Skip to content
Implementation and teamwork

From Excel to Purchasing Software: Moving Data and Workflows

The short answer

Start with one source of truth, clean master data, and a pilot of open orders. Define the fields, their owners, and a cutover date so the new system becomes part of the team's actual workflow.

When a spreadsheet works

A spreadsheet can work well when one person manages a small number of straightforward orders and everyone knows where the files are. It is flexible, familiar, and quick to extend with a new field. Introducing software is not a goal in itself.

Problems appear when several file versions circulate, approvals live in chat, and receipts sit in another spreadsheet. Map the time spent finding the latest version and coordinating shared work before choosing a system.

Clean the master data

Start with the suppliers, projects, and items the team actually uses. Resolve duplicate supplier names, assign stable identifiers, and verify how each existing record maps to its new equivalent.

Not every historical column needs to migrate. Decide which fields support action or control and which belong in the archive. Flag missing information and assign an owner; never invent plausible-looking values to make an import pass.

  • Project and project identifier
  • Supplier and supplier identifier
  • Purchase order number and date
  • Item, unit of measure, and quantity
  • Price, status, and document links

Decide what moves and what stays in the archive

Separate open orders from completed ones. For open orders, preserve the context of what has been received, invoiced, and left outstanding. Importing only the original total can lead to double counting.

Keep the historical archive accessible and searchable, with a way to connect original files to the new records. Confirm the actual import format and how errors will be reported before preparing the final file.

Run a pilot and reconcile the figures

Use one project and a small order sample. Compare record counts, amounts, and outstanding quantities, including orders with multiple lines, partial deliveries, and cancellations.

Ten source orders should not become eleven through a duplicate import. Matching counts alone can also hide a missing order and a duplicated one. Compare identifiers as well as totals.

Set a cutover date and support the team

Choose the date from which new orders will be created in the system and identify who may still edit the old file. If parallel operation is necessary, set clear rules so it does not create competing versions.

Train people for their roles, record edge cases, and review any workarounds after the first week. Novapro implementation is shaped around the company's structure, data, and working process.

Frequently asked questions

Do we have to import the entire history?

No. You can begin with master data and open orders while retaining an accessible historical archive. Base the decision on your control requirements and the information needed to follow up on active orders.

How do we avoid duplicate imports?

Use stable identifiers, clean the source data, and define how existing records will be handled. Compare identifiers, quantities, and balances after import, rather than relying on the overall total alone.

Want to see this workflow in the application? Explore the relevant Novapro capability →

Keep reading

Construction Budget vs. Actual: What to Review Throughout a Project

How to Choose Construction Procurement Software

Construction Purchase Requisition Template: Which Fields Do You Need?