Construction & Project Management Document Control

Consolidating More Than 5 Million Documents from 16 Legacy Archives for a Large Construction Group

A large construction group with more than 100 active projects held more than 5 million documents across 16 separate legacy document databases. Docwize was used to inventory the entire estate, map every document to a project register, and consolidate the archives into one active project repository, one corporate repository and one long-term archive.

Illustration showing an engineering team reviewing drawing revisions and approvals across project phases

5 million+ documents inventoried

Every record from 16 legacy databases loaded into one staging set with folders, drawing registers and cross-references.

16 source databases consolidated to 3

Active projects, corporate documents and long-term archive — with regions as logical folders, not separate systems.

100+ active projects mapped to a register

Documents attributed to projects through a metadata-driven lookup, with unmapped records surfaced as a worklist.

Claims-ready retrieval

Revision, recipient, issue date and status retained per document; retrieval by project, date, type or party regardless of when the project closed.

The challenge

Over two decades, each region and business unit had accumulated its own document database with its own folder taxonomy and document types. Day-to-day collaboration happened in SharePoint and project management tools, but when a claim surfaced after handover, what mattered was proof of which revision was current, who received it and when it was issued — and that evidence was spread across systems with no single retrieval path. SharePoint licensing scaled with active users, making it expensive to keep full access open across dozens of completed projects. The group needed a purpose-built long-term repository, but first it needed to know what it actually held.

The approach

Rather than migrate database by database and hope the result made sense, Docwize built a complete inventory of the legacy estate first, exposed it through dataviews, and used the analysis to design the target structure before moving anything.

•Load every document record from all 16 source databases — together with folder paths, drawing registers and cross-references — into a single migration staging set of more than 5 million rows
•Build a metadata-driven project register (reference number, name, division, region, status, source database) and map each document to a project through its contract-code field, classifying the remainder as unmapped or unassigned
•Publish dataviews summarising each project: document count, last active year, documents loaded after the contract end date, archive presence, upload source, and how many documents carry folder, drawing or cross-reference enrichment
•Analyse folder commonality and folder depth across databases to design one standardised folder structure and one set of document types for all regions
•Consolidate into three databases — active projects with regions as logical folders, corporate documents, and an archive for closed projects and legacy sites — with ingestion via the site-capture API, email robot rules and document controller upload

What the analysis revealed

Folder taxonomies had quietly diverged

Commonality analysis showed the same document types filed under differently named folders in different databases. A single standardised set of document types was agreed before migration rather than reconciled afterwards.

Project attribution was the hard part

More than a third of documents carried a project reference that did not resolve to the project register. An unmapped-projects worklist let the group's document controllers resolve mappings systematically instead of by memory.

Closed projects were still receiving documents

Counting documents loaded after each contract's end date revealed which nominally closed projects were still active for claims or close-out — informing which stayed in the active repository and which moved to archive.

Outcome

The group now has a single register of its document estate and a target structure grounded in what the data showed rather than what any one region assumed. Every document issued through Docwize carries its revision, recipient, issue date and status, so when a claim arrives years after handover, documents can be retrieved by project, date range, document type or issuing party in minutes. Collaboration tools stay where they work; the governed record lives where it is purpose-built and retained at a cost that does not scale with headcount on completed projects.

Sitting on decades of project documents across systems you no longer trust?

Docwize inventories what you have, maps it to your projects, and consolidates it into a governed long-term repository. Get in touch to discuss your archive.

Get in touch View all case studies