Casey

that specific error makes annual customers look more valuable per month than monthlies, which then tempts you to discount annual harder. it compounds quietly.

at 74 customers i would track both and care mostly about cash, honestly. mrr is the number for talking to people who benchmark companies, and it is also the number that makes you feel worse than your bank account does when annuals are a big share. the operational question you actually face this quarter is whether you can pay the bills in a month where nothing renews.

Just checked. There is in fact a NAT Gateway. I have never knowingly used it, I do not have anything in a private subnet, and it has apparently been sitting there since February being paid for out of my credits.

Mine is post-change, so this is the one I needed to know. Upgrading tonight rather than finding out on Saturday.

Worth saying those list prices are for the base finishes. The coloured versions of the Gaggia carry a premium over the stainless one, which catches people out when they price it from a photo.

Pen and paper genuinely works for a household of two, provided you both look at the same page once a week. What an app adds isn't insight, it's a shared source of truth neither of you has to transcribe. That's worth $100 to some couples and nothing at all to others.

What I did after mine, for what it is worth: moved to the paid plan on day one and mentally reclassified the credits as a discount rather than a countdown.

The entire problem with a free plan that ends is that it creates a cliff on a date you will not be thinking about. On a paid plan with credits applied there is no cliff - the balance drains, the bill starts, and the alerting you have already set catches it. Same money, no single night where everything can disappear.

Run the margin arithmetic on the roll-over question before you decide anything else, because it is the clause that quietly decides whether this is a business or a hobby.

Hours that do not roll over: you are paid for capacity, unused months are margin, and that margin is what funds the months where everything catches fire. Hours that roll over: you have written a liability onto your own books that gets called in at the worst possible time, usually as "we have banked eighteen hours, we would like a new feature", which is not support at all.

If a client will only sign with roll-over, cap it. Three months maximum, expires after that, stated on the invoice every month so it is never a surprise.

Automate the part that requires no decisions and stop tracking the rest. Separate account for fixed bills, standing transfer on payday, and whatever's left in the day-to-day account is spendable. I stopped categorising entirely three years ago and my finances improved, which is not what any budgeting app wants me to say out loud.

I rented the same flat for six years then bought two streets away, so I have both sets of numbers in one spreadsheet. Years one and two cost me clearly more than renting once fees, a new consumer unit and a washing machine were in. It crossed over somewhere in year four and by year six it was not close, mostly because the rent on my old flat went up twice while my payment did not. Five years is roughly where this stops being a gamble in my experience.

And it is not a smooth one percent. It is nothing for three years and then a boiler and a roof in the same autumn, which is why the buffer matters more than the average.

The free tiers are mostly demos. What I'd do in your position is take the month-long trial of a paid one and run it properly for a full cycle, including the awkward joint-account part, then decide. A month of real use tells you more than any comparison thread, including this one.

Every app worth using supports it, they just don't explain it during onboarding. Some auto-detect the transfer and some make you match it by hand every month.

Delamination after seven months in standing water isn't really a brand problem, it's an adhesive-and-water problem, and every glued shoe has it. What I look for now is a rubber outsole that wraps up over the toe rather than a foam edge with a rubber strip stuck onto it.

This is what I suspected and did not want to hear from myself. The evenings argument is the one that lands.

The re-auth churn is mostly upstream of whichever app you use, because aggregators are still scraping a lot of Canadian institutions rather than using a sanctioned API. The framework meant to fix that was legislated in 2024, oversight moved to the Bank of Canada in the 2025 budget, and read access was targeted for early 2026 with payment initiation later. It has slipped, and the big banks are the mandatory participants, so a credit union is likely to be last in line regardless.

Not the case here but I can see how that would look identical.

The aggregator had the ex-date a week off from the company page. That was it.

We sat at about 34 percent for three years and it was fine but joyless, in the sense that nothing broke and nothing exciting happened either. The people I know who got into trouble were not at a particular percentage, they were at a high percentage with variable income and no buffer, which is exactly the combination you have described. If you do it, underwrite the payment against the stable income alone and treat the freelance side as a bonus that fills the buffer.

bars for the month, line for the trailing three. that is a five minute change to the sheet and it fixes the thing i actually complained about.

four components instead of one net figure is one extra column and it answers the question i actually have, which is whether i am leaking or just not adding.

And as of this July the CDR scope widened past banks and energy into non-bank lenders and buy-now-pay-later, so coverage is meaningfully better than it was a year ago.

Separate from the change question: check whether your specialist can be covered as out-of-network under your new plan and what that actually costs after the deductible, and whether there is any continuity-of-care provision for treatment already in progress. Some plans have one, some do not, and it is worth one phone call to the insurer before you assume nine months of full price. Get whatever they tell you in writing through the member portal message system.

This is the trap I fell into for years, valuing redemptions against fares I would never have bought and congratulating myself.

Also check whether the payout ratio is quoted on adjusted earnings. A company that has adjusted out the same charge every year for a decade is not adjusting, it is hiding a cost.