Plan future changes to income and outgoings, with a date on every one

This is the distinction most tools blur, and it is the difference between a forecast you trust and one you argue with. A scenario is a parallel branch for something that might happen. Should we downsize. What if one of us went part time. You build it alongside your real plan, compare the two, and only one ever becomes true.

A planned change is not a branch. It is a single dated fact inside your one real plan, for something you already know is coming, with no decision left to make about it. Most tools leave you two bad options: ignore it and let a flat line lie to you, or maintain a second version of your finances for something that was never in doubt. CrestCast treats it as what it is, an event with a date.

The other half of the picture is scenario comparison, for the genuine what-ifs. Both sit inside the same forecast, alongside everything else CrestCast does.

Everything starts from something you have already entered. You pick the item, pick a date, and pick what happens to it. Against anything recurring: a new amount, a percentage change, the fact that it stops, a rate change on a mortgage or a loan, or a larger overpayment from that date on.

Separately you can add one-off future events that are not attached to a recurring item at all:

• Money in on a date. A bonus, a gift, a tax refund, a redundancy payment.
• Money out on a date. A wedding, a new roof, a car deposit.
• A one-off overpayment aimed at a debt you choose, rather than at your debts in general.
• Selling an asset, which is a bigger piece of machinery than it looks and gets its own section below.

One smaller option removes a common bodge. A bill can be set to be charged only in certain months of the year, so a cost that lands in term time only needs no invented monthly average.

This is the one that usually goes wrong by hand. Selling a house or a car inside a forecast means four things happen at once, and by hand you have to remember all four and keep them agreeing. In CrestCast you pick the asset and the month.

• The proceeds arrive as income in the month of the sale.
• The mortgage secured on that asset is settled out of the proceeds.
• The asset stops contributing from the sale month onward.
• Both the cash balance and net worth reflect all of the above, in the same pass.

The mistake is invisible when you make it by hand. Forget to clear the mortgage and your net worth quietly gains a couple of hundred thousand pounds. Clear the mortgage but forget the proceeds and the cash line falls off a cliff. Both look like a working forecast, and you will not spot which is right from the shape of the chart.

Pay is the input people most want to change on a date, so it gets its own schedule rather than being treated as a generic amount. On any date you either set a new salary, entered as a new gross or a new take-home depending on which number you know, or you stop that income. Income tax, National Insurance, pension contributions and student loan repayments are all recalculated from that month, so a rise of £4,000 does not turn into £4,000 more in your pocket on the chart.

With more than one job you choose which job the change applies to. Underneath, tax and student loan are worked out against one set of thresholds across all of your income and then apportioned back per job, which is what actually happens and what a per-job calculation gets wrong.

Entering zero is how you model redundancy. It is a bleak thing to type into a form and it may be the most useful planned change in the app, because the answer you want is not a feeling about whether you would cope. It is a month.

Say the mortgage is £220,000 with the rate fixed at 4.6% until March 2028. Nursery is £950 a month and stops in September 2027 when your youngest starts school. And you already know a promotion lands in July 2027, taking the gross from £48,000 to £54,000.

Those are three dated facts, not three scenarios. You schedule the rate change for March 2028, tell the nursery bill it stops in September 2027, and put the salary change in the schedule for July 2027. Your one plan now has a step up in take-home that summer, a £950 a month improvement that autumn, and a jump in the mortgage payment in early 2028 you can see coming from years out.

A planned change is not a note on a calendar and it is not a reminder. It goes into the forecast engine, so it lands in the cashflow, the profit and loss and net worth at the same time, from one entry. A dated overpayment increase re-amortises the debt. A stopped bill stops appearing as an expense. A sale moves an asset out of net worth and the proceeds into cash.

Milestones are drawn on the forecast charts by default and derived for you rather than typed in. Every dated change you have scheduled is flagged, alongside a debt clearing, a promotional rate ending and a mortgage deal ending. The chart is not just a line, it is a line with the reasons on it. To read those changes month by month, that is what cashflow forecasting is for.

In a spreadsheet a future change is not an event. It is a formula with a date condition inside it, repeated down every affected row. Get it wrong on one row out of forty and you have a nursery bill that half stops, which no error message will ever tell you about.

Then it has to be right three times over, because the cashflow tab, the profit and loss tab and the net worth tab each carry their own version of the condition. A sale is the worst case: proceeds in one tab, the mortgage clearing in another, the asset disappearing from a third, all keyed off a month you typed in three places.

Then six months pass and nobody remembers which cells carried the condition. The rows that were missed do not break. They keep producing plausible looking numbers, which is the failure mode that costs you money, because you never go looking for it. A dated event you enter once cannot drift out of agreement with itself.

Not everything takes a dated change, and the boundary is worth seeing before you build a plan around one. Here is what takes a scheduled change and what does not.

• Full support: income sources, bills including joint bills, contributions into a joint account, and salary.
• Mortgages and loans: rate changes and a dated increase to an overpayment, scheduled from the Schedule a Change flow rather than from inside the mortgage form itself.
• A partner or household member's income: takes a dated change through Schedule a Change, though not from their own edit screen.
• Assets: a start date and a dated sale, and nothing else.
• Transfers between household members: no dated changes at all.
• Credit cards and student loans: cannot take a scheduled change. The app offers the option and then disables it, with the reason stated on screen.
• Charged only in some months: exists on bills only. An income that only arrives in term time cannot be expressed that way.

Two quirks of the list itself. A scheduled change to a joint account contribution, and a change to a partner or household member income, both affect the forecast correctly but do not currently appear in the list of planned changes. They work. They are just not listed. And there is no bank connection anywhere in CrestCast, so nothing is detected for you: a planned change is something you tell it, deliberately.

Dated changes matter most when a household is running several at once, which is the normal case. One salary changes in the summer, childcare falls away in the autumn, a fixed rate ends the following spring. Individually each is easy to reason about. Together they are not, and that is where people stop forecasting and start hoping. All of them land in the same shared household forecast, so both of you work from the same set of dates. If the plan is to put the freed-up money somewhere useful, cash routing is what decides where the surplus goes once a cost drops away.