The PQC Migration Budget Builder Turns Deadlines and Estate Counts Into a Defensible Budget Request

In May I wrote that nobody can tell you what PQC migration costs before discovery, and that the useful response is a structure rather than a better guess: a bounded year-one ask that can be itemized today, a ranged estimate for everything after discovery with the assumptions attached, and stage gates where the range narrows as the inventory fills in. For a large enterprise the article put year one at $3.5 million to $7.5 million and gave unit costs for the years that follow. The article gave the method and the unit costs; the spreadsheet was left to the reader.

The PQC Migration Budget Builder is that spreadsheet, built as a free tool on this site. You select the obligations that apply to you, enter what you know about your estate, state how much of the work existing refresh budgets already pay for, and the tool builds the request: an itemized year-one ask, a three-scenario estimate for the years after discovery, gross and net incremental cost side by side, annual cash flow with a CapEx and OpEx split, the cost of starting one, two or three years later, and a board paper in the order the article recommends.

Everything runs in your browser. Nothing is sent or stored.

Nobody can tell you what your migration will cost before discovery, so the tool builds the request you can defend before discovery, with every assumption visible, sourced and editable. The rest of this article covers what goes in, what comes out, where every number comes from and where the result will be wrong about your estate.

What the Budget Builder Is

Five steps and a results page. The first four collect inputs: organization, obligations, estate, refresh alignment and schedule. The fifth turns them into tables, charts and a draft paper. A headline strip at the top shows the year-one ask, the expected program total, the net incremental cost and the priority deadline, and it updates as you type, so you can see which input moves which number. You can save the inputs as a small JSON file, load them later, copy the paper as text, print it, and export the budget tables as CSV.

The numbers rest on four sources, and every constant in the tool is labelled with one of them:

• A, my May 2026 cost article, for the year-one line items, the testing ratio, the vendor and program-management lines, the cross-check bands and the order of the board paper.
• F, Phase 0 of the Applied Quantum PQC Migration Framework v3.0, for the size profiles, the 12-month buffer before a deadline, the refresh-alignment logic and the board indicators.
• T, the Global PQC Migration Timeline on this site, for the deadlines.
• C, calculator assumptions. These are mine, they are not published anywhere else, and each one says what it was calibrated against. Every one of them can be overridden in the assumptions register on the results page.

I had a different tool half-built last year: a PQC return-on-investment calculator that multiplied an exposure figure by a loss figure and reported a payback period. I never published it. On its own default inputs it reported a return of more than eighty to one, which no finance director would believe, and the result depended entirely on two inputs nobody can defend, the probability that a cryptographically relevant quantum computer exists in a given year and the loss it would cause you. That is the arithmetic of the quantum panic industry, and I did not want my name on a version of it.

The Budget Builder has no loss model. Its business case rests on dated obligations, which is where the real deadlines come from anyway: regulators, supervisors, card schemes and procurement rules have set dates, and boards fund obligations.

The same tool runs on QuantumAcademy.com for course participants, in an edition written without the first person and without references to this blog. The model is identical.

Who Should Use It

The tool was built for the person who has to put a number into next year’s budget cycle before the cryptographic inventory exists: a CISO, a CTO, a program lead, or the architect they have asked to do the work. It is also for the finance partner who receives that number and needs to check it against something, which is what the cross-checks are for, and for a board or risk-committee member who wants to know what a credible request looks like before one arrives.

The PQC Migration Advisor on this site answers a different question, which algorithms to replace, with what and by when. The Budget Builder covers the money.

It is less useful at two ends. If you have completed discovery, you have better numbers than any calculator, and you should use them; the tool’s job is the period before that. If you run a small organization with a handful of SaaS applications and no HSMs, the fixed lines dominate and the small-organization size factor is an extrapolation from the article’s large and mid-size figures, so treat the result as an order of magnitude.

How to Use It, Step by Step

Step 1. Organization

Choose the size profile that fits your estate. There are four, taken from the framework’s right-sizing profiles: small (under 2,000 employees), mid-size (2,000 to 20,000), large (20,000 to 100,000) and hyperscale or multi-entity (100,000 and more). The profile scales the lump-sum lines, governance, the discovery team, PKI modernization, by a factor of 0.12, 0.30, 1.0 or 2.0. The factors are calibrated so that a mid-size organization lands on the article’s mid-size figures ($250,000 to $700,000 pre-discovery, $300,000 to $1 million for discovery staffing) and extrapolated for the small and hyperscale profiles.

