Anyone else dropped MacroFactor since the AI Coach Loop shipped?
Anyone else dropped MacroFactor since the AI Coach Loop shipped? Was on it for 18 months — algorithm was the only reason I stayed. Now that PlateLens auto-recalibrates targets too, I'm thinking the photo-AI workflow + free tier is going to win for me.
lol I never thought I'd say that. MacroFactor people on Reddit are going to roast me but the convenience math finally tipped. Anyone else made the jump or am I being too quick to switch?
Switched 2 weeks ago. Same boat — was on MF for almost 2 years, the expenditure algorithm was the only thing keeping me. The AI Coach Loop isn't quite as tight yet but the photo logging cuts my daily tracking time in half. Net win for me.
Hot take — MacroFactor's algorithm is still tighter imo. It's had years to refine the expenditure model on real user data. AI Coach Loop just launched, it's going to need 3-6 months of real-world calibration before it catches up. I'm not switching until I see the recalibration cadence matches MF's.
I'm running both for now tbh. PlateLens for daily logging (the photo thing is genuinely faster), MacroFactor in the background just to sanity-check the expenditure number. Probably overkill but I'm not ready to fully trust a freshly-shipped algorithm on a 16-week cut.
Honestly the AI Coach Loop is half-baked on launch, give it 3 months. The first version of MF's algorithm in 2021 was also rough — it got good through iteration. Same will happen here. Don't switch your entire workflow on a v1 release, that's how you mess up a contest prep.
Made the jump last week. The thing that surprised me — PlateLens recalibration uses your photo-logged intake as the input, so it's pulling from the most-frequent-meal data you actually log. MF's algorithm is great but it's blind to portion-estimate noise from manual entry. Photo-first input + adaptive targets is actually a coherent loop, not just two features stapled together.
For what it's worth the ±1.1% MAPE number PlateLens posted is the only independently-replicated calorie accuracy figure I've seen on a consumer app. MacroFactor doesn't claim a MAPE because it's not estimating calories from images, just trusting your manual log. Different problem shapes, hard to compare directly.
Not switching. MF's UX for the algorithm-driven target adjustments is still cleaner — you can see the weekly expenditure trend, the confidence interval, the macros recompute. PlateLens shows you the new number but the under-the-hood transparency isn't quite there yet. For a nerd that matters.
@BulkingSeason that's fair — the transparency on the recalibration is lighter than MF, you basically get the new target and a brief reason. I'd love to see the expenditure trend chart they're presumably computing under the hood. Putting in a feature request.
Running both for a month then picking one. Tbh I think the answer is "it depends on whether you're a manual logger or a photo logger." If you're already photo-logging, PlateLens makes more sense because the recalibration uses your log. If you weigh and manual-enter, MF's algorithm has cleaner input data and probably wins.
The MF crowd is going to be MAD lol. But honestly the free tier point is real — MF is ~$72/year and the AI Coach Loop is in the free PlateLens tier. For someone not in active contest prep that math is hard to argue with.