What a financial plan is actually for.
A couple sits down across the table. They are sixty-three and sixty-one. One of them says, with some pride, "We have a plan."
What they have, when we look at it, is a portfolio and a number. The portfolio is well-built — low-cost funds, a reasonable allocation, two decades of disciplined contributions. The number is a probability: 87% chance of success, generated by a tool their old advisor ran once, three years ago. They have not opened the report since.
That is not a plan. It is a snapshot of a portfolio with a confidence interval printed on it. The distinction matters, because the thing they actually need — and the thing most people mean when they say the word — is something else entirely.
What people think a plan is
Ask ten households what a financial plan is and you'll get two answers. The first is the binder: a forty- or eighty-page bound document, full of charts, that arrives once and then lives in a drawer. The second is the score: "we're at 87%."
Both are artifacts of planning. Neither is a plan.
The binder is a record of a calculation. The score is a summary statistic about that calculation. You can hold both in your hands and still not know what to do next March when you actually retire — which account to draw from, when to claim Social Security, whether to convert anything to Roth this year, what the year your pension starts will look like. The binder doesn't tell you, and the score certainly doesn't.
What a plan actually is
A plan is a set of decisions, dated. That's the whole thing.
It is the answer to "given everything we know about your income, your accounts, your taxes, and the years ahead — what should you do, in which year, and why." It is specific. It names accounts. It names years. It changes when the facts change. And it is worth something precisely to the degree that it changes a decision you would otherwise have made on autopilot.
Here is where a plan earns its keep — the decisions that have a right answer and a default answer, and where the gap between them is measured in real dollars:
- When to claim Social Security. The benefit at 70 is roughly 77% larger, per month, than the benefit at 62 for someone with a full retirement age of 67 — a difference driven by delayed retirement credits and the early-claiming reduction. The default is "claim when I stop working." The planned answer depends on longevity, the other spouse's benefit, and the tax picture in the years between.
- Which account to draw from first. The default heuristic is "taxable, then tax-deferred, then Roth." The planned answer is computed year by year, against the marginal cost of the next dollar from each account — and in high-income years the heuristic is often wrong.
- Whether to convert to Roth, and how much. A temporarily low-bracket year — say, retirement at 63 before Social Security and RMDs begin — is an opportunity that closes. The default is to do nothing. The planned answer fills a bracket deliberately, watching the IRMAA surcharge two years forward.
- Pension versus lump sum, the survivor year, the timing of a home sale. Each is a decision that lands in a specific year and interacts with the tax line in that year.
None of these is answered by "87%." Every one of them is answered by a plan.
The cost of the default
Take the couple above. Suppose they retire at sixty-three with $1.9M in a 401(k), $300K in a Roth, and $250K in cash, and they spend the next several years on autopilot: claim Social Security the month they retire, draw from the 401(k) because that's where the money is, convert nothing.
Now run the same household with the decisions made deliberately. Defer Social Security and bridge the gap with cash and taxable drawdowns. Use the low-bracket years before RMDs to convert into the top of a low bracket. Sequence withdrawals against the actual marginal cost in each year rather than the heuristic.
The two households own identical portfolios. They earn identical returns. The difference between them is entirely decisions — and over a thirty-year retirement that difference routinely runs into six figures of lifetime tax and a materially larger guaranteed income floor in late retirement, when it matters most. The portfolio didn't create that gap. The plan did.
This is the part that gets lost. People shop for returns and pay attention to fees, both of which matter, and then leave the largest controllable variable — the sequence and timing of their own decisions — entirely to the default. The default is not neutral. It is just a decision nobody made on purpose.
Why the number isn't the plan
The probability-of-success score has a real use: it tells you whether the plan, as a whole, survives a range of futures. It is a good thermometer. It is a terrible map.
When the score says 87%, it cannot tell you why the 13% of failed scenarios failed — whether they failed on pure return-shortfall risk or on a specific, fixable timing decision. It cannot tell you which year to change. A plan that is only a number is a plan you cannot act on, because the number isn't a decision; it's a grade on decisions you may not have explicitly made. We've written before about why the same probability can mean very different things depending on what the model underneath it actually knows.
The score is downstream of the plan. It is not a substitute for it.
What this means at the kitchen table
The reason any of this matters is that the conversation a household actually wants is not "what is our probability of success." It is "can we keep doing what we're doing," and "what changes when one of us isn't here," and "what does the year the pension starts look like." Those are decision questions. They have dated, specific, dollar-valued answers. A plan is the document that holds those answers and updates them as the facts move.
If the only thing your current plan can tell you is a number, it isn't doing the work. The number is the easy part. The decisions are the plan, and the decisions are what change the outcome.
We built Foundry Planning around that conviction: the deliverable a client takes home should be the decisions, in their own years, in plain language — not a portfolio projection with a confidence interval stapled to the front. If you want to see what decision-first planning looks like in the room, the pricing page is the place to start.
What's next
This is the first of three short notes on the same idea from different angles. The next looks at the artifact itself — what a clean client deliverable actually contains, and the binder it replaces. The third ties them together: the plan is the deliverable — they were never two separate things.
Working notes on planning, written for the people who do the work. If there's a planning conversation you'd like covered, support@foundryplanning.com.