fyva.

v3 · confidence scoring

Week of 17 Mar 2026

Total SKUs

14,284

all products

Star

—

Monitor

—

Pause

—

Kill

—

Zone

Star

—

Monitor

—

Pause

—

Kill

—

Import Product Data

Upload your Shopify CSV exports. The importer reads your column headers, auto-maps known Shopify fields, and lets you manually map any unrecognised columns before loading into the dashboard.

Direct Shopify API — via proxy

Point to your deployed Cloudflare Worker proxy URL. The proxy holds your secret token — it never touches the browser. See the setup guide below.

Proxy URL

Route

Live connection — fetch data automatically

Paste a public CSV URL (Google Sheets published link, Coupler.io export, or your proxy endpoint) and the dashboard fetches it on every load. No manual file upload needed.

Products CSV URL

Analytics CSV URL

Saved URLs load automatically on next open.

or upload CSV files manually below

Upload files manually

⇣

Products CSV

Shopify Admin → Products → Export
Contains inventory, variants, tags, dates

⇣

Analytics CSV

Analytics → Reports → Sales / Sessions by product
Contains sessions, orders, checkout sessions

— or use sample data to try the importer —

Load sample Products CSV Load sample Analytics CSV Load both samples

How to set up a live connection

Three routes to get a URL that this dashboard can fetch automatically.

1 Sign up at coupler.io → New importer → Source: Shopify
2 Enter your Shopify store URL and grant access
3 Destination: Google Sheets → select your sheet
4 Schedule: Weekly, Monday 6am IST
5 In Google Sheets: File → Share → Publish to web → CSV → Copy link
6 Paste that link into the Products CSV URL field above
7 Repeat for your Analytics sheet if you have a second Coupler importer for sales data

Cost: ~$19/mo. Handles Shopify's 2 req/sec rate limit and pagination for 14,000 products automatically.

Deploy this as a Cloudflare Worker (free tier, <2 min deploy). It holds your secret API token and the dashboard calls it safely.

// Cloudflare Worker — deploy at workers.cloudflare.com
export default {
  async fetch(request) {
    const SHOP  = 'fyva.myshopify.com';
    const TOKEN = 'your-admin-api-token'; // set as env var
    const url   = request.url;
    const type  = new URL(url).searchParams.get('type');

    const endpoint = type === 'analytics'
      ? `https://${SHOP}/admin/api/2024-10/orders.json?status=any&limit=250`
      : `https://${SHOP}/admin/api/2024-10/products.json?limit=250`;

    const resp = await fetch(endpoint, {
      headers: { 'X-Shopify-Access-Token': TOKEN }
    });
    const data = await resp.json();

    return new Response(JSON.stringify(data), {
      headers: {
        'Content-Type': 'application/json',
        'Access-Control-Allow-Origin': '*'
      }
    });
  }
};

Your Worker URL becomes: https://fyva-proxy.yourname.workers.dev?type=products — paste that into the URL field above.

Use Google Apps Script to pull from Shopify and write to a sheet automatically — then publish the sheet as CSV.

// Google Apps Script — Tools → Script editor
function syncShopify() {
  const SHOP  = 'fyva.myshopify.com';
  const TOKEN = 'your-token'; // Script Properties → TOKEN
  const sheet = SpreadsheetApp.getActive().getSheetByName('Products');

  const resp = UrlFetchApp.fetch(
    `https://${SHOP}/admin/api/2024-10/products.json?limit=250`,
    { headers: { 'X-Shopify-Access-Token': TOKEN } }
  );
  const products = JSON.parse(resp.getContentText()).products;

  sheet.clearContents();
  sheet.appendRow(['Handle','Title','Type','Tags','Variant Price','Variant Inventory Qty','Published At']);
  products.forEach(p => p.variants.forEach(v => {
    sheet.appendRow([p.handle, p.title, p.product_type, p.tags, v.price, v.inventory_quantity, p.published_at]);
  }));
}
// Set trigger: Time-driven → Weekly → Monday 6am

1 In your Google Sheet: Tools → Script editor → paste above
2 Store your token: Project Settings → Script Properties → TOKEN
3 Set trigger: clock icon → Add trigger → Weekly → Monday 6am
4 File → Share → Publish to web → Sheet: Products → CSV → Copy link
5 Paste that link into the Products CSV URL field above

Formula Settings

All scoring parameters live here. Changes apply to all products when you click Apply & Rescore. Every value drives the live scoring engine — not a spreadsheet, the actual formula running in this dashboard.

