Inventory all systems
Every production system catalogued with criticality (1 to 5), age, team owner, technology stack, business function. Spreadsheets fail above 100 systems. Catalogue tooling becomes essential.
Independent educational resource. Not affiliated with CAST, SonarSource, or any code-analysis vendor. Data sourced from CISQ, Stripe, McKinsey, and DORA reports.
You do not have one codebase. You have 50, or 200. This page covers how to measure, compare, and prioritise debt across an enterprise portfolio, with the governance models and KPIs that make portfolio strategy work in practice.
From scattered systems to a coherent portfolio view in four steps.
Every production system catalogued with criticality (1 to 5), age, team owner, technology stack, business function. Spreadsheets fail above 100 systems. Catalogue tooling becomes essential.
One primary framework (TDR is typical) plus the eight dimension assessment for breadth. Run on a fixed schedule. Make scoring uniform across teams or aggregation breaks.
A 50% TDR on the revenue path matters more than 50% TDR on a marketing site. Weight by criticality times revenue impact. Unweighted aggregation is misleading.
Roll up by business unit, technology stack, team, or cost centre. Different aggregations for different audiences. Executives see business unit. Architects see technology stack.
Two dimensions: business criticality and debt level. Four quadrants, four actions. Add your systems to see the prioritised list.
Allocate dedicated programme, 12 to 18 month horizon
Continue minimal allocation, monitor for drift
Sunset, migrate users, retire before remediating
Routine hygiene only, do not over invest
| System | Criticality | Debt | Recommended action | |
|---|---|---|---|---|
| Order processing | 5 | 2 | Protect and maintain Critical and healthy. The asset working as designed. Continue minimal allocation, monitor for drift. | |
| Customer portal | 4 | 4 | Urgent investment High criticality plus high debt is the most expensive quadrant. Allocate dedicated resources, target 12 to 18 month remediation horizon. | |
| Auth service | 5 | 1 | Protect and maintain Critical and healthy. The asset working as designed. Continue minimal allocation, monitor for drift. | |
| Legacy reporting | 2 | 5 | Consider retirement Low business criticality with high debt. Retire, sunset, or migrate users before investing in remediation. | |
| Internal admin tool | 2 | 4 | Consider retirement Low business criticality with high debt. Retire, sunset, or migrate users before investing in remediation. | |
| Marketing site | 1 | 3 | Maintain minimum Low criticality, low debt. Routine hygiene only. Do not over invest. |
Intel published their portfolio modernisation framework in 2023. The case study is widely cited because the outcomes are publicly documented.
Retired, consolidated, or migrated to standard platforms
Portfolio simplification across the modernisation programme
The framework structure Intel published
The retirement count matters as much as the modernisation count. Many enterprise debt programmes assume every system must be modernised. Intel demonstrated that retiring or consolidating one in three systems was a faster path to portfolio health than refactoring all of them.
Six metrics that tell you whether portfolio level debt strategy is working.
| KPI | Cadence | Purpose |
|---|---|---|
| Portfolio TDR trend | Quarterly | Headline metric for executive review |
| Systems above debt ceiling | Monthly | Identifies systems needing intervention |
| Budget allocated to debt reduction | Quarterly | Tracks investment as % of engineering budget |
| Mean dependency age, critical systems | Monthly | Leading indicator for security debt |
| DORA metrics, aggregated | Monthly | Cross team velocity and reliability tracking |
| Systems retired or migrated | Quarterly | Portfolio simplification progress |
Portfolio strategy without governance becomes documentation theatre. Three governance habits make portfolio debt management work in practice.
Executive sponsor present, every team owner presents 5 minutes, dependencies surfaced, budget allocated for next quarter. The cadence matters more than the format.
Engineering budget has a debt line, separate from feature work, separate from infrastructure. Typically 15 to 25% of total engineering budget. Visible and protected.
Published, written down, agreed at executive level. Criteria include: revenue impact, regulatory deadline, security exposure, capacity unlock. Removes politics from prioritisation conversations.
Three steps. First, inventory: every production system catalogued with criticality, age, team owner, and tech stack. Second, consistent measurement: pick one framework (TDR plus assessment scorecard is the typical combination) and apply uniformly. Third, weighting: criticality multiplied by debt level produces a priority score. Aggregate by portfolio, business unit, or technology stack for executive reporting.
Pick one primary framework for cross-system comparison and one supplementary framework for context. TDR is the typical primary because it produces a single comparable percentage. The eight dimension assessment is the typical supplementary because it surfaces architectural and process debt that TDR misses. Allowing each team to pick their own framework defeats portfolio aggregation.
Intel published their framework in 2023. They eliminated 665 applications, reducing the enterprise landscape by 30%. The approach: standards (every system must meet a baseline), roadmaps (every system has a 3 year plan), target architecture (a published reference for new builds and modernisations), and quarterly portfolio review with executive attendance. The retirement count is as important as the modernisation count.
The weighted average TDR across all systems, weighted by business criticality or revenue impact. A simple unweighted average is misleading: a 50% TDR on the order processing system matters more than a 5% TDR on the marketing site. Weighted aggregation reflects the actual cost the enterprise carries.
Quarterly portfolio review with executive sponsors. Each team owner presents: current state, trajectory, planned interventions, blockers. Cross team dependencies surfaced and prioritised. Budget allocation discussed and committed for next quarter. The governance cadence matters more than the format. Skipping a quarter is the typical failure mode.