Choose your sector. It does two things: it suggests which obligations to look at on the next step, and it sets the per-person multiplier used in one of the cross-checks (1.5× for financial services and payments, 1.0× for everything else). Enter your annual IT budget and the number of people you serve, customers, citizens or members. Both feed cross-checks only; neither changes a cost line. Set the program start year, usually your next fiscal year. The currency selector changes the symbol only. Every unit cost in the tool is a US-dollar figure for a large enterprise, and I have not attempted regional labor adjustments.

Step 2. Obligations

Obligations set the schedule, which is why this step comes before the estate. The tool carries 33 regimes in five regional groups, curated from the 162 entries in the timeline as of October 2026. Each regime lists its instruments, its binding level (law, binding policy, supervisory guidance, draft, non-binding roadmap, corporate target) and its dated milestones, with a link to the coverage on this site. Open a region, tick what applies, and three dates come out:

• The priority-wave target. The earliest milestone of the kind “priority systems migrated” among your regimes, less a buffer. The buffer defaults to 12 months, the minimum in the framework’s model risk appetite statement. If you select the EU coordinated implementation roadmap, whose high-risk milestone is the end of 2030, the target becomes 2029.
• The full-migration target. The earliest “full migration” milestone among your regimes, less the same buffer. The EU roadmap’s end of 2035 becomes 2034.
• The first upcoming obligation. The next plan, inventory or procurement milestone after your start year. Milestones that fall before your start year are listed separately, with a note to report them in the year-one plan as done or in progress.

Two checkboxes record exposure, which sets no deadline: data that must stay confidential for more than ten years, which is the harvest-now, decrypt-later case, and long-lived signatures and trust anchors, which is the trust-now, forge-later case. They change the paper’s exposure paragraph and the cost-of-delay text, not the cost lines.

The binding level is shown because it varies more than most people expect. The US federal civilian schedule under OMB Memorandum M-26-15 and Executive Order 14412 binds agencies, with high-value assets due on PQC key establishment by the end of 2030 and the remaining systems by 2035, and a procurement rule for contractors has been directed. The EU roadmap is a coordinated plan with no legal force of its own; Germany, the Netherlands and Sweden have published national documents that follow it. PCI DSS v4.0.1 is an industry standard whose annual cryptographic inventory requirement has been mandatory since March 2025.

Supervisory letters to banks in the EU, Saudi Arabia, Israel, Singapore and Hong Kong are not laws, and a supervised bank ignores them at its peril. Decide which of these count as obligations for you; the tool takes whatever you tick.

Step 3. Estate

Enter what you know. Rough counts are fine at this stage; the point of the first stage gate is to replace them with the inventory’s figures. If you have no counts at all, one button fills in typical counts for your size profile, and you edit from there.

Applications come in four complexity tiers, following the split in my 120,000 Tasks article: modern applications with cryptography behind a library ($50,000 to $100,000 each), standard enterprise applications with configurable cryptography ($100,000 to $200,000), complex integrations and custom protocols ($200,000 to $400,000), and legacy systems with embedded cryptographic logic or a locked-in vendor ($400,000 to $750,000). Count only the applications that need migration work. The tool adds testing and validation at 25 to 40 percent of application migration on top, and it reminds you that the algorithm change itself is 10 to 20 percent of the total; the rest is integration, testing, vendor and operational work.

Cryptographic infrastructure takes two counts: HSMs and cryptographic modules to upgrade or revalidate ($50,000 to $200,000 per module, with FIPS 140-3 validation taking 12 to 24 months) and TLS, VPN and SSH endpoints ($300 to $1,500 per endpoint for configuration, certificates and firmware). If you already have a figure for HSM hardware replacement or for the extra CPU, memory and bandwidth that larger keys and handshakes will need, there is an optional panel for both. The tool does not invent those figures, because there is no defensible unit price for them.

Your starting position adjusts year one. Without a usable asset inventory the tool budgets $500,000 to $2 million (large enterprise) to build one before discovery; a partial inventory halves that; a mature CMDB removes it. Discovery tooling already licensed removes the licence line. Internal cryptographic skills scale the external advisory line: strong teams at 1×, security engineers without PQC depth at 1.5×, no in-house capability at 2×.

