Quick answer.
IT operational excellence is the disciplined removal of work that doesn't produce value — duplicate processes, oversized application portfolios, low-impact automation, manual reconciliation between systems that should be integrated. Across the engagements we've run, the typical recoverable inefficiency in a mid-cap IT operation sits between 15% and 25% of total run cost.
Why most CIOs misdiagnose this.
The default approach to IT efficiency is one of two postures: cost-out (a top-down percentage cut applied uniformly) or transformation (a multi-year program led by a brand-name systems integrator). Both approaches share a flaw: they begin with the answer and work backward to the problem. Cost-out picks numbers before the analysis. Transformation picks technology before the diagnosis.
Operational excellence runs the other direction. The diagnosis comes first; the action falls out of it. The diagnosis is unglamorous — process mining, ticket-data analysis, application-portfolio inventory — but it is where the recoverable money is actually visible.
"The documented process and the actual process are reliably different. The work isn't in writing better documents — it's in fixing the actual."
— Margit Györfi, CPO, AventarioThe IT landscape assessment: a step-by-step.
- Application inventory. Pull every active application, system, and platform. Owner, business purpose, run cost, user count, integration footprint. Expect to find 15–30% more applications than the CIO expected.
- Process mapping. The top 20 IT-touching business processes (incident, change, fulfilment, onboarding, offboarding, etc.). End-to-end, not within-team.
- Cost allocation. Run cost by application, by process, by team. Most organizations cannot answer this within 20% accuracy at first attempt.
- Pain-point inventory. Where does work get stuck, redone, escalated. Talk to the people doing the work; the executive view is consistently wrong about this.
- Synthesis. A heat map: high-cost / high-pain / low-value combinations are the action queue.
Process mining: what it is, what it finds.
Process mining is the practice of reconstructing actual process behaviour from system event logs, rather than from how someone says the process works. The two are reliably different. A process documented as a five-step incident-management workflow turns out, in the event log, to involve 23 distinct paths, of which the documented one accounts for 38% of cases. The other 62% is the work to investigate.
The typical finding: 10–20% of process volume runs through paths that, on inspection, add no value — duplicate approvals, unnecessary handovers, manual rekeying between systems, exception handling for cases the process should have handled automatically.
Shadow IT: the hidden iceberg.
Every CIO underestimates shadow IT. The official application inventory shows 180 applications; the SaaS-management tool, when deployed, finds another 60–90. Each of those is paid by someone, used by someone, integrated with something — and none of it is governed. The first job is visibility. The second is triage: which shadow systems are filling a real need (and should be brought into the official portfolio) and which are duplication (and should be sunset).
Application portfolio rationalization: the four-quadrant decision.
- Keep. The application is fit-for-purpose, well-used, well-integrated, low-risk. Most of these are the obvious cases.
- Kill. No longer used at the volume that justifies the run cost. Migrate users to existing alternatives; decommission.
- Migrate. Used, but on an end-of-life platform or with overlapping functionality elsewhere in the portfolio.
- Modernize. Strategic, used, but the implementation is dated and the run cost has crept up. Replatform.
The decision rule we use: any application that scores low on usage and high on run cost is in the kill quadrant by default. The burden of proof is on keeping it, not on removing it.
Automation readiness: what to automate, what to leave.
The default automation backlog in most IT organizations is too long and badly prioritized. The screening question is not can this be automated; nearly everything can. The question is does the business case survive contact with the implementation cost. The criteria that matter:
- Volume. The process runs often enough that automation pays back inside 12–18 months.
- Stability. The process is not going to change shape inside the next 24 months.
- Standardization. The process variants are bounded; the automation doesn't need to handle 47 edge cases.
- Data integrity. The systems involved expose stable APIs, not screen-scraping.
An automation candidate that fails any one of these is usually not worth doing — even though it could be done.
The digitization business case the board approves.
Boards have learned to discount IT business cases. The cases that survive scrutiny share three properties: a baseline measured rigorously rather than estimated, a recovery path with proof points at 90, 180, and 365 days, and a governance commitment that the savings will be tracked into the operating budget rather than reinvested invisibly. The third is the one most often missing.
How Aventario approaches this.
Our Operational Excellence engagement runs the diagnostic — landscape, process mining, ticket analysis, portfolio rationalization — and stays through the elimination. The benchmark target is ≥20% efficiency gain, measured against a baseline we establish before any change is made. Outcome-based engagement; the fee is tied to realized, signed-off improvement.
FAQ.
What is IT operational excellence?
The disciplined removal of work that doesn't produce value across the IT operation — duplicate processes, oversized application portfolios, manual reconciliation, low-impact automation candidates.
How much is realistically recoverable?
15–25% of total run cost across the engagements we've measured, with a 6–18 month realization window depending on portfolio complexity.
What is process mining?
The reconstruction of actual process behaviour from system event logs, rather than from documented process maps. It typically reveals significant divergence between the two.
Julian Robida is Research Lead at Aventario. Margit Györfi (CPO) contributed expert input drawn from 25+ years of running IT engagements across pharma, automotive, financial services, and the public sector. Aventario is a boutique consultancy in Vienna; we have negotiated over €3B in IT contract volume and delivered more than 500 engagements across DACH and beyond.