Every GST business contact database in India is a join. The GST register publishes no phone numbers and no email addresses, so a file containing a GSTIN in one column and a mobile number in the next has combined two datasets. Understanding which second dataset — and how the match was made — tells you more about what you are buying than any accuracy percentage on the sales page.
There are four joining methods in common use. They have very different fill rates, very different error profiles, and very different compliance exposure. This guide describes each one, what it looks like in a sample file, and the questions that reveal which you have been sold.
Method 1: MCA cross-reference
The Ministry of Corporate Affairs publishes a company email address in company master data, alongside registered address, incorporation date, capital and director details. Because both GST and MCA registrations are linked to PAN, a GSTIN can be matched to a CIN with high confidence.
| Property | Assessment |
|---|---|
| Match reliability | High — PAN-based, not name-based |
| Fill rate on the GST universe | Low. Only companies and LLPs are in MCA at all |
| Field quality | Email is filed by the company itself; often a CA's or founder's address |
| Freshness | Updated when the company files; can be years old |
| Phone numbers | Not provided by MCA |
This is the most defensible enrichment available, and the least commercially useful on its own, because the great majority of GST registrants are proprietorships and partnerships that have no MCA presence whatsoever. The full boundary is mapped in GST data vs MCA company data.
How to spot it in a sample: email fill rate is decent for private limited companies and near zero for proprietorships; phone column is empty or sourced differently.
Method 2: Directory and listing matching
Business directories and marketplaces — IndiaMART, JustDial, TradeIndia, trade association listings, local directories — carry phone numbers that businesses published themselves. Providers match a registered business name and address against these listings.
| Property | Assessment |
|---|---|
| Match reliability | Moderate to poor. Name and address matching is fuzzy by nature |
| Fill rate | Moderate, and heavily skewed toward businesses that market online |
| Field quality | The number is real, but may belong to a different branch or a since-closed listing |
| Freshness | Depends on when the listing was last updated by the business |
| Coverage bias | Systematically over-represents traders and manufacturers, under-represents services |
The failure mode here is the false positive: two businesses with similar names in the same district get merged, and you call the wrong company. It is invisible in a fill-rate statistic and obvious the moment your sales team starts dialling.
How to spot it in a sample: pick five records and search the business name plus city. If the number you were sold is the first result on a directory site, you know the method.
Method 3: Website and public-web extraction
Numbers and emails published on the business's own website, extracted at scale and matched by domain, name or address.
| Property | Assessment |
|---|---|
| Match reliability | Good when a domain is confidently linked to the entity, poor otherwise |
| Fill rate | Low across the newly-registered segment — most new small businesses have no website |
| Field quality | Usually a genuine business contact, often a generic info@ inbox |
| Freshness | As fresh as the website |
| Compliance posture | Strongest of the four, when limited to genuine business contact points |
How to spot it in a sample: emails skew to role addresses (info@, sales@, contact@) on the company's own domain rather than free webmail.
Method 4: Third-party contact panels
Purchased or licensed contact databases, matched on name, address or PAN. This is where most of the volume in cheap "GST leads with mobile numbers" files comes from, and it is the least disclosed.
| Property | Assessment |
|---|---|
| Match reliability | Unknown to the buyer, and usually to the reseller |
| Fill rate | High — which is precisely why it is used |
| Field quality | Unverifiable without dialling |
| Provenance | Frequently undocumented several layers back |
| Compliance exposure | Highest. You cannot demonstrate lawful sourcing for data whose origin nobody can name |
The question that ends the conversation
"For the mobile number column specifically, name the source and tell me the date it was collected." A provider using method 1, 2 or 3 can answer immediately. A provider using method 4 will offer a sample, a discount, or a sentence about proprietary technology. That is your answer.
What an honest field table looks like
Any provider can produce this. Very few do unprompted. Ask for it before you ask for a price:
| Field | Source category | Last checked | Fill rate | Known limitation |
|---|---|---|---|---|
| GSTIN | GST registration | Per refresh cycle | ~100% | Status changes after collection |
| Legal name | GST registration | Per refresh cycle | ~100% | May differ from trading brand |
| Registration date | GST registration | Per refresh cycle | ~100% | — |
| Principal address | GST registration | Per refresh cycle | ~100% | Often a residence or CA's office |
| Company email | MCA master data | Filing date | Companies/LLPs only | May be the auditor's address |
| Business phone | Directory match | Collection date | Varies widely | Fuzzy name/address match |
| Website | Web extraction | Collection date | Low in this segment | — |
Note that the "last checked" column is per field, not per file. A single file-level date tells you nothing, because the registration columns and the phone column were almost certainly collected months apart.
Fill rate is the number that matters
Providers quote accuracy. Accuracy is unfalsifiable at purchase time. Fill rate is not, and it drives your actual economics:
A file of 10,000 newly registered businesses at ₹15,000 looks like ₹1.50 per record. If the phone column is 30% populated and a fifth of those are wrong or unreachable, you have roughly 2,400 usable records — ₹6.25 each. A file at ₹40,000 with 65% fill and better matching yields around 5,500 usable records at ₹7.27, with less wasted rep time on dead numbers. The cheaper file is not obviously the cheaper file.
Run this calculation before every purchase. The method, including how to test fill rate from a sample rather than taking it on trust, is in the B2B data quality checklist, and pricing structures across the market are compared in what GST lead data costs in India.
The compliance dimension you cannot skip
A proprietor's mobile number is personal data. It does not stop being personal data because it was matched to a business registration, and the "publicly available" carve-out in India's data protection framework is narrower than vendors imply — it turns on data the individual themselves made public, not on data that became inferable through a match.
Practically, this means your procurement needs to cover: documented source per field, a lawful basis you can articulate, a suppression and correction workflow, and a contractual position with the vendor on where the data came from. All of it is in the DPDP checklist for B2B data buyers, and the separate telecom rules that govern actually calling or messaging those numbers are in the TRAI and DND guide.
Common questions
Is any GST contact database "government verified"? No. Contact fields do not come from government sources, so no part of that phrase applies to them. The registration fields are government-published; the phone number sitting next to them is not.
Which method gives the best data? For companies and LLPs, method 1 combined with method 3 is the most defensible. For the proprietorship long tail — which is most of the register — method 2 is usually the only option that produces volume, and it needs verification before use rather than after complaints.
Should I buy a file or use an API? A file is a snapshot that starts decaying immediately; an API lets you check status at the moment of use. For anything where currency matters, the API and CRM integration guide covers the trade-offs.
How do I check a specific GSTIN myself? Free, on the official portal. The procedure and what each returned field means is in how to look up a company from its GST number.