Data Quality & Cleansing
Data cleansing services take the customer, vendor and product records your teams already hold — in a CRM, an ERP or a spreadsheet export — and return them deduplicated, standardised, validated and documented, with a reviewer’s sign-off on every batch. As a Dhaka-based company, we work inside your own systems, for international companies, at a price per 1,000 records, with a QA lead who re-checks a sample of every batch before it goes back to you.
One exit at each step. No long-term commitment at any of them. We start with a free sample: send 500 records and get them back cleaned, with a change log and a count of the duplicates, invalid emails and missing fields found. The full file is then profiled and priced per 1,000 records in writing, before any record is touched.
You get a written verdict — manual cleansing, automated rules with review, a pipeline instead, or nothing yet. If the sample shows little to fix, you stop here.
2 days 01 Scope · free 500-record sample
3–5 days 02 Diagnose · profiling audit, credited to the pilot
1 week 03 Pilot · $1,200 fixed, one batch
In batches 04 Production · from $3,600, quoted after the pilot
Monthly 05 Managed Ops · priced monthly, optional
What is included in data cleansing services?
Data cleansing services include six deliverables: an audit that counts what is wrong, deduplication on the match keys you agree, standardisation of names, addresses and dates, validation of emails, phones and postcodes, enrichment of missing fields from sources you approve, and a change log with a rejected-records list. You get a file you can load, not a report about one.
Audit and profiling
Every field profiled before work starts: duplicate rate, invalid emails, missing fields and format mix, and a written count to check the result against.
Database deduplication
Duplicates found on the match keys you agree — email, phone, company plus address — merged into one surviving record, merge rule recorded.
Standardisation
Names, job titles, company names, addresses, phone numbers and dates put into one agreed format, so filters, mail merges and imports stop breaking.
Validation and verification
Email syntax and domain checks, phone format checks, postcode-to-city checks; records that fail are flagged in a rejected list, never silently deleted.
Enrichment
Missing company, title or address fields filled only from sources you approve and can name; nothing scraped, nothing guessed, every added value marked.
Change log and sign-off
A row-level log of every change, the rejected-records list, and a QA lead’s sign-off on a re-checked sample, delivered with every batch.
Not included: Third-party contact lists · enrichment from sources you have not approved · scraping · cleaning that must recur every week, which is Managed Ops, priced monthly.
Sales and marketing
CRM contacts and leads deduplicated, emails validated and titles standardised before the next campaign or territory split.
Finance
Vendor and customer master records merged, bank and tax fields validated, so payments reach one correct entity.
Operations
Product, SKU and supplier lists standardised across ERP, e-commerce and spreadsheets so stock and orders match.
HR and payroll
Employee and candidate records deduplicated, addresses and IDs validated, with personal fields masked while reviewers work.
What does data cleansing change in a CRM record, before and after?
Data cleansing changes a CRM file record by record: duplicates are found on the match key you agree, names, companies and phone numbers are put into one format, emails are validated, and anything that fails is listed rather than deleted. In the example below, three versions of one contact become one surviving record and one invalid email is set aside.
Before · CRM export, 4 records
| Name | Company | Phone | |
|---|---|---|---|
| jane doe | ACME ltd | jane.doe@acme.com | 5550142 |
| Jane Doe | Acme Limited | JANE.DOE@ACME.COM | (555) 0142 |
| J. Doe | Acme | jane.doe@acme.com | — |
| Mark Lee | Northwind | mark@northwind | 555 0199 |
3 versions of one contact
1 invalid email
Rules agreed in writing
Match key: email + phone
Formats: name case, company suffix, phone
Validation: email syntax and domain
Reviewer re-checks a sample · a person signs
After · 1 surviving record, 1 listed
| Name | Company | Phone | |
|---|---|---|---|
| Jane Doe | Acme Ltd | jane.doe@acme.com | 555-0142 |
Rejected listMark Lee · mark@northwind · invalid email domain, listed for you to check, never deleted
Change log3 records merged on email + phone · company and phone formats standardised · every change with record ID, rule and reviewer
Managed Ops cleans new and changed records every week or month, with a monthly quality report.
