Corrections
A corrections policy with nothing logged against it is just a claim. Here's the actual list, with what was wrong and whether it would have changed a decision.
When a page here is wrong, the page gets fixed and the error gets logged below with the date. Nothing is quietly swapped out and nothing is deleted from this list later. If you find something wrong, send it. Half the entries below came from outside readers.
One distinction worth drawing. A figure that changes because the law changed is not a correction, so you won't find those here. Those pages state which tax year the number applies to, and the number moves when the year does. A correction means we published something that was wrong when we published it.
The log
-
A $42,000 account was said to last a year and a half at $55,000 a year of spending
It lasts about nine months. The worked example gave Maria a $42,000 taxable brokerage account against an annual need of $55,000 and then described her spending it down "over roughly the first year and a half." Dividing one by the other gives 0.76 years. The sentence now says nine months.
The same paragraph also glossed over something the rest of the page is careful about. The $55,000 is what she needs to spend, so the amount she withdraws has to be larger than that to cover the tax on the withdrawal itself. That's now stated where the example makes the claim.
Would it have changed a decision? Yes. The point of that example is how long the taxable account buys you before you start touching the IRA, and the error doubled it. Anyone reading it to estimate their own bridge would have planned for twice the runway they had.
-
ACA premium tax credits were described as based on taxable income
They're based on modified adjusted gross income. The page said so correctly in one section and incorrectly in six other places, including the intro, the meta description and two FAQ answers. All six now say MAGI.
This mattered more than a vocabulary slip, because for marketplace purposes MAGI is deliberately broader than taxable income. It adds back tax-exempt interest and the untaxed portion of Social Security. So municipal bond income counts toward your subsidy eligibility even though it's never taxed, which is exactly the situation of the household that page is written for. The contrast is now spelled out rather than assumed.
Would it have changed a decision? Yes, for a reader holding municipal bonds or already drawing Social Security. Planning a year's income against taxable income instead of MAGI can understate the figure the subsidy cliff is measured against, and the cliff is a cliff.
-
The target ignored the years between retiring and claiming benefits
The calculator subtracted Social Security and pension income from the spending your portfolio has to cover, which is right, but it did so for every year of retirement. If you retire at 60 and claim at 67, that income doesn't exist for seven years, and the portfolio has to carry the whole bill in the meantime.
The tool now asks when benefits start and funds the gap between the two ages as a separate line in the target.
Would it have changed a decision? Yes, and in the dangerous direction: it understated the target. For someone retiring at 60, claiming at 70 and expecting $30,000 a year of benefits, the missing bridge was roughly $300,000.
-
The page carried FAQ structured data for questions a visitor couldn't see
The calculator sent search engines a block of questions and answers in its markup that appeared nowhere on the page itself. Google's own structured data guidelines require that the content be visible to the person visiting, so this was a straightforward violation on our part, not a grey area.
The questions are now rendered on the page, and they're generated from the same source as the markup so the two can't drift apart later.
Would it have changed a decision? No, nothing stated was inaccurate. Logging it anyway, because a site that only admits the errors making it look careless rather than the ones making it look sloppy isn't really keeping a log.
-
The saved PDF printed the investment return under the "withdrawal rate" label
The summary you can save from the calculator has a row for each assumption. The withdrawal rate row was showing the expected investment return instead, so a PDF built with a 4% withdrawal rate and a 5% return read "withdrawal rate 5.0%". The figures used in the calculation were always correct. Only the printed label was wrong.
Would it have changed a decision? Possibly, for anyone who saved the PDF and later worked from it instead of the tool. The number in the calculator was right, so the on-screen result was never affected.
-
The homepage target read as an answer rather than an estimate
The figure in the hero calculator was labelled "what your portfolio has to cover" and shown to the dollar. Both of those overstate what the arithmetic behind it can support, since it rests entirely on a withdrawal rate that reasonable people set anywhere from 3.5% to 4.5%.
It's now labelled an estimated portfolio target, and the card shows what the same spending needs across that whole range. On the default example that's a spread of roughly $290,000, which is the honest size of the uncertainty.
Would it have changed a decision? Not the arithmetic, which was always right. But a single exact number invites people to treat the difference between $1.1 million and $1.2 million as meaningful when it sits well inside the margin, and that does change behaviour.
What counts as a correction
Three categories get logged here: a figure that was wrong, a claim that was wrong, and a definition that was wrong. Anything that would have left a reader with a false belief about how something works.
Three categories don't. A number that changed because a contribution limit or a tax bracket changed isn't a correction, it's an update, and those pages carry the tax year the figure applies to. A typo that doesn't alter meaning gets fixed silently. And a page rewritten to explain the same thing better isn't correcting anything.
Where the line is genuinely unclear, the entry goes in the log. An over-inclusive corrections page costs us nothing except some embarrassment. An under-inclusive one would make this whole page worthless.
Reporting something
Use the contact page. The most useful report names the page, quotes the sentence, and says what the right figure or definition is. That can be checked and fixed the same day. A report saying a page "seems off" takes considerably longer to act on, but send it anyway, because the alternative is us not knowing.
We don't pay for corrections and we don't publish the names of people who send them unless they ask us to.
Retire BFF is an independent publisher, not a registered investment adviser, tax adviser, or financial planner. Corrections logged here fix errors in general information about how the rules work. None of it is advice about your situation, corrected or otherwise.