Skip to content
InteliSense IT, Navigating Change, Delivering Value

Data migration and data quality

The system goes live and
nobody trusts the numbers.

A new system can only be as good as the information you put in it. The hard part is not moving the data. It is deciding what deserves to move, agreeing what it means, and being able to prove the balances before anybody signs anything.

See how it works

We will tell you when the safest migration is a smaller one.

Sound familiar?

Bad data does not get better because it arrives successfully.

These are the things finance directors and operations managers raise in the first ten minutes of a call about migration.

  • Nobody can say if the data is right

    Everyone assumes somebody else has checked it. Nobody has.

  • The same customer exists four times

    Different spellings, different addresses, and three of them still get invoices.

  • Data was left until testing

    The design is finished, the build is underway, and nobody has looked at the actual records yet.

  • Finance will not sign off the balances

    The numbers in the new system do not agree with the numbers in the old one, and nobody can explain the gap.

  • Stock on the screen is not stock on the shelf

    If people cannot trust quantities on day one, they go back to their own spreadsheets.

  • We are told to bring twenty years across

    Nobody has asked who will use it, or what it will cost to carry it forever.

How it works

From legacy records to numbers finance will sign.

Seven steps. The first one is the one most projects skip, and it is the one that saves the most time later.

  1. Look at the data early: During design, not near testing. The records tell you what the business really does.
  2. Decide what deserves to move: Each dataset gets a decision: migrate, archive, keep for reference, or retire.
  3. Fix what is worth fixing: Duplicates, gaps and inconsistencies, with a named person in the business deciding what is correct.
  4. Agree what each field becomes: Mapping is a business decision written in technical form, so the business reviews it.
  5. Rehearse the whole run: Timed and repeatable, so the first full migration is not on go-live weekend.
  6. Prove the numbers: Balances, stock and record counts bridge from the old system to the new one, with exceptions owned.
  7. Cut over, then stabilise: Go-live is a decision made on evidence. Data ownership carries on afterwards.

What customers say

People who have been through it.

Around 80% of our business comes from existing customers and referrals. These are their words, not ours.

80%

of our business comes from existing customers and referrals.

All customer stories

The best I’ve ever experienced in over 20 years of collaboration.

Programme Manager, Hill & Smith Plc

Four decisions

What you are being asked to approve, and when.

Migration should never be a single leap of faith at the end. These are the four points where you decide whether to carry on.

What is moving, and who owns it
Gate 1Before any cleansing or mapping starts

What is moving, and who owns it

We have looked at the real data, every dataset has a decision recorded, and each one has a named owner in your business.

Confirmed at this gate

  • What the data actually looks like, measured rather than assumed
  • Migrate, archive, reference or retire, recorded per dataset
  • Agreement on which system is the trusted source
  • A named business owner for each area of data

Decision

  • Proceed
  • Reduce what moves
  • Resolve ownership first

Swipe, drag or use the arrow keys to move between decisions.

Check the fit

Six questions before migration scope is agreed.

A routing aid rather than a quote. Your answers come with you into the conversation, so nothing has to be repeated.

Question 1 of 6

0 of 6 answered

1. Why is the data moving?

Select one option.

Common questions

Questions leaders ask about data migration.

Part of the InteliSense Delivery System

Controlled Data Migration, built from the projects behind it.

CDM is the reusable capability our teams use for the work on this page: profiling, mapping, repeatable extract and load, rehearsal, validation and reconciliation, built on Microsoft technologies including Azure Data Factory and the Power Platform. It supports the discipline. It does not replace your ownership of what the data means.

How we deliver

A conversation, not a pitch

What does your data actually look like today?

Planning an implementation, heading towards cutover, or already live with data nobody quite believes. Start there and we will tell you what we would look at first.

Extensions (ISVs)The migration tooling we use on every programme.See what we have built

Related