4×4× proposal-result hypothesis
Under review
The 4× proposal-result claim is evidence-pending and is not a verified outcome.
No qualifying receipt is currently bound.
How this was measured
Claims & Evidence
CUT THROUGH
THE NOISE.
See what RightWork measures, what it can support, and where the evidence still needs to catch up.
No verified outcome claims are published yet.
Scanning for high-fit work
See RightWork in action
From feed noise to a better next move.
Every stage keeps the decision visible, explainable, and yours.
01
Scan
We scan more jobs, then surface the signals worth your attention.
Rank signal
02
Rank
Fit, economics, and visible risk make the shortlist easier to trust.
03
Shortlist
Keep the work with the strongest case for your time.
04
Tailored proposal
Shape a reviewable proposal around the opportunity and your proof.
Provenance
Provenance before promises.
Under reviewRightWork keeps provenance, methods, and outcome claims separate so you can inspect what is known.
Read the provenance recordThe RightWork signature formula
Client need+
Relevant proof+
Specific plan+
Clear next step+
Reviewable draft
Built to make the reasoning visible before a user approves a proposal.
The record
Claims get a fair hearing.
Large claims stay close to their status, method, and limits. Nothing becomes verified by repetition.
Evidence status changes only when its source record changes.
Every material claim
Read the claim, then the method.
Plain-language summaries stay compact. Measurement details remain one click away.
VerifiedEvidence recordUnder reviewNot yet proven
Material claim
Signature-formula proposal quality
No proposal-result lift is published.
Plain-language claim
RightWork is testing whether its signature formula produces stronger proposals.
How this was measured
- Method
- Score baseline and treatment drafts on the frozen rubric; report the paired median score difference and reviewer agreement.
- Cohort / date
- Registry review 2026-08-01. 30 paired drafts across at least three service niches.
- Baseline
- The same freelancer's baseline draft for a comparable brief, reviewed in randomized order.
- Outcome
- Paired rubric scores and reviewer agreement, not win rate or revenue.
- Limitations
- A rubric improvement is not a win-rate, client-result, satisfaction, or revenue claim. Exclusions: Materially different brief, Invented proof, Unblinded review, Missing paired draft.
Material claim
4× proposal-result hypothesis
No qualifying receipt is currently bound.
Plain-language claim
The 4× proposal-result claim is evidence-pending and is not a verified outcome.
How this was measured
- Method
- For every eligible participant, divide treatment net collected revenue by baseline net collected revenue; report the aggregate paired estimate, uncertainty, pursuit and delivery hours, Connect spend, refunds, and all zero baselines separately.
- Cohort / date
- Registry review 2026-08-01. At least 30 eligible paired participants and a lower 95% interval above 4.00 before achieved-result copy.
- Baseline
- The same participant's predeclared baseline window before the RightWork-assisted treatment window.
- Outcome
- A predeclared paired estimate with uncertainty, refunds, time, Connect spend, and zero baselines.
- Limitations
- No achieved `4X` claim from code, local tests, founder-manual work, one anecdote, or an uncalibrated pilot. Exclusions: Manual founder channel, Major pricing or profile change, Material availability or seasonality shock, Missing receipt, Mixed currency.
Material claim
Six-figure Upwork expertise
No reconciled founder evidence receipt is currently bound.
Plain-language claim
Six-figure Upwork expertise is provenance under review, not a promise of user earnings.
How this was measured
- Method
- Sum net collected earnings over the declared period after refunds; reconcile exports, screenshots, currencies, account linkage, and unexplained variance.
- Cohort / date
- Registry review 2026-08-01. One complete reconciled founder packet; a single packet never becomes a customer-outcome claim.
- Baseline
- None; this is a provenance claim rather than a product-effect comparison.
- Outcome
- A reviewed founder record, not a customer performance result.
- Limitations
- Founder earnings are optional provenance, not evidence that RightWork users will reproduce the result. Exclusions: Off-platform earnings, Unrelated account, Booked-only value, Unreconciled currency, Hidden material refund.
Material claim
Opportunity ranking and weed-out quality
Decision quality and final outcomes remain separate measurements.
Plain-language claim
RightWork helps freelancers choose better opportunities and see why.
How this was measured
- Method
- Report final won/lost outcomes and calibration by frozen score band; report pursuit efficiency separately. Do not calculate causal lift until a predeclared paired design exists.
- Cohort / date
- Registry review 2026-08-01. Five sessions for comprehension; 30 final outcomes across three score bands for calibration review; 30 paired participants before causal language.
- Baseline
- The same participant's baseline human decision workflow before RightWork-assisted review.
- Outcome
- Decision quality, evidence coverage, acquisition source, and later final outcomes.
- Limitations
- A source implementation or calibrated score distribution does not prove that RightWork causes more wins. Exclusions: Incomplete receipt, Duplicate opportunity, Test or demo data, Material strategy change.
Material claim
Jobs scanned, highlighted, and time saved
No paired time-saving result is currently published.
Plain-language claim
RightWork is testing whether guided review reduces wasted pursuit time.
How this was measured
- Method
- For each participant, calculate pursuit minutes per pursued opportunity in baseline and treatment; report the paired median difference and interval without imputing missing time.
- Cohort / date
- Registry review 2026-08-01. Five users for feasibility; at least 30 paired participants for a public numeric claim.
- Baseline
- The same participant's baseline pursuit workflow over the predeclared baseline window.
- Outcome
- Paired pursuit minutes per opportunity, reported without imputing missing time.
- Limitations
- Five-user results are directional; no public numeric time-saving claim before the minimum paired sample. Exclusions: Missing time, Outside study window, Material availability change, Duplicate opportunity.
Material claim
Earnings and value signals
Signals have no attached outcome cohort.
Plain-language claim
Observable economics help users compare opportunities; they do not prove income or ROI.
How this was measured
- Method
- Use only typed listing inputs such as budget, effective rate, client basics, competition, and visible risk wording.
- Cohort / date
- Current product surface · no approved outcome cohort
- Baseline
- No causal baseline established for earnings or value outcomes.
- Outcome
- Signal definitions and coverage can be audited separately from downstream earnings.
- Limitations
- Signals support review but do not establish replies, interviews, wins, income, or return on investment.
Material claim
Future performance claims
Claim-specific study required.
Plain-language claim
Replies, interviews, wins, income, and client-result claims each need their own evidence.
How this was measured
- Method
- Predeclare the cohort, comparison, metric, consented receipts, exclusions, uncertainty, and independent review.
- Cohort / date
- No cohort assigned
- Baseline
- A claim-specific predeclared baseline or comparator is required.
- Outcome
- Only the named metric may be reported, with its observation and maturation window.
- Limitations
- No guaranteed outcomes, automatic provider submission, or autonomous messaging.
A clearer decision
Simple pricing.
Start with clarity.
See the current plans and your account-specific terms before checkout.
See pricingYour next move
CHECK THE EVIDENCE.
CHOOSE YOUR WORK.
Review plans
For freelancers who want the reasoning before the pitch.