CrestCast vs ProjectionLab

The first difference is the clock. ProjectionLab's help centre answers the month-by-month question directly: at the moment it operates on a year-by-year basis, and the reason it gives is speed, because year-by-year computation is what lets the whole simulation redraw the instant you change something. Across a plan measured in decades that is the right call, and a single day is noise. CrestCast is built on the opposite decision. The day is the unit, and the month, the quarter and the year are added up from the days rather than estimated separately.

That matters for a specific class of question. An annual premium leaves on the 16th and payday is the 27th. A bonus lands in the month a school fee is due. A 0% deal ends four days before the balance is cleared. Averaged over a year, none of this exists. Day by day it is the difference between a plan that holds and a fortnight of unplanned overdraft.

The second difference is who the product thinks the user is. ProjectionLab is set up as one plan, and you tell it whether you are planning as an individual or as a couple. Since 2022 it has let couples define separate and joint ownership for income, assets and expenses, and it breaks some yearly summaries down by person, so it is not a single-person tool with a spouse bolted on. CrestCast starts from the household instead: both incomes, joint and personal accounts kept apart and then added up, children, and one subscription with a login for each of you. Both of you can open it and see the joint picture, the household picture, or your own.

The third is that CrestCast is UK the whole way down rather than a UK layer over a general model. Income tax and National Insurance, the student loan plans, ISAs and pensions, all inside the one subscription. That subscription is priced in pounds too, so what you pay does not move with the exchange rate.

And CrestCast can act on the gap between two dates rather than only drawing it. A spare cash waterfall is a rule you write once: hold a minimum in the current account, sweep what is above it into the pots you choose, and if a month comes up short draw the difference back from the savings you name, in the order you name them. The forecast then runs with that rule in force, which is what turns a projection into a plan you could follow.

Choose CrestCast if the question is the near one, and it is a household question. Can we afford the bigger house. What happens to us if one of us goes part time. When is the mortgage actually gone, and would overpaying beat saving. Which day of the month does the account get tight. Saving two versions of your plan and putting them side by side works because the model underneath is a full three-way statement rather than a balance curve.

Choose ProjectionLab if the question is the long one. It runs a whole-life plan with life expectancy milestones for you and your partner, and answers when you could stop working and what the drawdown after that looks like, with Monte Carlo simulation and historical backtesting behind it. Its UK work is real and recent, with pension withdrawal methods for UK defined contribution pensions added during 2026. If retirement modelling is what you are shopping for, that is the tool that sells it, and CrestCast does not.

Unusually, neither product connects to your bank. ProjectionLab's help centre states that linking financial accounts directly is not supported, and it offers third-party plugins that pull balances from Lunch Money, YNAB and Monarch Money. CrestCast reads nothing from your accounts either. You enter what it knows, or hand it the CSV your own bank exported, which is read on your device and never uploaded. Here is why I built it that way.