What do you do about 15 years of project data that nobody can find?
Nathan Voogt
•
June 30, 2026
It's a question I hear in almost every conversation with construction and engineering firms.
Not whether they still have their project records, they do. The real question is whether anyone can actually find and use them.
For many organizations, years of completed projects are locked away across SharePoint, file shares and cloud storage. The information exists, but it's fragmented, difficult to search and increasingly expensive to keep. Instead of being a valuable business asset, it becomes a growing operational and compliance risk.
That's the challenge I see time and time again.
Why does construction project data become stranded?
Fragmentation is rarely a deliberate choice. A firm grows, data accumulates across SharePoint, file shares, cloud storage, and whatever tools each project happened to use - and when the job closes, it just stays where it is. No trigger, no process, no one with the authority to move it.
Fifteen years later: petabytes of records that are technically yours but practically inaccessible.
How can you search across SharePoint, file shares and cloud storage at the same time?
This is one of the first things people ask me. Without a unified layer on top, you can't - not easily. Finding records for a closed project means querying each system separately, and an hour later you still can't be certain you've found everything.
For firms working on long-lifecycle infrastructure, where a project from 2011 might be directly relevant to a claim or a bid today, that's a real operational problem.
How long do construction companies need to keep project records?
Truth is, it's longer than most systems are designed to handle. Retention obligations can run 10, 20, even 50 years after project completion - and neither SharePoint nor cloud storage is built to ensure a file readable today is still readable in 2045.
That gap between the obligation and the tool notionally meeting it, is where most of the risk sits.
What happens when a construction firm has 11 independent business units and no shared data governance?
We worked with a large US construction and engineering firm in exactly this position. Eleven business units, operating at around 80% decentralization. Each with its own data practices, its own systems, and its own informal answer to "what do we do when a project closes?", which mostly meant doing nothing.
Data had accumulated for 15 years with no consistent governance. Administrative files sat in SharePoint, while engineering data lived on file shares. Legacy project data had been migrated to cloud storage at some point, but never properly decommissioned from the original location, so duplicate copies were spread across multiple systems, driving up costs. There was no lifecycle trigger; no event that automatically initiated an archival transfer. Things moved only when a storage bill got noticed or SharePoint performance degraded enough to raise a ticket.
Three problems were compounding at once. Storage costs were climbing. Users couldn't find records without querying multiple systems individually. And there was no governance framework that could hold consistently across all eleven units.
Preservica became the single long-term intelligent archive across all of them.
"For the first time, a single search surfaced records from across the full estate - SharePoint, cloud storage, migrated legacy systems".
Policy-based triggers, configured around project closure and other lifecycle events automated data movement without central IT involvement in every transfer. Each business unit kept its operational autonomy; the guardrails applied consistently across all of them. For the first time, a single search surfaced records from across the full estate - SharePoint, cloud storage, migrated legacy systems - with security controls enforcing per-unit access. Storage costs dropped significantly. The search problem went away. And governance became something the business could actually demonstrate, not just claim.
Is historical project data ready for AI?
Well no, or for most firms, not yet. But 15 years of project records is an enormous body of institutional knowledge: specs, cost data, site conditions, engineering decisions. Once records are preserved in open formats with a consistent metadata layer, querying past projects by geography, spec type, or engineering precedent becomes achievable.
That's where this firm is headed. Getting the archive in order is the prerequisite to AI success.
The conversation I can offer you.
If this sounds all too familiar - the fragmented silos, the storage cost pressure, the lack of lifecycle governance, the SharePoint that's struggling under the weight of content it's holding indefinitely - I'd like to talk.
Not because I think every organization needs the same solution, but because the problems are usually more solvable than they look from the inside. The data is already there. The obligation to keep it isn't going away. The question is just whether it works for you or against you.
Get in touch with our team or read the full case study on this construction firm's approach.
