How to Build a Defensible Capital Improvement Plan

What a capital improvement plan is — and what "defensible" adds

A capital improvement plan (CIP) is a multi-year schedule of an agency's planned capital projects and the funding sources to pay for them. The Government Finance Officers Association (GFOA) recommends that a CIP cover a period of at least three years, and preferably five or more, and that it be tied to the agency's strategic plan and reviewed annually. A CIP is distinct from a single-year capital budget, which appropriates money for the current fiscal year; the CIP is the plan, the budget is the year-one commitment that flows from it.

A defensible CIP is one whose every project and rank can be traced back to evidence — measured condition, a forecast of how that condition will change, and a risk-based reason this project beats the next one. Defensibility is what survives a city council's questions, an auditor's review, and a grant reviewer's scoring. With the U.S. facing an estimated $3.7 trillion investment gap and an overall grade of C (ASCE 2025), and federal surface-transportation funding under the IIJA expiring September 30, 2026, the gap between "a list of projects" and "a defensible plan" is increasingly the gap between funded and unfunded.

How to build a capital improvement plan in 8 steps

Building a capital improvement plan is the process of turning what an agency owns and how it is aging into a prioritized, funded, multi-year program. The eight steps below move from raw asset data to an adopted plan. The early steps establish the evidence; the later steps turn it into decisions a council can defend.

1. Build a complete asset inventory

Start with a single, complete inventory of the assets the plan will cover — roads and pavement, bridges, water and sewer pipes, facilities, runways, and the rest. The inventory is the denominator for everything that follows: you cannot prioritize, forecast, or fund what you have not counted. Most agencies already hold much of this in an EAM or CMMS and a GIS; the task is to consolidate it into one cross-asset view rather than re-key it.

2. Assess current condition

Attach a current condition rating to each asset — for example a Pavement Condition Index (PCI) for roads, a sufficiency or NBI-style rating for bridges, or a likelihood-of-failure score for buried pipe. Condition is the evidence that justifies a project. A plan built on condition data is far harder to challenge than one built on the squeakiest wheel.

3. Forecast deterioration

Project how each asset's condition will decline over time using deterioration modeling. This is the step that separates a static snapshot from a plan: it tells you not just what is bad today, but what will become expensive tomorrow if deferred. AI deterioration models that fuse condition history, environment, and consequence forecast more sharply than static straight-line curves — see how they compare.

4. Quantify the need and the cost of waiting

Roll the forecast up into a quantified deferred maintenance backlog measured against a state-of-good-repair target. Because deterioration accelerates as condition drops, a dollar of timely preservation routinely prevents several dollars of reconstruction — and that cost-of-waiting figure is often the single most persuasive number in the whole plan.

5. Prioritize by risk

Rank candidate projects by risk — the probability of failure times the consequence of failure — rather than by date requested or department clout. Risk-based prioritization makes the order defensible: every rank has a reason. For the scoring mechanics, read how to prioritize infrastructure projects when the budget won't cover everything.

6. Model funding scenarios

Use scenario budgeting to show how different funding levels change network condition and backlog over the plan horizon — a constrained budget against the unconstrained need, and several points in between. Scenarios turn the plan into a decision tool: leadership can see the consequence of each funding choice before committing. Learn the technique.

7. Align funding sources and grant strategy

Map each year of the plan to its funding sources — operating revenue, bonds, and grants. A plan that already identifies a condition-backed, risk-ranked project pipeline is also a grant-ready pipeline, which matters as the IIJA window closes and discretionary competition tightens.

8. Adopt, publish, and revisit annually

Bring the plan to the council or board for formal adoption, publish it, and revisit it every year as condition is reassessed and budgets change. The GFOA frames the CIP as a living, annually reviewed document, not a one-time exercise. Annual refresh is also what keeps the same data serving the audit (GASB 34) and, for DOTs, the TAMP under MAP-21.

Where software replaces the spreadsheet

For decades the CIP lived in a spreadsheet, which works until the inventory grows, the condition data changes, and someone on the council asks why project A outranks project B. Capital planning software is the owner-side layer that automates the evidence: it pulls inventory and condition from the systems you already run, forecasts deterioration, ranks by risk, runs funding scenarios, and produces an audit-ready record of how every number was reached.

InfraMind is AI capital planning software that does exactly this on top of the EAM and GIS you already run — it does not replace systems like Esri; it is the condition-to-capital planning layer above them. The result is a plan that is faster to build, easier to update annually, and defensible by construction. Explore the InfraMind platform or the capital planning software overview, and when you are ready to evaluate, the RFP requirements guide and How to Buy page show the procurement path.

Need to convince a council or board? See Building the Business Case for Capital Planning Software.