Ask

How long should I give something before deciding it does nothing? Magnesium and fish oil, six weeks in

Ran a proper stop trial on magnesium after four months because I could not tell whether it was doing anything. Kept a one line sleep note every morning for two weeks on, two weeks off, two weeks on, without telling my partner which phase I was in so I could ask her whether I had been snoring or restless. Result was a small but consistent difference in how long it took me to fall asleep and no difference in anything else. That was worth knowing, and it cost me six weeks and a notebook rather than another year of maybes.

96 · in/supplements ·

Is it normal to be starving all afternoon if I eat protein mostly at dinner

Also worth checking the boring possibilities before the clever ones. Is the afternoon hunger real hunger or is it boredom, stress or dehydration at a predictable time of day? Mine turned out to be 3pm at a desk with nothing interesting happening, and a glass of water and a ten minute walk handled a decent chunk of it.

276 · in/macros ·

Sign up bonus never posted eight weeks after I hit the spend

Use the secure message centre rather than the phone for the first contact. You get a written record with a timestamp, which matters if you have to escalate later, and the person answering messages is usually looking at the same screens as the phone agent. If the answer is unsatisfactory, then call.

241 · in/points-and-miles ·

900 companies at $200/mo or 40k solo users at $9, which do i probe first

900 at $200, and the reason is that 900 is a countable number.

You can build a list of all 900 in a week. You can email every one of them personally over a quarter. You need 12 to clear $2,400 MRR, and 12 is a number you can picture. The whole business fits in a spreadsheet you can actually read.

The 40,000 at $9 is not a market you can reach with 10 hours a week - it is a market you reach with content, SEO, ads and a funnel, all of which are full-time jobs and none of which you are currently doing. You would need 270 customers for the same revenue, at a price point that supports zero human contact.

The honest cost of the narrow one: sales cycles in regulated niches are long, procurement is real, and someone will ask for a security questionnaire. If that sounds worse to you than writing blog posts for two years, that preference is legitimate data.

39 · in/before-you-code ·