How we score sweeteners.
Every sweetener on SweetSpot is scored 0–100 against ten evidence-weighted metrics. The methodology is open. We publish what we measure, where the evidence comes from, and how we weight it. If we are wrong we want you to be able to tell us, with citations.
Three principles
Before any of the metrics, three commitments shape the work:
No advertising. No affiliate revenue. No brand sponsorships. Funded by user subscriptions and nothing else. Brands cannot influence scores or pay for placement.
Sources are tiered. Randomised controlled trials and meta-analyses outrank observational studies; observational outranks animal models; animal outranks mechanistic in vitro work. We tell you which tier any claim is resting on.
Every published score has a public list of the studies it relies on. When the evidence shifts — as it did for erythritol in 2023 and aspartame in 2023 — we update the score and document the change.
The ten metrics
Each sweetener is scored on ten dimensions. Five are surfaced on the free tier; the full ten power the SweetSpot Premium scan and detail views.
Closeness to clean sucrose taste. Penalises bitter, metallic and liquorice tails.
Postprandial glucose and insulin response. Includes fructose-load adjustment for liver effects beyond GI.
Source and processing intensity. Whole-food and minimally-processed score higher; the score is informational, not a safety claim.
Cariogenicity. Non-fermentable sweeteners and xylitol's anti-caries effect score highest.
Long-term health evidence aggregate: cancer signals, cardiovascular, metabolic disease, microbiome.
GI tolerance at typical doses. Sugar alcohols and high-FODMAP sweeteners score lower.
Microbial composition and SCFA production. Newer literature, lower confidence — flagged transparently.
Separated from taste quality so you can see when a sweetener is sweet but lingers wrong.
Land, water and processing footprint per 100g sucrose-equivalent sweetness.
Population subsets with absolute contraindications (PKU, dog households for xylitol, etc.).
Evidence tiers
Not all studies are equal. Where conflicting findings exist, the higher-tier source determines the call.
- AMeta-analyses & systematic reviewsCochrane, Annual Reviews, peer-reviewed systematic reviews registered on PROSPERO.
- BRandomised controlled trialsPre-registered, in humans, with relevant duration and dose.
- CObservational human studiesProspective cohort > case-control. Adjusted for confounders. Replicated where possible.
- DAnimal & mechanistic studiesUsed to support, not establish, claims. Translational caveats noted.
- ERegulatory positionsFDA, EFSA, WHO/IARC, JECFA. Cited as positions, not as evidence on their own.
Conflict-of-interest policy
SweetSpot is funded entirely by user subscriptions. We do not accept payments from food companies, ingredient suppliers, or trade associations. We do not run advertising. We do not use affiliate links.
When we cite a study with industry funding, we say so. When we cite a study with a known critical commentary, we link the commentary too. Where a finding is preliminary — Suez 2022 on the microbiome, Witkowski 2023 on erythritol — we score the sweetener as if the finding may matter, and we tell you so.
If we are wrong about a sweetener, we want to be told. Send corrections with citations to support. We will publish substantive updates with a dated changelog on each sweetener page.