Your first forecast

There are three ways in: the quick start, a copy of a demo household, or a blank plan you fill in yourself. If this is your first plan, pick one of the first two. Both put a working forecast in front of you in a minute or less, and a blank form is much easier to fill in once you have seen what it produces.

Your own forecast, in about a minute

The quick start is on the home page, and again inside the app under New plan. It asks for:

1. Your age, from a dropdown that starts at 35
2. The age you plan to retire, or a tick to say you already have
3. The state you live in, for the tax tables
4. What you have saved and invested, in total, and roughly where it sits
5. What you earn in a year, before tax
6. What you spend in a year, everything included

Only the three amounts are typed; the ages and the state are dropdowns that start on a plausible answer, and where your savings sit is one tap you can skip. That is enough to run the same engine every other plan runs: real federal and state tax tables, 1,000 simulated market histories, the lot. You land on Projections with a forecast built from your own numbers, and the plan is yours from that moment. It is saved, it is in the sidebar, and every field in it is editable. Nothing is read-only and there is no copy to make.

What it had to guess, and how to fix it

A handful of answers do not describe a household, so the rest is filled in for you. Every guess is listed on a chip at the top of the plan. Click one to open the field it belongs to, or tick it once you have checked it, and it goes away. When the last one goes, so does the banner.

Each guess is also marked Assumed on the Inputs tab, beside the field it belongs to, so you can see which numbers are yours and which are ours while you are working through them. Clicking a marker confirms that one, the same as ticking its chip.

What was guessed	The guess it made
Social Security	Estimated from your income and how long you will have worked, and adjusted for when you are assumed to claim it: never before you stop working, at 62 if the savings would not comfortably bridge the gap to then, otherwise at 67. Someone already retired and 62 or over is given the median benefit instead.
How your savings split	Divided across a 401(k) and a brokerage, following the tile you picked. Half and half if you picked none.
Cost basis	What you paid for the taxable holdings. Assumed to carry a decent unrealized gain, growing with your age and the size of the account.
Allocation	A stock share set from your age, with a cash buffer that is a larger share of a small balance than of a large one.
Spending in retirement	The same as what you spend today, for the whole plan.
Household	Single, with no spouse.
Birth month	January, so the age you typed is the age this plan uses all year.
State of residence	California, if you left the dropdown where it started. State tax rules differ enough to move the whole forecast.

A quick-start plan cannot be shared until you have been through that list. The share control explains itself rather than handing you a link: a share link is you showing someone an outcome, and the parts of this one nobody has checked are ours rather than yours. Confirming them takes a few clicks and makes the link worth sending.

Two things worth knowing about the list itself. The guesses lean conservative on purpose, so filling in the real numbers usually makes your forecast better rather than worse. And a debt is not one of them: a mortgage or a loan payment is inside the spending figure you gave, which means it never ends. Adding the loan itself under Liabilities is often the single largest correction you can make.

Or start with a demo plan

Open the app and look at the left sidebar. Under Demo plans there are four complete households already built: a family with a mortgage and kids, an engineer trying to retire at 37, a couple whose house is most of their plan, and a Wall Street exit at 55.

Click one. Every tab works immediately, because the plan is already full of numbers. Nothing you do to a demo can break anything: they are read-only, and the moment you try to edit one, the app offers to make you an editable copy instead.

Poke at a demo for two minutes before you build your own. It is much easier to recognize what a projection is telling you when you did not also just spend twenty minutes typing in balances.

Building one by hand

New plan is the + in the sidebar’s Plans header. You can also start from a copy of any demo, or from the quick start above, both of which are less work than starting blank.

Six things need to be there before a projection will run:

1. Household. Single or couple, and each person’s birth month and year. The birth month is not decorative: it decides RMD ages, early withdrawal penalty ages, and when the extra standard deduction at 65 kicks in.
2. State of residence. Required, because state income tax is modeled for all 50 states and DC. There is no national average to fall back on.
3. Planning-till age. How long the plan should run. The default is 90. It must be at least 15 years past the person’s current age.
4. Savings and investments. Balances by account, and cost basis on taxable accounts. Savings and investments only: not your house, not your cars.
5. Income and expenses. What comes in and what goes out, and for which years.
6. Portfolio allocation. Stocks, bonds, and cash. The default is 60/20/20. This is your target for the portfolio as a whole, and it is what new savings buy. Each account also carries its own mix, set when you add it, and the projection grows each one at the rate its own mix earns. A rebalancing strategy sits next to it: the default steers toward the target without forcing a taxable sale, and Let it drift never trades to a mix at all. How rebalancing works.

If Projections, Cash Flow, Details, and Outcomes are blocked, one of the first three is missing. Inputs will tell you which.

What you will not be asked for

A rate of return. There is no box for it, and that is deliberate. Typing “7%” into a retirement calculator is how you get a forecast that has quietly assumed the market never has a bad decade at the worst possible moment. Returns, inflation, and taxes are all simulated or taken from statutory tables instead. See where the market numbers come from.

A tax rate. Same reasoning. See how taxes are estimated.

Where to look first

Once a plan runs, two tabs are worth your time in this order:

Projections shows one possible future in detail: a line for your liquid assets, with income, spending, and taxes as yearly bars underneath. The line is jagged, and that is not a bug. It is one real simulated path rather than a smooth average, and watching it is the point. When the tab loads it flashes through fifteen neighboring simulations before settling, so you can see how differently things could have gone.

Outcomes shows the distribution instead: the chance of success across all 1,000 paths, and a fan showing where the middle of the pack lands each year. This is where the headline number lives. What that percentage means is worth reading before you act on it.

Then try breaking it

The forecast is more useful as a comparison than as a prediction. Clone your plan and change one thing: retire three years earlier, spend $500 a month more, move the allocation to 40% stocks. The market is held constant across plans, so the two are genuinely comparable and any difference you see came from your edit, not from a different roll of the dice.