Microsoft Project Online End-of-Life: What You Need to Know in 2026

If you opened your Microsoft 365 Admin Center any time in the last twelve months, you've seen the banner: Microsoft Project Online will be retired on September 30, 2026. As of today, that leaves roughly five and a half months before the service stops working — and most PMO directors we talk to are still underestimating how much work the migration actually is.

This isn't another retirement where Microsoft quietly moves you to a successor SKU and everything keeps running. Project Online's shutdown is a hard cutover. The PWA (Project Web App) interface, the enterprise resource pool, the OData reporting endpoint, and the Project Online Desktop Client's online features all stop on the same day. Whatever lives inside Project Online needs to live somewhere else by then — or it's gone.

This post walks through exactly what's happening, what Microsoft is (and isn't) offering as a replacement, why the obvious successor products may not fit your organization, and a 30-60-90 day action plan you can start running today.

The Official Timeline

Microsoft published the retirement announcement through the Microsoft 365 message center (MC812729) on July 10, 2024, and has updated the Project Online retirement notice several times since. The key dates:

July 2024 — Initial retirement notice published
October 2025 — New Project Online Professional and Premium subscriptions no longer available for purchase
April 2026 (now) — Existing tenants can still use the service but cannot add new PWA instances
September 30, 2026 — Hard shutdown. PWA interfaces disabled. OData endpoint returns 410 Gone. Desktop Client online sync stops working.
Post-retirement — SharePoint sites associated with projects remain (they're SharePoint Online content, not Project Online content), but the scheduling engine, resource pool, and custom enterprise fields are unrecoverable.

There is no extension, no deprecation-mode read-only period, and no migration tool from Microsoft. The service simply stops.

What Actually Happens to Your Data

This is the question that should keep every PMO director awake, because the answer is more nuanced than Microsoft's FAQ suggests.

Project Schedules

Your project plans live in Project Online's SQL database, which Microsoft operates. You do not have direct access. Your only extraction path is:

Export each project as a .mpp file via the Project Online Desktop Client, or
Pull XML via the OData feed before September 30, 2026.

Both paths work today. Neither works after the shutdown. If you have 500 active projects and you've never exported any of them, that's 500 manual exports to schedule.

Enterprise Resource Pool

This is the one that's hardest to recover. The ERP — your enterprise resource names, calendars, cost rates, skill codes, custom resource fields — exists only inside Project Online. There is no officially supported export format. You'll need to pull it via OData and reconstruct it in whatever successor system you choose.

Timesheet History

Gone, unless you export it. Timesheet data drives most PMOs' billable-hours reconciliation, audit evidence, and utilization reporting. If your SOX or ISO 9001 scope includes timesheet retention, you need to export at least three years of history to an archival format (CSV or Parquet) and document the chain of custody before the cutover.

Custom Fields, Views, and BI

Enterprise custom fields, lookup tables, and any Power BI dashboards that hit the OData feed will all need to be rebuilt. The fields themselves are definition-only data (easy to export as metadata); the problem is re-mapping them into whatever platform you move to.

SharePoint Project Sites

These are safe. They live in SharePoint Online and aren't affected by the Project Online retirement. Your documents, issues lists, risk logs, and team sites will continue to work. What breaks is the link between those sites and the Project Online schedule — the "sync issues to SharePoint" feature simply stops.

Microsoft's Recommended Path (and Why It May Not Work)

Microsoft is pointing customers toward two products: Microsoft Planner (the new unified Planner, which absorbed Project for the Web in 2024) and Microsoft Project Plan 3 / Plan 5 desktop licenses. On paper this sounds like continuity. In practice, both products have significant gaps for anyone running Project Online today.

The Gaps in Microsoft Planner

No enterprise resource pool. Planner has individual assignees, not a shared resource pool with cost rates and calendars. If you do resource leveling, capacity planning, or billable-rate reporting, this is a cliff, not a gap.
No portfolio-level dependency tracking. Planner handles within-project dependencies, but cross-project "critical chain" rollups — the reason most organizations picked Project Online in the first place — aren't there.
No PMO governance layer. Project Online's stage-gate workflow, portfolio approvals, and demand-management forms don't exist in Planner. You'd rebuild all of it in Power Apps and SharePoint lists.
Limited reporting. Planner's reporting is scoped to individual plans. Enterprise Project Type (EPT) rollups, portfolio RAG dashboards, and capacity heat maps don't have an equivalent.

The Gaps in Project Plan 3/5 Desktop

Project Plan 3/5 is essentially Project Professional with a cloud-connected scheduling engine (Project for the Web's backend). It solves the "I need a Gantt chart and critical path" problem but it's a single-project tool, not a portfolio system. There is no PWA equivalent, no timesheet approval workflow, and no enterprise-wide reporting. It's a step backward for any org that adopted Project Online specifically for the PWA side of the product.

Microsoft has signaled that portfolio features will "come to Planner over time," but as of April 2026 they aren't there. Betting your 2026 Q3/Q4 operations on a product roadmap that's five months out from your hard deadline is a risk most PMO directors won't accept.

Your Real Options

Strip away the Microsoft-preferred narrative and you have three honest paths forward:

Option 1: Stay on Microsoft, Accept the Downgrade

Move to Planner + Project Plan 3/5 and accept that you'll lose portfolio features, enterprise resource management, and governance workflows until Microsoft adds them back. Supplement with Power BI + custom SharePoint/Power Apps for the gaps. Cost: the Microsoft licensing bill stays roughly flat, but you pay in custom-development time and reduced PMO capability.

