This is the shape of what comes back. Request a live sample and the same columns arrive filled in for GST numbers taken from your own list.
The fourth row is a cancelled registration with no phone number against it. It is there because that is what a real delivered file looks like. A provider who shows you five immaculate rows is showing you a brochure rather than a sample, and the gap between that brochure and the file you receive after paying is where most of the disappointment in this industry lives.
The second is the masked phone column. Live samples return complete numbers. We mask them here because publishing working business telephone numbers on a public, indexable web page is exactly the sort of thing a data provider should not do, and a company that would do it to demonstrate a point would do it with your file too.
Choose numbers you already know something about. That is the whole point of the exercise. If the returned legal name, status and phone number match what you already know about those five businesses, the remaining forty-nine thousand rows are worth trusting. If they do not, you have learned that for free.
The table above shows six columns because six fit comfortably on a page. A delivered file carries more: trade name alongside legal name, the full principal place of business with PIN code, taxpayer type, date of registration, email address where one exists, and the effective date attached to any cancelled or suspended status.
Your own original columns sit to the left of all of it, untouched and in their original order, so a file that already drives a dashboard or a mail merge keeps working exactly as it did before.
A few suggestions from customers who evaluate data for a living. Include at least one business you know has closed or deregistered, because how a provider handles that row tells you more than five active ones. Include one very small proprietorship, since coverage is always thinnest at that end and a file that only performs on large companies may not suit your market. And include one business whose registered legal name you know differs from its trading name, to confirm you are being given filed data rather than something scraped off a website.
| GSTIN | Legal Name | State | Status | Phone | Constitution |
|---|---|---|---|---|---|
27ABCDE1234F1Z0 | Sample Traders Pvt Ltd | Maharashtra (27) | Active | +91 98XXX XXXXX | Pvt Ltd |
29ABCDE5678G1Z6 | Example Industries LLP | Karnataka (29) | Active | +91 80XXX XXXXX | LLP |
07ABCDE9012H1ZQ | Demo Enterprises | Delhi (07) | Active | +91 11XXX XXXXX | Proprietorship |
33ABCDE3456J1ZM | Illustration Textiles | Tamil Nadu (33) | Cancelled | — | Pvt Ltd |
24ABCDE7890K1Z5 | Placeholder Chemicals Ltd | Gujarat (24) | Active | +91 79XXX XXXXX | Public Ltd |
Illustrative rows. The GST numbers are placeholder values, the company names are invented and the phone numbers are masked. A live sample returns real values for GST numbers you supply.
Contact coverage is not evenly distributed, and knowing the shape of it before you order prevents most disappointment. As a general pattern, registrations belonging to private limited companies, public limited companies and LLPs carry contact details more consistently than small proprietorships, because the registration was more likely handled by an accountant who completed every field.
Newer registrations also tend to carry better contact data than very old ones, for the obvious reason that less time has passed for the number to change. And metropolitan registrations generally do better than remote ones. None of these are rules, but they are reliable enough that if your target segment sits at the thin end of all three, we will tell you before you order rather than after.
Two practical points. Import the GSTIN as a text field rather than a number, or your spreadsheet software will mangle codes beginning with a zero – state codes 01 through 09 are the ones that break. And map registration status to a field your team can actually filter on, because the ability to exclude cancelled registrations in one click is most of the value of having the column at all.
Registration status changes continuously, and a contact file ages faster than most buyers expect. For an active outbound operation, refreshing the segment you are working on a few times a year keeps the status field meaningful. For a database enrichment use case, a scheduled refresh costs less per record than repeatedly buying new files, which is why it exists as an option.
Occasionally a registration record carries a number that is present but incomplete – a digit short, a landline without its STD code, or a mobile entered with a country prefix duplicated. These are flagged rather than silently corrected, because guessing a missing digit is exactly the kind of helpfulness that produces a wrong number. Where the fix is unambiguous, such as a duplicated country code, it is applied and noted. Where it is not, the row is marked so you can decide whether it is worth chasing.
If you are buying data for the first time, show the returned sample to whoever will actually be calling it. Callers spot problems that buyers miss, because they know immediately whether a legal name looks like something they can open a conversation with and whether the constitution type tells them who to ask for. Five minutes of that conversation before you order is worth more than any amount of vendor documentation.
Save the sample you receive. When the full file arrives, spot-check the same rows against it. They should match, and if they do not, that is a question worth asking immediately rather than three weeks into a calling campaign. It is a two-minute check that catches the rare but real problem of a file being assembled to a different specification from the sample that sold it.
One last suggestion. Ask for the sample in the same file format you intend to work in, rather than whichever is convenient. Column handling, character encoding and the way leading zeros survive an export all behave differently between CSV and Excel, and it is far better to discover a formatting problem on a five-row sample than on the full file.
Yes, and it does not require card details or a call. It exists because it is the fastest way for both sides to find out whether this is worth doing.
Five to ten is the usual range. That is enough to judge quality and small enough that we can turn it round quickly.
Yes, and you are not charged twice for them.
Then you have learned something valuable at no cost, and you should not place the order. We would rather tell you that up front than sell you a file you cannot use.
The fastest way to judge a data provider is to test it on records you can already verify yourself.