AI Execution Strategy
Incidents - Status
Jul 9, 2026, 01:17 PM
Execution Strategy
Execute a 'Fix the Flow First' Strategy to Eliminate Rework and Cut Resolution Time
The incident process is critically unstable, with 40% rework and an average resolution time of over 10 days, inflated by a 7-day administrative delay. The immediate priority is to eliminate process loops, starting with the damaging 'Work in Progress' to 'Open' regression. We will then enforce strict criteria for the 'Pending' state and correct the 7-day closure delay. This foundational stability is required before any advanced automation or AI can be effective.
Critical Confidence: High
Current process instability makes performance unpredictable, erodes user trust, and wastes significant resolver effort. Delaying these foundational fixes will only amplify waste and make future automation initiatives fail.
1Priority Actions
Priority
1
Eliminate Administrative Closure Delay
Automation Impact: Very High Effort: Low 0-30 days
Why now
The 168-hour (7-day) delay from 'Completed' to 'Closed' affects 93% of incidents, artificially inflating resolution metrics and masking true performance. This is the fastest, highest-impact change to make.
Business outcome
Achieve accurate end-to-end resolution time reporting, leading to a 30-40% reduction in the reported average ticket duration and focusing efforts on true bottlenecks.
Scope
Modify the system business rule that governs the transition from 'Completed' to 'Closed' status. Reduce the delay from 7 days to 4 hours.
Owner
Platform Engineering / ITSM Admin
Dependencies
Approval from Change Advisory Board
Risks
Minor risk of users disagreeing with a resolution after the shorter window closes, which can be handled by a new incident process.
Success measures
Average duration of 'Completed' to 'Closed' transition reduced to < 4 hours.
Overall Average Duration metric reflects true resolution time.
Evidence
Transition: 'Completed - Closed', Avg duration (hrs): '168.87', % total Work Items: '93.49'
Priority
2
Eradicate 'Work in Progress' to 'Open' Rework
Process Impact: High Effort: Medium 0-30 days
Why now
This rework loop affects 17% of incidents and represents a total failure in initial triage and assignment. It causes significant delays and user frustration. Fixing this is the most critical step to stabilize the workflow.
Business outcome
Reduced resolution time for complex tickets, improved first-touch resolution rate, and a more stable, predictable process.
Scope
Conduct a root cause analysis on the 182 incidents that moved from 'Work in Progress' back to 'Open'. Focus on initial data quality from self-service forms and triage guidelines for service desk agents.
Owner
Incident Process Owner
Dependencies
Service Desk Lead participation
Access to incident data for analysis
Risks
Analysis may reveal deeper issues with service catalog design or team capabilities, requiring a larger scope.
Success measures
Reduce the 'Work in Progress -> Open' transition count by 80%.
Decrease in average duration for variants containing this rework loop (e.g., Variant 10).
Evidence
Transition: 'Work in Progress - Open', % total Work Items: '17.43'
Variant 10 ('Work in Progress -> Open -> Work in Progress...') has a 350.56 hr average duration, nearly double the primary happy path.
Priority
3
Govern the 'Pending' State
Governance Impact: High Effort: Medium 30-60 days
Why now
The 'Work in Progress' <-> 'Pending' loop affects over 38% of incidents, creating significant delays (47 hours average wait). Lack of clear rules for when to use 'Pending' leads to process ambiguity and extended resolution times.
Business outcome
Reduced ticket 'ping-pong', shorter active resolution times, and clearer accountability for ticket progression.
Scope
Define and publish strict, mandatory entry/exit criteria for the 'Pending' status. Implement automated reminders and escalations for tickets in 'Pending' for more than 48 hours.
Owner
Incident Process Owner
Dependencies
ITSM tool configuration for automated reminders
Risks
Agent resistance to more rigid process controls.
Success measures
Reduce the average duration of the 'Pending -> Work in Progress' transition by 50%.
Decrease the volume of incidents that cycle between 'Work in Progress' and 'Pending' more than once.