Operational technology and embedded devices get a scope selector and their own figure. The article was explicit that OT is a separate workstream with its own estimate, never a percentage of the IT migration, and the tool follows that: enter a number if you have one, and if you do not, the totals exclude OT and the paper says so in plain words.

Step 4. Refresh Alignment and Schedule

Part of any PQC program is planned refresh redirected to quantum-safe equipment. The framework asks you to present gross program cost and net incremental cost separately, and calls the gap between them the strongest single number in the budget conversation. Step 4 is where you set that gap, as the share of each line that existing budgets already fund. The defaults are starting points, each with a stated reason:

Move every slider to what your own refresh plans say. A CFO will ask.

The schedule has three inputs. The years the program needs at a sustainable pace default to six for a large enterprise, inside the article’s five to seven and the framework’s four to fifteen. The share of implementation in the priority wave, the systems regulators put first, defaults to 30 percent.

The rush premium defaults to 15 percent of implementation cost per year of compression, applied to the wave that is compressed. The mechanism is the article’s and the framework’s, parallel workstreams, premium external rates and expedited hardware, but the rate is mine, and you should set it to your own experience. A discount rate, default 8 percent, is used only if your CFO wants the cash flow discounted.

Step 5. The Budget Paper

The results page shows, in order: the headline figures; any warnings, such as a compressed priority wave or an OT scope you marked as significant without entering a figure; the three-scenario range against the article’s reference scenarios; cash flow by year with the CapEx and OpEx split and the net present value; the year-one ask itemized line by line; the post-discovery estimate by category; gross and net incremental cost per line; the cost of starting one, two or three years later; the cross-checks; and the board paper itself.

The assumptions register at the bottom lists every unit cost with its source and lets you change it, and the whole page recomputes.

The paper follows the order from the article: a one-page executive summary with the regulatory driver, the approach, the year-one ask and the total range; the detailed budget with annual cash flows, contingency as a separate line, year one itemized and later years by category; a risk section with the cost of delay, the exposure paragraph, the six cost traps and the nine benefit buckets from the article; and a governance section with the accountable executive, the steering committee, the stage gates and the five board indicators the framework uses, Coverage, Trust, Inventory, Vendors and Agility. Copy it as text into your own template, or print it.

A Worked Example for a Large EU Bank Starting in 2027

The example that loads with the tool is a large financial institution with a $100 million IT budget and two million customers, 104 applications needing migration work (60, 30, 10 and 4 across the four tiers), 15 HSMs, 2,000 endpoints, a partial inventory, security engineers without PQC depth, and three obligations ticked: the EU coordinated implementation roadmap, DORA, and PCI DSS v4.0.1. The program starts in 2027.

The schedule comes out first. The EU roadmap’s high-risk milestone is the end of 2030; with the 12-month buffer the priority-wave target is 2029. A six-year program at a sustainable pace would reach the end of its priority wave in 2030, a year late, so the tool compresses that wave by one year and prices a 15 percent premium on the 30 percent of implementation inside it. The full-migration target is 2034, two years of slack against the six years needed, so the rest of the program runs uncompressed.

Two milestones are already behind the start date, DORA’s application in January 2025 and the PCI DSS inventory requirement from March 2025, and the paper lists them as items to report as done or in progress.

The cross-checks on this example all point the same way. The expected total is 6.8 percent of the IT budget over the program years, below the 8 to 15 percent that comparable transitions such as ERP, cloud and zero trust have taken, and below the article’s expected scenario of 7.1 percent. It is $20 per customer against a comparator of $31.50, the US federal estimate of about $21 per person served times the 1.5 I apply for financial services.

The conservative scenario is 1.45 times the expected one, below the 1.5 to 2× growth the article expects post-discovery estimates to absorb over a program’s life, and the tool flags it. The blended CapEx share is 43 percent, below the 60 to 70 percent rule of thumb, because at these defaults the program is labor-heavy.

My reading of the four checks together is that for a bank of this size the estate counts are light, and the right move is to widen the conservative scenario or revisit the application tiers before the paper goes anywhere.

The Assumptions Behind Every Number

The table below is the short version of the assumptions register. The full version, with the cash-flow profiles and the CapEx shares by category, ships with the tool as MODEL.md.

