Build a Procurement Shortlist from IndiaMart Suppliers (2026)

Published April 27, 2026 · 1584 words · For ops

Thirdwatch's IndiaMart Scraper makes Indian supplier shortlisting a one-call workflow at $0.002 per record — pull every supplier in your category, score by rating-weighted trust × MOQ fit × location, return a ranked Top-10. Built for procurement managers and sourcing analysts who need to compress days of supplier discovery into minutes of structured ranking.

Why build a procurement shortlist with IndiaMart

Indian procurement still runs on phone calls and word-of-mouth for most categories. According to the Government of India's MSME ministry data, more than 6.3 crore (63 million) registered MSMEs operate in India and the long tail is diffused across thousands of categories — making manual supplier discovery genuinely hard. IndiaMart aggregates roughly 10 million of those suppliers into a single searchable directory, which makes it the most efficient single starting point for any India-sourced procurement exercise.

The job-to-be-done is structured. A factory's procurement manager needs five to ten verified suppliers for stainless steel pipes 304 grade, within driving distance, with usable MOQs, all GST-compliant. A startup's ops team needs the same shape for their packaging, IT equipment, or office supplies categories. A government compliance team verifying an existing supplier panel needs a fresh data snapshot to refresh certification records. The Thirdwatch actor returns the raw input; one short pandas script produces the ranked shortlist.

How does this compare to the alternatives?

Three options for generating a procurement supplier shortlist in India:

Paid sourcing services curate supplier panels but lag IndiaMart by months and apply opaque scoring you don't control. The IndiaMart Scraper actor page gives you the live raw feed; the scoring layer is downstream pandas you can tune to your category, location, and risk tolerance.

How to build a procurement shortlist in 4 steps

Step 1: How do I authenticate against Apify?

Sign in at apify.com (free tier, no credit card), open Settings → Integrations, and copy your personal API token. Every example below assumes the token is in APIFY_TOKEN:

export APIFY_TOKEN="apify_api_xxxxxxxxxxxxxxxx"

Step 2: How do I pull the candidate supplier pool for a category?

Pass several specific queries (not one broad term) and the target city or state. Specific queries return tightly relevant results; combining several covers the long tail.

import os, requests, pandas as pd

ACTOR = "thirdwatch~indiamart-supplier-scraper"
TOKEN = os.environ["APIFY_TOKEN"]

resp = requests.post(
    f"https://api.apify.com/v2/acts/{ACTOR}/run-sync-get-dataset-items",
    params={"token": TOKEN},
    json={
        "queries": [
            "stainless steel pipes 304 grade",
            "stainless steel pipes 316 grade",
            "GI seamless pipes",
            "ERW MS pipes",
        ],
        "location": "Mumbai",
        "maxResultsPerQuery": 100,
    },
    timeout=600,
)
df = pd.DataFrame(resp.json())
print(f"{len(df)} listings across {df.company_name.nunique()} unique suppliers")

Four queries × 100 results = up to 400 listings. After dedupe, expect 150-250 unique suppliers per category sweep.

Step 3: How do I compute a procurement-fit score per supplier?

Combine rating-weighted trust, GST presence, vintage (member_since), and listings count into a single score:

import datetime as dt
import numpy as np

now_year = dt.date.today().year

agg = (
    df.dropna(subset=["company_name"])
    .groupby("company_name", as_index=False)
    .agg(
        n_listings=("product_name", "count"),
        city=("city", "first"),
        state=("state", "first"),
        gst_number=("gst_number", "first"),
        phone=("phone", "first"),
        member_since=("member_since", "first"),
        supplier_rating=("supplier_rating", "first"),
        rating_count=("rating_count", "first"),
        catalog_url=("catalog_url", "first"),
    )
)

agg["rating"] = pd.to_numeric(agg.supplier_rating, errors="coerce")
agg["nrate"] = pd.to_numeric(agg.rating_count, errors="coerce").fillna(0)
agg["years_active"] = (now_year - pd.to_numeric(agg.member_since, errors="coerce")).clip(lower=0).fillna(0)
agg["has_gst"] = agg.gst_number.fillna("").ne("").astype(int)

# Composite score
agg["trust"] = agg.rating.fillna(0) * np.sqrt(agg.nrate)
agg["score"] = (
    agg.trust * 1.0
    + np.log1p(agg.years_active) * 5.0
    + agg.has_gst * 8.0
    + np.log1p(agg.n_listings) * 2.0
)

shortlist = (
    agg[(agg.nrate >= 20) & (agg.has_gst == 1)]
    .sort_values("score", ascending=False)
    .head(10)
)
print(shortlist[["company_name", "city", "rating", "nrate",
                 "years_active", "n_listings", "score"]])

The composite score weighs rating-confidence highest, then GST verification, then years of IndiaMart presence, then catalog depth. Tune the coefficients to your category's risk profile — high-spend categories should weight years_active more heavily.

Step 4: How do I export the shortlist for RFQ outreach?

