Data Assessment: Can You Trust Your Numbers?

Assessing your organisation’s data: can you trust your numbers?

Many business decisions rest on numbers whose source nobody knows precisely: an Excel file edited by several people, or a notebook that hasn’t been updated for weeks. A data assessment answers a simple question: can you rely on your numbers for pricing, purchasing and collection? Here is what to check and how to check it.

Digital maturity assessment 6 min read

By Nour Aldein Tleti. Updated 7 October 2026.

How do you know whether your company’s data is reliable?

Reliable data is accurate, complete, free of duplicates and up to date, with one approved source that everyone knows. The simplest test is to ask two people for the same number, such as this month’s sales or a customer’s balance, and compare the answers. If they differ, the problem lies in the data or in how it is collected, and it must be fixed before you build reports or buy a system.

Five criteria for data quality

Accuracy: Does the recorded figure match reality? Compare a sample of stock balances with the shelves, and a sample of customer balances with their statements.
Completeness: Are the essential fields filled in? A customer without a phone number or a product without a purchase price leaves the analysis incomplete.
No duplicates: The same customer recorded under two or three names with different spellings means their balance appears split and you don’t know their true value.
Timeliness: When was the figure last updated? A sales report compiled once a month doesn’t help with a daily decision.
Consistency: Are dates, units and names written the same way in every file? Inconsistency makes merging and comparing difficult.

Where data problems usually hide

Several copies of the same file: A “Final sales” file and a “Final sales 2” file on two different computers, and nobody is sure which is newer.
Data on phones: Customer numbers and orders sit in personal WhatsApp chats, never enter any record and never appear in any report.
Two currencies without a rate and date: Amounts in Syrian pounds and in dollars in the same column, or conversions without the exchange rate used, make month-to-month comparisons misleading.
Late entry: Invoices entered at the end of the week from memory or from piled-up papers, so details get lost and errors creep in.
Free-text fields: Typing the city or product as free text instead of choosing it from a fixed list produces many spellings of the same thing.

Who owns the data?

An owner for each type of data: Name one person responsible for the quality of customer data, another for stock and another for accounts.
Written entry rules: Agree how names, numbers, currencies and dates are written, and put the rules where the people entering data can see them.
Clear permissions: Not every employee needs to edit everything. Restricting edits reduces errors and clarifies responsibility.
An edit log: Use tools that record who changed a figure and when, so an error can be traced to its source and fixed quickly.

Steps to assess and fix your data

01
List your data sources
Write down every notebook, file, program and page where customer, sales, stock and accounting information is recorded.
02
Test with a sample
Pick a small sample of customers, products and invoices, and compare them with reality and with the other sources.
03
Classify the problems
Separate errors from gaps, duplicates and delays, because each type needs a different fix.
04
Clean before you migrate
Correct and standardise the core data before moving it into any new system; otherwise you will move the same mess.
05
Settle on one source
Adopt one source for each type of data and phase out side files gradually.

Read also

Digital maturity assessment

Frequently asked questions

Do we have to clean all our old data?

Not necessarily. Focus on what you need in daily work, such as active customers, products in circulation and open balances, and keep the rest as an archive.

How do we record amounts in two currencies correctly?

Record each amount in its original currency, together with the exchange rate used, its date and the source you have approved internally, so you can recalculate and compare later. Also check current local legal requirements on accounting and currency.

Is Excel enough for reliable data?

It may be enough for a small company if it is a single shared file with clear entry rules and a backup. Problems start when files and copies multiply.

How often should we assess data quality?

Make sample checks a regular habit, for example monthly for sensitive data such as balances and stock, and repeat the full assessment before any new system project.

Where do we start if the data is a complete mess?

Start with the customer or product list, because everything else rests on it. Standardising that alone visibly improves your reports.