Evidence
Transition: 'Pending - Work in Progress', Avg duration (hrs): '47.32', % total Work Items: '38.22'
Transition: 'Work in Progress - Pending', % total Work Items: '27.68'
Priority
4
Remediate 'Priority 5 - Planning' Misuse
Data Impact: Medium Effort: Low 30-60 days
Why now
Over 54% of Incidents are logged as the lowest priority, 'Planning'. This devalues the priority field, hinders effective work prioritization, and likely contributes to long queue times as nothing is treated as urgent.
Business outcome
Improved ability to prioritize critical work, more accurate SLA measurement, and a reduction in initial response times for genuinely high-priority incidents.
Scope
Review the definition of incident priorities. Communicate and train teams on correct usage. Consider making the field mandatory with clear guidance or removing the 'Planning' option for the Incident type.
Owner
Head of Service Management
Dependencies
Internal Communications team support
Risks
Users may start defaulting to a higher priority if 'Planning' is removed, requiring monitoring.
Success measures
Reduce the percentage of incidents logged as '5 - Planning' to below 10%.
Observe a corresponding decrease in the 'Open -> Work in Progress' queue time (currently 63 hours).
Evidence
Field Usage for Priority: '5 - Planning' accounts for 54.7% of all incidents.
Transition: 'Open - Work in Progress', Avg duration (hrs): '62.99'
2Phased Plan
1
Stabilize the Core Flow
Eliminate the most damaging rework loops and administrative delays to create a predictable process baseline.
Why this phase
This phase attacks the largest sources of waste and unpredictability. A stable process is the mandatory foundation for any further optimization or automation.
Included priorities
Priority 1Priority 2
Entry criteria
Leadership approval of the execution strategy.
Exit criteria
Rework from 'WIP' to 'Open' is reduced by 80%.
'Completed' to 'Closed' transition time is under 4 hours.
Overall process rework is below 20%.
Expected outcomes
A significant reduction in process variation.More accurate reporting of Mean Time To Resolution (MTTR).Increased confidence in the process from stakeholders and resolvers.
2
Standardize and Govern
Enforce process discipline and improve data quality to enable consistent execution and prioritization.
Why this phase
With a stable flow, we can now focus on making it consistent and ensuring work is prioritized correctly based on business impact.
Included priorities
Priority 3Priority 4
Entry criteria
Phase 1 exit criteria have been met and sustained for 2 weeks.
Exit criteria
Average time in 'Pending' is reduced by 50%.
Usage of 'Priority 5' for incidents is below 10%.
The top 3 process variants account for >70% of total incident volume.
Expected outcomes
Reduced process 'ping-pong'.Faster response to high-priority incidents.A clear 'golden path' for the majority of incidents.
3
Accelerate with Automation
Leverage the stable, standardized process to introduce intelligent automation for common, repetitive incidents.
Why this phase
The process is now reliable enough to be a candidate for advanced automation without the risk of automating chaos. This is where we unlock significant efficiency gains.
Included priorities
Entry criteria
Phase 2 exit criteria have been met and sustained for 30 days.
Exit criteria
A pilot automation for a high-volume subcategory (e.g., 'Password Reset') is live and handling >50% of its target incidents without human touch.
A business case for further automation opportunities is approved.
Expected outcomes
Reduced cost per incident.Faster resolution for end-users on common issues.Increased capacity for resolvers to focus on complex, high-value work.
3Sequencing Principles
Stability Before Speed
We must first eliminate rework and create a predictable process. Chasing speed on an unstable foundation will fail.
Address Governance Before Automation
Fixing data quality (e.g., Priority field) and process rules (e.g., 'Pending' state) is a prerequisite for successful automation.
Targeted Impact Over Broad Changes
Focus on the specific transitions and variants that cause the most delay and rework, rather than attempting a full process re-engineering.
4Do Not Do Yet
Implement AI-powered automated resolution
The current process has 164 variants and 40% rework. Applying AI now would automate a chaotic process, leading to poor outcomes and wasted investment. Stabilize and standardize first.
Redefine all Incident SLAs
The process duration is too unpredictable (Std Dev > Avg) to set meaningful SLAs. Doing so now would create noise and alert fatigue. Establish a stable baseline in Phase 1 and 2, then define realistic SLAs.
Launch a full-scale agent retraining program
The primary issues are systemic process flaws (rework loops, admin delays), not individual performance. Fix the process, then provide targeted training on the new, simplified workflow.