A bank deciding whether to lend to someone starts by pulling their credit history. You, standing at the counter of an electronics shop in Rawalpindi with a customer asking for twelve months on a refrigerator, cannot.
That asymmetry is the central risk of installment retail in Pakistan, and no amount of instinct closes it. What does close it — slowly, and only if you are deliberate about it — is building the one credit file nobody else has: your own.
What the bureau is, and why it is not available to you
Pakistan does have a central credit information system. The State Bank of Pakistan's published material on its electronic Credit Information Bureau (eCIB) describes it as a database that collects borrower credit data from its member financial institutions — banks, development financial institutions, non-bank financial companies and microfinance banks — and makes Credit Information Reports available to those member institutions for their own credit assessment. SBP's published eCIB FAQ further states that the report is a confidential document and that access to generate it sits with member institutions.
A retail shop is not a member institution. There are also licensed private credit bureaus operating in Pakistan, whose access arrangements are a matter for those bureaus and the institutions that subscribe to them.
The practical consequence for your counter, as of November 2026, is simple and worth stating plainly rather than hoping around: no bureau report is available to you at the point of sale. Everything below assumes that constraint rather than trying to work around it.
Sources worth reading yourself: SBP's eCIB FAQ and its list of eCIB member institutions.
Five signals a counter can actually verify
You cannot see a credit score. You can see these, and each takes minutes.
1. Identity, properly. Original CNIC in hand, photograph matched to face, name matched to signature. The overwhelming majority of total losses — where the goods and the person both vanish — begin with identity that was never actually checked.
2. Address that someone could reach. A utility bill or tenancy document, plus a landmark description good enough for your staff to find the house without the customer guiding them. Treat an unverifiable address as a decisive negative, not a minor gap.
3. Stability, not income. Shopkeepers often ask what someone earns and get a number that cannot be checked. Stability is more useful and more verifiable: how long at this address, how long in this work, how long on this phone number. A modest income in the same place for six years is a better risk than a high income three months into a new city.
4. A phone number with history. Ask how long they have had the number. A long-held number is an asset the customer does not want to abandon, which makes it a quiet form of security. A number activated last week is worth noting.
5. Their own account of why installments. Let them explain. "My fridge died and I get paid monthly" is a different proposition from "everyone else refused me." Customers tell you a great deal when simply asked, and it costs nothing to listen.
None of these is decisive alone. Together they separate the customer who is inconvenient from the customer who is dangerous.
Your real asset: the payment history you already own
Here is the part most shops waste. Every installment account you have ever run is a credit record — and it is better data than a bureau report, because it describes behaviour with you, on goods like yours, in your own market.
The condition is that it must be retrievable at the moment of the next sale. A register full of closed accounts is a historical document. A customer file that surfaces when that person walks back in is a credit check.
So make one rule absolute: nobody gets a new plan until their previous plan has been looked at. Not remembered — looked at.
What to capture per account, so that the history is usable later:
- Was each installment on time, late, or missed — not just whether the account eventually cleared.
- How many follow-up contacts it took, and whether promised payments were kept.
- How the account closed: paid in full, early settlement, restructured, repossessed, or written off.
- Who the guarantor was, and whether the guarantee had to be invoked.
A customer who paid everything but needed eleven reminders and three broken promises is not the same risk as one who paid twelve installments untouched. Both show as "cleared" in a register. Only the first will cost you again.
A four-tier model that staff can actually apply
Scores are for banks. At a counter you need categories that a salesperson can read in two seconds and act on:
| Tier | What it means | What happens at the counter |
|---|---|---|
| On track | Paying on schedule, no broken promises | Normal terms |
| Needs attention | Occasional lateness, pays after a reminder | Normal terms, tighter follow-up |
| At risk | One to two installments behind, or repeated broken promises | Higher advance, shorter tenure, stronger guarantor |
| Defaulter | Stopped paying or unreachable | No new credit without full settlement |
Why the worst status matters more than the current one
A customer who defaulted for four months and then cleared their balance will show as "on track" today. Current status is honest about the present and silent about the pattern — and the pattern is what you are actually pricing.
So keep the worst status the account ever reached, permanently, alongside the current one. Someone who once sat in the defaulter tier is not barred from buying again, but they should be buying on a higher advance and a shorter tenure. Lifetime risk memory is standard in the Markaz Flow account record for this reason: a clean present should not erase an expensive past.
The same logic applies to how you close accounts. Distinguish "paid in full", "settled early", "restructured", "repossessed" and "written off" instead of collapsing them into "closed". Those five words are your credit history.
It compounds, which is why it is worth starting badly
The uncomfortable truth is that this file is worthless on day one and formidable in year three. A shop with four years of recorded behaviour on two thousand customers has something no bureau can sell it and no competitor can copy.
Which means the only real mistake is waiting for a tidy system before you start recording. Capture identity properly, record payment behaviour as it happens rather than reconstructing it, and refuse to open a second plan without reading the first.
For the pre-sale checks themselves, read our guide to verifying a guarantor who will actually pay, and for what to do once an account is already behind, our five-step recovery system. If you would rather this history built itself as you trade, start free on Markaz Flow.