Drop the top 10 to a CSV ready for RFQ population, and dial out via Twilio for phone-first outreach (the Indian SMB norm).

shortlist[["company_name", "city", "state", "phone",
           "gst_number", "rating", "nrate", "years_active",
           "n_listings", "catalog_url"]].to_csv("rfq-shortlist-pipes.csv", index=False)

# Optional: open Twilio call to top 10 for verbal RFQ
from twilio.rest import Client
twilio = Client(os.environ["TWILIO_SID"], os.environ["TWILIO_TOKEN"])
for _, row in shortlist.iterrows():
    if pd.isna(row.phone) or not row.phone:
        continue
    twilio.calls.create(
        to="+91" + str(row.phone),
        from_="+1XXXXXXXXXX",
        url="https://your-twiml-endpoint/rfq-pipes-stainless.xml",
    )

A 10-supplier shortlist with phone-dialed RFQs typically returns 6-8 quotes within 48 hours — the canonical Indian SMB sourcing rhythm.

Sample output

A single record from the dataset for one supplier listing looks like this. The shortlist scoring stitches many such rows together.

{
  "company_name": "Tata Steel Ltd.",
  "product_name": "Stainless Steel Pipes 304",
  "price": "Rs 180 / Kg",
  "moq": "500 Kg",
  "product_url": "https://www.indiamart.com/tata-steel/stainless-steel-pipes.html",
  "catalog_url": "https://www.indiamart.com/tata-steel/",
  "image_url": "https://5.imimg.com/data5/SELLER/Default/.../stainless-steel-pipes.jpg",
  "location": "Mumbai, Maharashtra",
  "city": "Mumbai",
  "state": "Maharashtra",
  "phone": "8044464742",
  "gst_number": "27AAACT2727Q1ZV",
  "supplier_rating": "4.5",
  "rating_count": "1240",
  "member_since": "1998"
}

A typical shortlist row aggregated from many such records looks like this:

The composite score is the column to sort on; the underlying components let a procurement manager justify the ranking to a sceptical CFO.

Common pitfalls

Three things go wrong in IndiaMart-driven shortlisting. Single-listing suppliers — a supplier that appears only once in your dataset (low n_listings) might be a true specialist or a stub profile; visit the catalog_url to verify before including. Year-of-founding noise — member_since is the year the supplier joined IndiaMart, not the year the business started; a 1998 supplier number doesn't always mean a 28-year-old business. Treat as a soft signal, not absolute proof of vintage. Composite score weight tuning — the coefficients in the score formula above are reasonable defaults but should be tuned for your category. Capital-intensive sourcing (machinery, large-volume metals) should weight years_active more; commodity sourcing (plastics, packaging) should weight n_listings and nrate.

Thirdwatch's actor returns 15 fields per record so the shortlisting logic can stay in your code where it belongs. The pure-HTTP architecture means a 400-record category sweep completes in under five minutes and costs $0.80, leaving comfortable budget for re-runs as the shortlist evolves.

Related use cases

• Find B2B suppliers in India with IndiaMart
• Monitor wholesale prices in India with IndiaMart
• Verify Indian suppliers' GST and MOQ at scale
• The complete guide to scraping business data
• All Thirdwatch use-case guides

Frequently asked questions

What's the right shortlist size for an Indian procurement RFQ?

Five to ten suppliers per category is the standard. Below five, you don't have meaningful price competition; above ten, evaluation effort grows faster than information value. The actor returns enough metadata per supplier (rating, member_since, GST, MOQ, location) that a 200-supplier raw pull collapses cleanly to a 5-10 shortlist with one pandas score-and-sort.

How do I weight rating versus rating count in shortlisting?

A common pattern is sqrt-weighted score: rating × sqrt(rating_count). This balances a 5.0/3-rating supplier (high rating, low confidence) against a 4.5/200-rating supplier (lower rating, much higher confidence). The 4.5/200 wins, which is correct for procurement risk. Add a hard floor of rating_count >= 20 to filter out unreviewed suppliers.

Should I include suppliers without a published GST number?

For most procurement workflows, no. A missing GST number signals either an unregistered business or an incomplete IndiaMart profile — both are higher risk than a registered supplier with verified GST. The exception is when sourcing from very small businesses where unregistered status reflects size, not non-compliance. Most procurement teams require gst_number for any contract above ₹50,000.

How do I cross-reference MOQ across suppliers when units differ?

MOQ comes back as a display string with mixed units (Kg, Pieces, Tonne, Roll, Meter). Build a parser that extracts numeric value and unit, then convert to a canonical unit per category. For metals, normalise to Kg. For textiles, normalise to Meters. The IndiaMart raw moq is preserved in case the parser misses an edge case.

Can I automate the entire RFQ generation flow?

Partially. The actor produces the supplier shortlist; the RFQ document generation is downstream. Most procurement teams export the shortlist to Excel, populate an RFQ template, and email it to suppliers. For phone-first outreach (common in India), use the public phone numbers the actor returns and route to your call-centre or Twilio dialer with the shortlisted ranking.

How fresh does the data need to be for a procurement decision?

A single fresh pull at the start of the sourcing exercise is enough. Indian SMB supplier listings are stable — most don't change company_name, GST, or location for years. Prices and MOQ shift more frequently but are rarely the deciding factor in shortlisting. For ongoing supplier monitoring (price changes, rating drift), schedule a weekly re-pull.

Run the IndiaMart Scraper on Apify Store — pay-per-supplier, free to try, no credit card to test.