Who is this checklist for?

Government CIOs, IT directors, city or county managers, agency program directors, and any leader scoping or recovering a digital transformation program. Federal, state, and municipal.

Can I use this checklist as a self-assessment?

Yes. The checklist is structured so that you can score each item green / yellow / red against your current program. The aggregate gives you an honest picture of which phase your program is actually in, which is often different from which phase the slide deck says it is in.

What if my program has skipped a phase?

That is the most common pattern. Most stalled programs have skipped Discovery, gone straight to Pilot, and are now stuck because Strategy was never done. The remediation is to circle back rather than push forward. The checklist makes this visible.

How long does it take to work through the checklist?

As a self-assessment, two to four hours of leadership time. As a program plan, the underlying work spans the program lifecycle, typically 18 to 36 months for a meaningful transformation.

Where does AI fit in the checklist?

AI fits inside Pilot and Scale, gated by Governance. We deliberately did not put AI in its own phase. Treating AI as a separate program from the rest of the digital transformation tends to produce poorly-governed AI deployments and undergoverned non-AI deployments.

Insight / Government / Checklist / April 2026

Government Digital Transformation Checklist

Q: Can I get this as a downloadable PDF?

Yes. Request our capability statement and we'll include a printable PDF version of the checklist as part of the package.

This is a working checklist. Twenty-five items, organized into five phases. It is the checklist we use ourselves when we walk into a federal, state, or municipal client engagement and need a fast, honest assessment of where the program actually is, as opposed to where the program plan says it is.

The checklist is not exhaustive. It is opinionated. It reflects the failure modes we have seen most often: skipping Discovery and going straight to Pilot, running Pilots without exit criteria, scaling without governance, declaring Strategy complete on the basis of a slide deck. If your program looks healthy on every item below, congratulations. If it doesn't, the gaps tell you where to go next.

Phase 1: Discovery (5 items)

Discovery is the work that happens before strategy. It is the answer to "what is actually true right now." Discovery is not a project plan. It is not a vendor RFP. It is the foundation everything else stands on.

Application portfolio inventory. Every running workload, its owner, its operational criticality, its current cost (capital and operating), its dependencies, its change velocity. Walked, not asserted. If the existing inventory has not been verified in the last 12 months, treat it as suspect.
Stakeholder map and political alignment. Every elected official, appointed leader, agency head, union or works council, and significant external partner whose support is required for the program. Not a slide; a working document that gets updated as the program runs.
Current-state cost baseline. Three-year actual operating spend on the in-scope portfolio, broken out by hardware, hosting, license, labor, and external services. Without this baseline, no future-state cost claim is defensible.
Compliance obligation mapping. Every framework that applies to the in-scope workloads (NIST 800-53, FISMA, HIPAA, CJIS, FedRAMP, StateRAMP, state-specific privacy law, state-specific AI law, accessibility / Section 508, language access). Mapped per workload.
Technical debt inventory. Every system flagged by IT, agency leadership, or the user community as fragile, end-of-life, or operationally painful. Not all of these are transformation candidates, but Discovery is where they get named.

Phase 2: Strategy (5 items)

Strategy is the documented choice about what to do, in what order, with what investment, gated by what success criteria. It happens after Discovery and before Pilot. Programs that invert these phases routinely waste 12 to 24 months.

Written program charter. One short document that names the program's goals, the scope, the leadership accountability, the budget envelope, and the success metrics. Signed by the executive sponsor (city manager, agency head, federal program executive). Updated annually.
Three-year roadmap. Phased, with named workstreams, named owners, and explicit dependencies. Cost ranges, not point estimates. Sensitivities called out. Reviewed quarterly.
Build-vs-buy framework. Per workstream and per major decision, a documented framework that scores build, buy, and hybrid options against total cost, risk, and strategic fit. Used as a decision artifact, not a slide.
Procurement vehicle map. For each workstream, the procurement vehicle that will be used (cooperative purchasing, state contract, GSA Schedule, direct RFP, sole-source justification). Verified with procurement and legal before committing to the workstream.
Workforce strategy. Roles affected by the program, training requirements, hiring requirements, role transitions, separation paths where applicable. Aligned with HR and (where applicable) union or works council.

Phase 3: Pilot (5 items)

Pilot is the proof that the strategy is real. Pilots that are not designed with explicit gates produce results that no one can act on. The five items below are the minimum hygiene for any government pilot.

Defined pilot scope. One workstream, one workflow, one quality measure. Documented. If the pilot scope is "an AI program" or "a cloud migration," the pilot will fail at the seams.
Baseline measurement. Captured before the pilot starts. Without baseline, "pilot success" is a vibe.
Go / no-go criteria. Written before the pilot starts. Specific. Quantitative where possible. Reviewed at the gate, not retroactively.
Stakeholder communication plan. Who hears about pilot progress, when, in what format. Includes the city council or agency oversight committee, includes the workforce, includes residents or end users where appropriate.
Pilot kill criteria. Explicit conditions under which the pilot stops. Programs that cannot kill failing pilots end up running them as zombies for years.

Phase 4: Scale (5 items)

Scale is the longest phase and the one most often skipped by programs that called pilot success "transformation." Scale is the actual work of changing the operating model, not just demonstrating that the technology can run.

Operating runbook. Per scaled workload: how it runs day-to-day, who owns it, how incidents are handled, how cost is managed. The transition from pilot to scale is a transition from "the project team owns it" to "the operating team owns it."
Workforce transition execution. Roles transitioned, training delivered, separations completed where applicable. The strategy made this commitment; scale is where it gets honored.
Cost optimization discipline. Especially for cloud-based scaling, a quarterly cost review with action items. Without this, year-two operating cost will exceed projections.
Continuous compliance. Per scaled workload, the controls are in place, the evidence is being collected, and the audit cycle has been completed at least once. Compliance is not a launch checkbox; it is a steady-state discipline.
Stakeholder transition. The program leadership that drove the transformation hands the operating responsibility to the standing organization. Without an explicit transition, the program leadership becomes operating staff and burns out.

Phase 5: Governance (5 items)

Governance is not a phase that ends. It is the discipline that runs through all five phases and continues after. The five items below are the governance hygiene any government digital transformation should be able to demonstrate.

Program steering committee. Meets regularly (monthly is a reasonable cadence during active transformation). Has decision authority on scope changes, budget reallocation, and gate reviews. Documented decisions.
Risk register. Maintained. Reviewed at every steering committee. Risks have owners and mitigation status. Not a slide deck.
Budget review cadence. Monthly burn review with finance. Variance commentary. Forecast updates. Year-over-year actuals.
Audit relationship. Internal audit, the inspector general (federal), the city auditor (municipal), or the state auditor is briefed on program plan and progress. Audit is a stakeholder, not an adversary discovering the program after a public incident.
Program closeout. When the transformation phase ends, an explicit closeout: lessons learned, formal handoff to operating organization, archival of program artifacts, post-mortem on what worked and what didn't. Programs without closeout have a way of never ending.

How to use the checklist

Score each of the 25 items green, yellow, or red against your current program. Green means done and demonstrably so. Yellow means in progress or partial. Red means not started or not real. Then count.

If your program is showing more than five reds, the program is not at the phase its plan says it is. The remediation is almost always to circle back to Discovery and Strategy, not to push harder on Pilot or Scale. This is unintuitive for leaders under political pressure to show forward motion, but skipping back is faster than the alternative of continued forward investment in a program without foundation.

Want this as a downloadable PDF?

Request our capability statement and we'll include a printable PDF version of this checklist, plus a worked example showing how we applied the framework on a representative engagement. Use the contact form and we'll respond within one business day.