The earlier archive Written for a different audience — operations teams, business stakeholders, new joiners. It is kept unchanged, with its original date, because a dated body of work says something a rewritten one cannot.

Post-Deployment Support: The Critical First 90 Days

It was October 2022 at Exclusive IT Solutions. We had just deployed a new inventory management system for a retail client.


The Launch Day We Thought Was Done

It was October 2022 at Exclusive IT Solutions. We had just deployed a new inventory management system for a retail client.

The deployment went smoothly. No errors. Perfect data migration. All systems came online.

We celebrated. We wrote "deployment complete" in our project tracker. We moved on to other projects.

A month later, the client called. Furious.

"The system doesn't work," their operations director said.

"What's wrong?" I asked. "Everything tested perfectly."

"It works technically," he said. "But it doesn't work for us."

He explained:

  • Store managers say it's too slow
  • Warehouse staff can't find the right screens
  • Sales associates don't understand the terminology
  • The reporting format doesn't match what they need for their weekly meetings

We had deployed the system. We hadn't deployed it for actual use.

The Metaphor: The Housewarming vs. Moving In

Let me explain with a metaphor that transformed my approach to deployments.

Deployment is like delivering a furnished house. You've built it, you've decorated it, you've handed over the keys. The house is complete.

But the real work starts when people move in. They discover:

  • The furniture arrangement doesn't work for how they live
  • The kitchen lacks a utensil drawer that would be perfect
  • The lighting in the bedroom is wrong for how they read
  • The closet organization doesn't fit their clothes

The house wasn't "wrong." It just wasn't adapted to how they live.

Deployment is the housewarming. Post-deployment support is helping them move in.

The Struggle: The "We're Done" Mindset

I'll be honest: we loved checking "deployment complete."

It felt like achievement. We celebrated. We moved to the next exciting thing.

But here's the truth: deployment is not the finish line. It's the starting line for real-world testing.

Systems work in controlled environments. They behave predictably. They're tested thoroughly.

Real-world use is completely different:

  • Edge cases you never tested
  • User behaviors you didn't anticipate
  • Performance characteristics you didn't see in testing
  • Integration issues that only appear under real load

The Transformation: The First 90 Days Framework

I implemented a new approach. Every deployment had a mandatory 90-day post-deployment support phase.

The "First 90 Days" Framework

Week 1-2: Intensive Support

What:

  • On-site presence for critical deployments
  • Rapid response to any issues (within 30 minutes)
  • Daily check-ins with key stakeholders
  • Real-time monitoring for problems

Why: Users need help immediately when they first start using a system. Small issues escalate quickly when users are frustrated.

Week 3-4: Stabilization

What:

  • 8-hour response time (not immediate, but fast)
  • Bi-weekly check-ins with stakeholders
  • Focus on identifying patterns in issues
  • Documenting edge cases and unexpected usage

Why: Users are getting comfortable. Patterns emerge. You need to capture them.

Week 5-8: Optimization

What:

  • 24-hour response time (standard SLA)
  • Monthly check-ins with stakeholders
  • Performance optimization based on real usage patterns
  • UI adjustments based on user feedback

Why: Real-world usage reveals what actually needs optimizing—not what you thought would.

Week 9-12: Documentation and Handover

What:

  • Update all documentation based on real-world use
  • Create runbooks for common issues and solutions
  • Train internal support teams to handle issues independently
  • Formally close deployment

Why: By month 3, the system should be stable. Internal teams should own support.

The Real-World Example: The Inventory Management System

Let me share how this framework transformed our approach to the inventory management deployment.

What Went Wrong (Without Post-Deployment Support)

Week 1-2 (We Thought We Were Done)

  • Store managers complained it was slow
  • Warehouse staff couldn't find right screens
  • Sales associates didn't understand terminology
  • Reports didn't match their needs

Our response: We were busy on other projects. We responded when we could. Issues accumulated. Users got frustrated.

Week 3-4

  • Users started finding workarounds (paper, spreadsheets, manual processes)
  • Usage was inconsistent across locations
  • Performance issues got worse as usage increased
  • Stakeholder trust eroded

Our response: We tried to catch up, but issues had accumulated. We were reactive, not proactive.

What We Did (With Post-Deployment Support)

When a similar client deployed six months later, we applied our first-90-days framework.

Week 1-2: Intensive Support

Issues identified and addressed:

  • Slowness: Database queries weren't optimized for actual usage patterns → Added indexes, restructured queries (resolved in 4 days)
  • Screen navigation: Warehouse workflow was different from what we assumed → Reorganized menu structure (resolved in 3 days)
  • Terminology: "SKU" vs "Product ID" confused sales staff → Renamed to "Product Number" (resolved in 1 day)
  • Reports: Weekly meeting format needed different data columns → Added custom report builder (resolved in 5 days)

Result: Week 2, users were using the system enthusiastically.

Week 3-4: Stabilization

Issues identified and addressed:

  • Mobile device issues: Store managers used tablets, system wasn't mobile-friendly → Added responsive design (resolved in 1 week)
  • Barcode scanning: Warehouse barcode scanners weren't compatible → Added device support (resolved in 3 days)
  • Permissions: Staff roles were too restrictive → Adjusted role permissions (resolved in 2 days)

Result: Week 4, system was stable across all use cases.

