BetterRank
How BetterScore is calculated
One number, per product, per job, from 0 to 100. Published weights, published confidence rules, and a stated reason behind every movement. If you cannot argue with a ranking, it is not a ranking — it is an advert.
The seven signals
These weights apply identically to every product in every job.
Verified rating, recommendation rate, sentiment
Resolved rate, response speed, user confirmation
Verified users, depth, anomaly checks
Primary-use votes, workflow match
Helpful replies, Q&A, founder participation
Recent improvements, sentiment trend, release activity
Follows, saves, likes, qualified profile visits
Ranking behaviour
The rules that stop volume and popularity from deciding everything.
Confidence-adjusted, not averaged
A raw average with 8 reviews is noise. Scores are damped toward the category median until there is enough evidence.
Recent evidence counts more
A problem resolved last month matters more than one resolved two years ago — but the history stays visible.
Volume raises confidence, not rank
More reviews let your quality signals count for more. They do not directly add points.
Different rank per job
A product can be #1 in one job and #6 in another. There is no single universal ranking.
Every movement is explainable
Each rank change records the signal that moved and the reason, and that reason is public.
Paid placement is inert
BetterBid spend contributes exactly zero points. The two systems share no inputs.
Worked example
Why a smaller product can outrank a more popular incumbent.
Product B wins on satisfaction, review trust, community and popularity — and still loses. Product A resolves problems 23 points better and fits the job 7 points better, and those two signals carry 35% of the weight between them while popularity carries 5%. That is the intended outcome: support quality and use-case fit should outweigh raw audience size.
Confidence
Sparse data creates fake precision. We label it instead of hiding it.
Why ranks moved recently
Live examples from the last few weeks.
Postloop #4 → #1 in Social Media Scheduling
Two recurring publishing problems confirmed resolved and eleven new verified reviews
8d ago
Tidewave #6 → #4 in Social Media Scheduling
Momentum signal up after three releases in six weeks
12d ago
Emberly #7 → #3 in Email Marketing
Verified review volume tripled after a support-led invite campaign
14d ago
Excluded from BetterScore entirely
- BetterBid spend, and any sponsored placement performance
- Subscription plan or amount paid to BetterSaaS
- Affiliate or referral relationships
- Reviews identified as vendor-authored or incentivised
- Reviews removed for manipulation, including their historical effect
- Editorial coverage or newsletter placement
How weightings change
Weight changes are recorded in an internal audit log with an actor, a timestamp and a rationale, and material changes are announced before they take effect. The most recent change raised problem resolution from 18% to 20% on 1 August 2026.
Ranking and moderation decisions leave an audit trail so a disputed change can be reconstructed rather than argued about.
Read the rules and appeals process →