Realistic total cost of ownership: high, because every gap becomes a Power Platform project with its own maintenance burden.

Option 2: Migrate to a Modern PPM Platform

Move to a purpose-built project and portfolio management platform that replicates everything Project Online did — Gantt, resource pool, timesheets, portfolios, governance — and adds modern capabilities (AI risk detection, real-time collaboration, mobile-first access) that Project Online never had.

This is where Onplana fits. Onplana was built specifically with this retirement in mind: it imports .mpp files, it imports the Project Online OData feed directly (no intermediate CSV gymnastics), and it preserves the PMO constructs — portfolios, gate reviews, resource pools, timesheet approvals — that Microsoft's own alternatives drop. We wrote a full comparison in MS Project vs Onplana: Complete Feature Comparison for 2026.

Pricing starts at $0 for teams up to 5 and scales linearly from there. Full details on the pricing page.

Realistic total cost of ownership: low to medium, depending on team size. The migration guide is at /ms-project-alternative.

Option 3: Build Your Own on SharePoint / Power Platform

Some large enterprises with strong internal Power Platform teams will attempt to rebuild PPM capabilities on top of SharePoint Lists, Dataverse, Power Apps, and Power Automate. It's technically feasible. It's also slow, fragile, and expensive — you're reinventing a PPM product with a team of two developers and a two-quarter deadline.

Realistic total cost of ownership: very high, because you're taking on product-engineering work that vendors have spent a decade building.

The Migration Risk Most PMOs Underestimate

The single biggest risk we see is not data loss. Microsoft gives you enough export tools to get the data out. The real risk is process continuity — the stage-gate reviews, the timesheet approval workflow, the portfolio steering-committee cadence, the way your organization has built ten years of habit around Project Online's specific UX.

A migration that moves the data but doesn't preserve the process is a migration that quietly fails six months later, when the CFO asks why utilization reporting is broken or the sponsor asks why the Q4 gate review didn't happen.

Any vendor evaluation needs to ask, specifically:

Can we import the Project Online OData feed directly?
Can we preserve our EPTs, custom fields, and lookup tables?
Can we recreate our stage-gate workflow (DRAFT → SUBMITTED → REVIEW → APPROVED → IMPLEMENTATION → CLOSED, or whatever your specific flavor is)?
Can MEMBER-role users approve gate reviews when designated, without giving them full admin access?
Can timesheets flow into billable-rate reporting with the same resource-rate logic we use today?

If a vendor can't answer all five affirmatively with a demo, they're not a real Project Online replacement — they're a task-management tool dressed up as one.

A Practical 30-60-90 Day Action Plan

If you haven't started yet, here's the plan we've seen work for teams of 50–500 users.

Days 0–30: Inventory and Decision

Week 1: Export a full OData snapshot of every active project. This is your backup. If everything else fails, you still have the data.
Week 2: Inventory your custom fields, EPTs, lookup tables, and resource pool. Document what's actually in use (most orgs have 40%+ legacy fields nobody queries anymore — prune before you migrate).
Week 3: Run a vendor shortlist (Onplana, the Microsoft-native path, Smartsheet, one other). Demo each against three of your real project plans, not generic demo content.
Week 4: Make the platform decision. Get sponsor and procurement sign-off.

Days 31–60: Pilot

Pick two or three real projects representing different EPTs. Migrate them to the chosen platform.
Recreate the stage-gate workflow, resource pool, and one portfolio rollup.
Run the pilot projects in parallel with Project Online for 30 days. Compare outputs: schedule variance, utilization, burn-down.
Identify process gaps before they become production gaps.

Days 61–90: Cutover

Migrate remaining active projects in waves (we recommend 20–40 per week to keep training manageable).
Export and archive all timesheet history and completed projects to long-term storage (S3 + Parquet is our default recommendation, or an Azure Storage equivalent).
Freeze Project Online writes. Redirect all user traffic to the new platform.
Retain read-only Project Online access for the final 30 days as a rollback path.

On September 30, 2026, decommission. Preserve your exported archives.

The Cost of Waiting

Every week you delay compounds three risks:

Vendor availability. Every mid-size PMO is moving at the same time. Implementation partners, trainers, and data-migration specialists are booking solid through Q3 2026. The teams that start in April are getting first pick. The ones starting in July are paying premiums and accepting suboptimal timelines.
Data loss probability. The longer you wait, the more you compress the export phase into a single frantic weekend. Rushed exports miss custom fields, miss timesheet history, miss the quirky workflow configurations nobody documented.
Business disruption. A 90-day migration done calmly is boring and predictable. A 30-day migration done in panic is a reputation event. PMO directors who wait get blamed for both the downtime and the data loss, even though Microsoft set the deadline.

What We'd Do in Your Position

If we were running a Project Online PMO today, our playbook would look like this: export a full OData backup this week, shortlist two successor platforms by the end of the month, pilot one of them in May, and be in active migration by the end of June. That puts cutover in August, with a full month of buffer before Microsoft pulls the plug.

If you want to see exactly how Onplana handles the Project Online import — the OData feed, the resource pool, the stage gates, the timesheets, all of it — the MS Project alternative page walks through the flow with screenshots, and you can start a free migration in the product without talking to sales.

The clock is ticking. September 30 is not negotiable. The teams that move first move calmly; the ones that wait move in crisis. You know which of those you'd rather be.

Next step: Start a free Onplana workspace and run your first test import at onplana.com/ms-project-alternative. No credit card required, and the first 5 users are free forever.