Scenarios are simple: optimistic is the sum of the low ends, expected the sum of the midpoints, conservative the sum of the high ends. Contingency is conservative minus expected, presented as a managed line, which is how the article says to present it to a CFO.

Building the tool also turned up two places where my own article does not add up, and I would rather say so here than have a reader find them. The year-one ask is stated as $3.5 million to $7.5 million, but its components, $1 million to $2.5 million pre-discovery and $2 million to $5 million for discovery, sum to $3.0 million to $7.5 million, and the optimistic column of the article’s table also sums to $3.0 million. The program-management line in the table, $2 million to $5 million over six years, is smaller than six years of the year-one director and program-office lines alone.

The tool follows the table and reports both year-one figures in its cross-check, and the article will get a correction note.

Where the Model Will Be Wrong About Your Estate

• Every unit cost is a 2026 US-dollar figure for a large enterprise. Labor rates in Mumbai, Munich and Minneapolis are not the same, and the tool does not adjust for them. Scale the unit costs in the register if you know your rates.
• The small and hyperscale size factors are extrapolations. The mid-size and large profiles are anchored to the article; the other two are not.
• The cash-flow profiles are fixed shapes: infrastructure in years two to four, applications rising then falling, program management flat. Your sequencing will differ.
• HSM hardware replacement, capacity uplift and OT are your figures or nothing. The tool does not estimate them.
• The deadline data is a snapshot from October 2026. Regimes change, and several entries are guidance or non-binding roadmaps, which the binding level shows. The version number in the tool’s footer tells you which data you are looking at, and I update the list when the timeline changes.
• The cost of delay prices compression and lost refresh alignment. It does not price enforcement action, lost contracts or breach losses, because I do not have defensible numbers for those and neither does anyone else.
• The board paper is a draft written from your inputs. Check every sentence before it goes to anyone; it states your assumptions as facts because that is what you told it.

How Far to Trust the Output

Treat the output as a planning aid. Nothing in it is a quotation, an audit, or legal, financial or regulatory advice, and the tool knows nothing about your contracts, your supervisors or your accounting policy. The obligations it lists are my reading of public instruments as of the data date; check the ones you rely on against the instrument itself, and with your counsel, before they appear in a board paper under your name. The figures are ranges built from published benchmarks and from assumptions I have labelled as mine, and they carry no warranty of any kind.

Nothing you enter leaves your browser: there is no account, no analytics on the inputs, and no copy of your paper anywhere except your clipboard or your download folder. The tool is published under a Creative Commons Attribution 4.0 licence, so you may adapt it for your own organization or your clients, with attribution. If you find an error in a figure, a source or a deadline, write to me; corrections go into the next version with the change noted.

Taking the Paper to Your CFO and Board

Lead with the obligation. The first paragraph of the executive summary names the regime, the date and the buffer, because that is what a board can act on; a Q-Day forecast is not. Then the year-one ask, itemized, because it is the only number in the paper you can defend line by line today. Then the range, with the three scenarios and the contingency as a managed line rather than as a confession of uncertainty.

Put gross and net on the same page. In the example, $4.3 million of a $40.9 million program is money already planned for PKI automation, HSM refresh and network renewal. For many organizations that share will be far higher once the refresh calendars are in front of the right people, and the net figure is the one that goes into the incremental budget request. The framework calls this the infrastructure modernization umbrella.

Fix the stage gates in the paper and keep them. The first gate is the end of discovery, when the inventory replaces the estate counts: load your saved inputs, change the counts, and the paper shows the delta. If the range narrowed, say so. If it widened, say that too, with the reason. A program that reports a wider range with a reason keeps its budget; one that quietly reissues a bigger total does not. Board-level PQC oversight covers what directors should expect to see at each gate.

Run the cross-checks before anyone else does. A finance partner will divide your total by the IT budget and by headcount within minutes of receiving the paper. If your number is below 8 percent of IT budget for a multi-year program with a complex estate, be ready to explain why, or raise the conservative scenario before you send it.

What Changes at the First Stage Gate

Discovery will replace most of the counts you enter today, usually upward: inventories find more applications, and more places where cryptography is used, than anyone listed beforehand. The structure of the request does not change. A year-one ask you can itemize, a range you can defend, a net figure your CFO can compare with money already planned, and a date that is not yours to move. The Budget Builder gives you that structure in an afternoon. The inventory, the vendors and your own refresh calendar supply the rest.

From the author