Component weights

Maximum points each signal can contribute. Must sum to 100. The weight bar below shows live distribution.

Signal 1

Sales velocity

How fast is it selling vs. catalog median. Discounted by velocity confidence (low sales = less trust).

Max points30pts

Signal 2 — confidence-weighted

Sessions-to-checkout rate

% of sessions reaching checkout, multiplied by session confidence. Low sessions = discount.

Max points25pts

Signal 3

Inventory health

Stock vs. ideal level. Zero stock is always zero points regardless of other signals.

Max points20pts

Signal 4

Size completeness

Variants in stock ÷ total variants. Hard override below the threshold in the Overrides section.

Max points15pts

Signal 5

Return rate guard

Penalises post-purchase dissatisfaction. 50%+ return rate scores zero on this component.

Max points7pts

Signal 6

Recency bonus

Benefit of doubt for newer products. Decays after the recency age cutoff set in Benchmarks.

Max points3pts

Weight distribution — must total 100

Total: 100 pts

Confidence thresholds

Below these thresholds, scores are discounted proportionally. This prevents small-sample noise from inflating scores — the fix you identified for the 30-session vs 1000-session problem.

Velocity confidence

Minimum sales for full trust

Products need at least this many sales in 30 days to earn full velocity points. Formula: MIN(1, sold_30d ÷ this value)

Full confidence at30sales

15 sales → conf 50%

Checkout confidence

Minimum sessions for full trust

Products need at least this many sessions before checkout rate is taken at face value. Formula: MIN(1, sessions ÷ this value)

Full confidence at200sessions

30 sessions → conf 15%

Benchmarks

The "median" reference values the formula compares each product against. Replace with MEDIAN() from your actual Shopify data after the first real data pull.

Velocity

Median catalog velocity

Products at or above this earn full velocity points. Replace 2.5 with your actual catalog median.

Median velocityunits/day

Checkout

Target checkout rate

Products at or above this earn full checkout points. 12% is the Indian apparel benchmark.

Benchmark rate%

Inventory

Ideal stock level

Stock above this earns full inventory points. Tune per category — lehengas vs. basic kurtas differ.

Ideal stockunits

Recency

Full bonus age cutoff

Products younger than this get full recency bonus. Bonus decays to zero over the same period beyond.

Full bonus underdays

Hard override rules

These fire before scoring. When triggered, the product is assigned directly to a zone and the score formula is skipped entirely.

Override 1

New product exemption

Products younger than this skip scoring and get status NEW. They enter normal scoring once they age past this.

Exempt if age belowdays

Override 2

Force KILL: high returns + low sales

If return rate exceeds the threshold AND units sold is below the minimum, product is forced to KILL.

Return rate above%

AND units sold belowunits

Override 3

Force PAUSE: size completeness

If fewer than this % of size variants are in stock, product is forced to PAUSE regardless of score.

Force PAUSE below% size completeness

Override 4

Seasonal hold tag

Products with this Shopify tag skip scoring. Managed via Shopify tags — applies immediately without rescoring.

Shopify tag name

Zone thresholds

Score ranges that determine zone assignment. Zones must not overlap — the dashboard enforces this on Apply.

STAR

Min:pts

Max:100

MONITOR

Min:pts

Max:69

PAUSE

Min:pts

Max:44

KILL

Min:0

Max:19

Applied — all products rescored.

Fyva Assistant

Ask about your products & scores

How many KILL products? Top 5 STAR products Worst checkout rate High returns products Low stock alerts Score distribution