Illustrative example. Yellow marks the step where a person decides; nothing is deleted without being listed for you.
How do data cleansing services work, from free sample to Managed Ops?
Data cleansing services with us run in five steps: a free 500-record sample, a profiling audit of the full file with a written price per 1,000 records, a pilot batch cleaned and signed off, production on the whole file in batches, then Managed Ops for the records that keep arriving. Every step ends with a result you can stop on.
01 2 days · free
Scope Send 500 records; they come back cleaned, with a change log and counts of duplicates, invalid emails and missing fields, before you pay anything.
02 3–5 days · credited
Diagnose The full file profiled field by field; match keys and formats agreed with you in writing; the price per 1,000 records and the batch plan quoted.
03 1 week · fixed price
Pilot One batch of up to 10,000 records cleaned, logged and re-checked on a random sample by the QA lead; you approve the rules before production.
04 In batches · quoted after pilot
Production The whole file in agreed batches, each with a change log, rejected list and sign-off; rules loaded into your CRM’s own dedupe tools where available.
05 Monthly · optional
Managed Ops New and changed records cleaned every week or month, a monthly quality report with duplicate and invalid rates, and rules updated as sources change.
How long does a first cleaned batch take?
A free 500-record sample comes back in two working days; the profiling audit takes three to five days, and a first pilot batch of up to 10,000 records about a week.
Turnaround Free 500-record sample in two working days; a pilot batch in about a week
Tools Spreadsheets, SQL, OpenRefine and your CRM’s native dedupe tools
Communication Report per batch, shared channel, one named contact
Data CRM exports, ERP master data, marketing lists, spreadsheets
QA Random-sample re-check on every batch, QA lead’s sign-off
Hosting Your CRM or your shared drive; no copies kept
Do you need manual data cleansing, automated rules or a pipeline?
Not every dirty file needs the same data cleansing fix. A few thousand records with judgement calls need reviewers; a hundred thousand records with repeatable errors need rules and a reviewed sample; records that get dirty again every month need a pipeline, not a one-off clean. Five questions show which one fits, before anyone quotes a price.
1. How many records need cleaning?
2. What is mostly wrong with them?
3. Where do the records live?
4. How often do new records arrive?
5. Do the records contain personal data?
Reviewers plus rules, piloted on one batch
5,000 to 100,000 records with duplicates and format errors: rules do the bulk, reviewers handle the exceptions, and a 10,000-record pilot batch shows the real rate.
Sample cleansing run log · CRM contact file
1
Records received48,200 contacts from a CRM export · 3 sources
2
Profiled11% duplicates, 6% invalid emails, 9% no title
3
Rules appliedDedupe on email plus phone, name and address formats
4
Reviewer sign-off2% random sample re-checked · 3 rules corrected
5
DeliveredClean file, change log, 2,900 rejected records listed
Illustrative run
Yellow = a person signs
Every change logged
How the verdict is decided: daily arrivals from several systems → a pipeline · under 5,000 records, or judgement calls → manual cleansing by reviewers · over 100,000 records with repeatable errors → automated rules with a reviewed sample · everything in between → reviewers plus rules, piloted on one batch.
Why choose us as your data cleansing company?
A data cleansing company is judged on the counts it can show before and after, not on adjectives. We profile every file first, price per 1,000 records in writing, and deliver data scrubbing services with a change log, a rejected-records list and a signed sample check on every batch, which supports the accuracy principle of the GDPR.
Without a counted process
!!!!!