Week 5-8: Optimization

Optimizations based on real usage:

  • Search performance: Users searched by description more than by SKU → Optimized text search (improvement: 60% faster)
  • Dashboard customization: Different locations needed different metrics → Added configurable dashboards
  • Bulk operations: Warehouse managers needed bulk updates → Added bulk edit functionality

Result: Week 8, system was performing better than at launch.

Week 9-12: Documentation and Handover

Documentation updates:

  • Updated user guides based on real workflows
  • Created troubleshooting runbooks for common issues
  • Documented edge cases and solutions
  • Trained client's internal IT team on system management

Result: Week 12, client's internal team owned support. We formally closed deployment.

The Results: Measurable Impact

User Satisfaction

Without post-deployment support:

  • Week 1: 2.1/5 (frustrated)
  • Week 4: 2.8/5 (still frustrated)
  • Week 8: 3.2/5 (marginally satisfied)
  • Week 12: 3.5/5 (somewhat satisfied)

With post-deployment support:

  • Week 1: 3.8/5 (good, thanks to immediate support)
  • Week 4: 4.5/5 (very good, issues resolved)
  • Week 8: 4.7/5 (excellent, optimized for them)
  • Week 12: 4.8/5 (excellent, stable)

Improvement: 37% higher satisfaction by week 12

Adoption Rates

Without post-deployment support:

  • Week 1: 60% adoption (many users resisted)
  • Week 4: 68% adoption (slow growth)
  • Week 8: 75% adoption (plateaued)
  • Week 12: 78% adoption (never fully adopted)

With post-deployment support:

  • Week 1: 85% adoption (enthusiastic start)
  • Week 4: 92% adoption (strong adoption)
  • Week 8: 97% adoption (near-universal)
  • Week 12: 98% adoption (fully adopted)

Improvement: 26% higher adoption by week 12

Support Requests

Without post-deployment support:

  • Week 1-12: 387 support requests (many repetitive issues that should have been fixed)

With post-deployment support:

  • Week 1-12: 124 support requests (issues identified and resolved early)

Reduction: 68% fewer support requests

Time to Stability

Without post-deployment support: 6 months to reach stable usage With post-deployment support: 2.5 months to reach stable usage Reduction: 58% faster to stability

The Metaphor: The Restaurant Opening

I like to explain this with another metaphor:

Deployment is like a restaurant opening night. You've prepared the food, you've trained the staff, you've decorated the space. The restaurant is ready.

The real test is the first customers. They discover:

  • The menu layout is confusing
  • The tables are arranged awkwardly for large groups
  • The music is too loud for conversation
  • The service isn't as smooth as it was in training

A restaurant that responds immediately to these issues? They get great reviews. They build regular customers.

A restaurant that ignores these issues, blames customers for "not understanding"? They get bad reviews. Regular customers don't return.

Post-deployment support is the restaurant responding to opening night feedback.

Common Mistakes to Avoid

Mistake 1: Treating Deployment as Finish Line

The Problem: We're done! Move on to next project.

The Fix: Deployment is starting line. 90-day post-deployment support is mandatory.

Mistake 2: Ignoring Early Issues

The Problem: That's just user confusion, they'll figure it out.

The Fix: Every issue is data. Early issues reveal what needs fixing. Fix them immediately.

Mistake 3: Not Adapting to Real Usage

The Problem: Users are using it wrong, they should use it as designed.

The Fix: Adapt to how users actually work, not how you think they should work.

Mistake 4: Reactive Instead of Proactive

The Problem: Wait for problems, then fix them.

The Fix: Proactive check-ins, monitoring, and optimization. Anticipate issues.

Mistake 5: Never Documenting Lessons

The Problem: We fixed issues, but didn't document what we learned.

The Fix: Document edge cases, solutions, and lessons learned. Apply to future deployments.

The Hard Truth

The deployment date isn't the finish line. It's the starting line for real-world testing.

I've seen deployments that were technically perfect but failed because users couldn't use them. I've seen deployments with initial issues that succeeded because post-deployment support made the system work for users.

A deployed system isn't successful. A used system is successful.

A Framework for Your Deployments

If you're deploying systems:

Week 1-2: Be There

On-site or virtually always available. Immediate support.

Week 3-4: Watch Patterns

Identify recurring issues. Fix root causes.

Week 5-8: Optimize

Based on real usage, not assumptions.

Week 9-12: Document and Handover

Update documentation. Train internal teams. Close deployment formally.


What's your approach to post-deployment? Do you support systems after launch?

#Deployment #PostDeploymentSupport #ChangeManagement #ITOperations #ProjectManagement

The archive

Every article, 2013–2026

Thirty dated articles on application support, delivery, infrastructure and data — plus the recent writing on CI/CD reproducibility and offline-first design.

The recent work is written for a specialist audience, and it is where the measurement discipline is set out in full.

Recent writingOpen Source Impact