / previous, compute KPIs and deltas. - Build the dashboard for the client:
- Claude Cowork: a live artifact - embed the current data so it renders at once; the
Refresh button re-runs steps 3-5 through
window.cowork.callMcpTool('<DataDoe tool name exactly as in your tool list>', args)and re-renders; declare exports_create / exports_get / exports_raw_download in the artifact's MCP tool list; parse the MCP envelope (content[0].text-> JSON, id fieldexportId). - Other clients: one self-contained HTML file with the data embedded; refresh = re-run the skill.
- Never embed a DataDoe key, MCP URL key or Anthropic key. Outside Cowork the Refresh button explains that live refresh runs inside Claude Cowork.
- Claude Cowork: a live artifact - embed the current data so it renders at once; the
Refresh button re-runs steps 3-5 through
- Report the path / show the artifact; offer the PDF and, if wanted, a routine that sends it.
- Offer the updated version (see below): "There's an updated version of this dashboard - it adds ordered revenue and TACoS to the scorecards, and each ASIN's lost Buy Box share and out-of-stock rate, so you can see why an ASIN isn't converting. Want it?" Build it only on a yes.
Output format
Modern dark style matching datadoe.com (background #0d1117, cards #161c2c, DataDoe green
#22c55e accents, blue #60a5fa for ad spend, Inter / system font), Chart.js 4 via CDN:
- Header: vendor pill (account name + marketplace) · "{current from} - {current to} vs {previous from} - {previous to}" · "Updated {date}" · Refresh · Download PDF.
- KPI scoreboard (5 cards): Total Vendor Sales · Ad Spend · Ad Sales · ROAS · Total Orders (Units); each with the value, a delta chip (+/- %) and "vs {previous value}".
- Trend chart: "Vendor Sales & Ad Spend - daily" - green area line for vendor sales (left axis), blue area line for ad spend (right axis), shared tooltip.
- Performance by ASIN (top 25): # · ASIN · Total Sales · Ordered Units · Glance Views · CVR (coloured pill).
- Footer: "Data via DataDoe".
PDF: Download PDF uses window.print() with print CSS (light background, A4 landscape) so the
user can restyle it for a client's brand and share it. If the user wants it sent on a schedule,
create a routine (scheduled task) that rebuilds the dashboard and emails the PDF to the
recipients they name - weekly or monthly.
Updated version (offer after the standard dashboard)
Same layout - the updated version adds two scorecards and the "why" behind each ASIN. Build it only when the user says yes (the Refresh button then re-runs the extra export too).
- Scorecards 6 and 7: Ordered Revenue (add sum
manufacturing_retail_ordered_revenue_amount, aliastotal_ordered_revenue, to the vendor daily export so both windows have it - customer orders; shipped revenue lags it) and TACoS (Ad Spend / Total Vendor Sales, change in points), both with the delta vs the previous 30 days. - ASIN table, two new columns for the top 25 ASINs:
- Lost Buy Box % -
lost_featured_offer(share of glance views where Amazon Retail lost the Featured Offer), averaged per ASIN weighted by dailyglance_views. - Out of Stock % -
manufacturing_retail_procurable_product_oos, weighted the same way. - Pull them with one extra export: current window, filter
child_asin in <top 25, comma-separated>,groupBy ["date", "child_asin"], aggregations avglost_featured_offer(lost_bb_rate), avgmanufacturing_retail_procurable_product_oos(oos_rate), sumglance_views(day_glance_views) - one row per ASIN-day, 25 x 30 rows fits in JSON. Weight in code by glance views and skip days with a null rate or 0 glance views - never average the daily rates unweighted. - Colour: amber from 10%, red from 25%.
- Lost Buy Box % -
- Reading hint under the table: low CVR + high Lost Buy Box -> a price / offer problem; low CVR + high Out of Stock -> an availability problem; both low -> look at content and reviews.
Worked example (illustrative)
German vendor account, newest vendor data 12 May. Current window 13 Apr - 12 May vs 14 Mar - 12 Apr: Vendor Sales €182,400 (+6.1%), Ad Spend €14,900 (+18%), Ad Sales €61,300 (+4%), ROAS 4.11x (vs 4.66x), Orders 9,870 units (+3.2%). The trend shows spend stepping up from 1 May while sales stay flat. In the ASIN table, ASIN C has the third-highest glance views but a 3.1% CVR (red) - the one to look at first (price, content or availability).
Quality self-check
- Did I anchor both windows on the newest date with vendor shipped revenue (not yesterday, not just the newest row) and use the same days for ads?
- Are ROAS and CVR computed from summed values, not averaged daily ratios?
- Did I use
ad_sales/ad_ordersand the manufacturing-view vendor columns only? - Ads missing -> "-" with a note, not zeros?
- Currency from the data? No credentials in the page?
- Did I offer the updated version at the end - and change nothing unless the user said yes?
Common mistakes
- Ending both windows yesterday - the last ~4 days have ad spend but no vendor sales, so sales and ROAS look worse than they are.
- Using the removed 14-day ad columns (
ad_sales_14d) - the export fails. - Averaging daily ROAS or CVR.
- Adding sourcing-view and manufacturing-view revenue together.
- JSON
limitabove 1,000. - Finding the newest date with
count- rows can arrive a day before their revenue, which ends the window on a €0 day. - Plotting Sundays without shipped revenue as 0 (draw gaps).
Notes
- Read-only. Vendor Central + Amazon Ads.
- Want more than the main factors? Ask in the same chat - filters, sorting or more marketplaces can be added; ordered revenue, TACoS, lost Buy Box and out-of-stock rates come with the updated version.
- A DataDoe skill, built on
amazon_vendor_sales_traffic_and_inventory_by_child_and_dateandamazon_ads_performance_by_campaign_by_date.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