- A “cleaned” file with no change log, so nobody knows what changed
- Duplicates merged on the wrong key, so two real customers become one
- Records deleted silently instead of listed as rejected for you to check
- An hourly price that nobody can convert into a price per file
With EICRA
Batch report · CRM contacts
Received48,200Duplicates11%Invalid6%Sample check2%VerdictBatch approved
Illustrative example
- A profiling count before the work and a matching count after
- Match keys and formats agreed in writing before any record changes
- A rejected list and a row-level change log with every batch
- A price per 1,000 records, quoted in writing after the sample
Is it safe to send customer records to a data cleansing team?
Sending customer records to an offshore data cleansing team is safe when access, transfer and deletion are settled in writing before any file moves. As a Bangladesh-based company, we work inside your CRM or a shared workbook you control, see personal fields only where the cleansing needs them, sign a data processing agreement first, and delete every working copy at sign-off. Reviewed By Eicra.com team
Which data cleansing agreements are signed, and when?
NDA — mutual, signed before any file, export or credential is shared with anyone.
DPA — processor terms under Article 28(3) of the GDPR and the equivalent national law, for any personal data.
International data transfers — standard contractual clauses or the transfer instrument your jurisdiction requires, signed before personal data moves.
Access — a least-privilege user in your CRM, or a shared workbook you own; credentials never leave your accounts.
Certifications — listed only when held; none are claimed on this page or in any proposal.
What data cleansing controls, IP and rework terms apply?
Your systems Work happens inside your CRM or your shared drive; every working copy is deleted at sign-off, in writing.
Personal fields Masked wherever a rule does not need them; reviewers see names and emails only for the fields being cleaned.
Change log Every changed value logged with the record ID, the rule and the reviewer, so any change can be reversed.
Ownership The clean file, change log, rejected list and rules are yours under the contract; our access is removed at sign-off.
Rework Free when a batch misses the agreed sample-check threshold; new fields or sources priced first as a change.
What proof do you get before you pay for data cleansing?
Before you pay for data cleansing, you get evidence instead of promises: measured commitments, a free 500-record sample cleaned before any quote, and a signed sample check on every batch. Client case studies with numbers are added here as clients give permission to name them.
2 days
For a free 500-record sample, returned cleaned with a change log and counts before any quote.
Every batch
Re-checked on a random sample and signed off by the QA lead, with a rejected-records list.
Per 1,000
Records: the price is written after the profiling audit, before any record is touched.
Case studies: client results with numbers are added here as clients give permission to name them. Ask on the scoping call for references in your industry.
What do buyers ask a data cleansing company?
How much do data cleansing services cost?
Manual data cleansing is priced per 1,000 records, quoted in writing after a free 500-record sample and a profiling audit, so the price reflects your duplicate and error rates. Automated cleansing with a reviewed sample is fixed per project: the pilot batch price is on the price cards at the top, and production is quoted after the pilot.
What is data cleansing and why is it important?
Data cleansing finds and fixes duplicate, invalid and incomplete records so every team, report and dashboard works from one correct copy. It matters wherever a record is typed in more than once: the same customer twice in the CRM, a vendor paid under two names, a candidate emailed at a dead address. Campaigns and payments go wrong because of them.
What is the difference between manual and automated data cleansing?
Manual cleansing is reviewers working record by record against agreed rules — right for small files and judgement calls, such as which of two companies is the real one. Automated cleansing is rules written by an engineer and run on the whole file, with reviewers checking exceptions and a random sample. Most files over 100,000 records need both.
Can you clean data inside HubSpot, Salesforce or Dynamics 365?
Yes. Where your CRM has native merge and dedupe tools we use them, so the change history stays inside your system; where it does not, we work on an export in a shared workbook you control and load the result back through the CRM’s own import. Nothing is copied to systems of ours.
Who owns the cleaned data, and how is our customer data protected?
You do. The clean file, change log, rejected list and rules are yours, and the work happens inside your CRM or a workbook you control. Our offshore team signs a data processing agreement before any personal field is seen, works through least-privilege access, and deletes every working copy at sign-off.
Send 500 records for a free cleaned sample before any quote.