Standard PW-METHOD · v1.2
This is the rule the scores are held to — written down, numbered, and citable by clause. Seven principles set the method; the tiers, the score bands, the published category weights and the correction loop are the reference apparatus that proves it. Every figure is a dated reading, never a live quote.
A PriceWorld score is one number from 1.0 to 5.0 — higher means cheaper all-in once renewals are counted. It is a rubric output, not buying advice. See the full bands ↓
Part 1 · The principles
The backbone of the standard. Each clause is a rule the scoring must obey; the reference sections that follow are filed against these clauses, so you can cite a figure back to the principle it answers to.
No opinion, no "feel". If we cannot show the figure and link the source it came from, it does not go on the record.
A reading (the raw figure + source + date) feeds a published rubric (a fixed formula that maps it to a 1–5 score), which feeds a weighted composite (published weights → one rating to one decimal).
See the evidence tiers ↓Each dimension is bucketed to a whole number 1–5. A 3.7 on a single dimension would imply accuracy we do not have; honest buckets beat invented decimals.
The weight tables are on this page. Changing a weight is a new version and a re-score of everything it touches — there are no quiet edits to a published score.
See the published weights ↓Across every category, the majority of weight sits on what you actually pay over time — entry, renewal, and total cost of ownership — because that is where the surprises hide.
See the published weights ↓Vendor prices change without notice. A figure carries the date it was read and the caveats that bound it; it is never presented as monitored or guaranteed-current.
See what a score does not say ↓Scores come from the rubric — there is no manual override. Where a link earns a commission we say so, and a low-scoring tool still gets its real figure and its real score.
See the correction loop ↓Part 2 · Evidence tiers Reference for Clause 2
Every figure carries a tier that says how much weight it can bear. Only a source-backed reading is eligible to be a headline figure; modeled fills a gap and is never the headline; needs-review is held and cannot move the record on its own.
| Tier | Headline eligibility | What it means | Standing |
|---|---|---|---|
| Source-backed | Headline-eligible | Read directly from the provider and reviewed onto the record — dated, with caveats attached. e.g. A renewal rate copied from the cart total on the day it was read. | On file |
| Modeled | Never the headline | A documented estimate where no direct reading exists yet — shown as modeled, never as fact. e.g. A 3-year cost projected from a published renewal schedule. | Modeled |
| Needs review | Held, not applied | Reported, disputed, or past its freshness window — opened as a pending item that cannot move the record on its own. e.g. A reader writes in that a price rose — logged, not yet published. | Pending |
Part 3 · Score bands Reference for Clause 3
The 1.0–5.0 composite maps to five bands. These are rubric outputs — what the math says, not what you should buy. The palette has no green by design: sound standing reads as filing-blue, an adverse standing reads as the oxblood record stamp, each carrying a word so the meaning never rests on colour alone.
| Range | Label | Standing | What the math says |
|---|---|---|---|
| 4.5 – 5.0 | Excellent value | Sound | Best-in-class once renewals and total cost are counted. |
| 3.5 – 4.4 | Good value | Sound | Above average on total cost of ownership. |
| 2.5 – 3.4 | Average | Neutral | Industry standard — no major surprises and no real savings. |
| 1.5 – 2.4 | Below average | Adverse | Notable hidden costs or weak value for the money. |
| 1.0 – 1.4 | Poor value | Adverse | Significant cost surprises or lock-in risk. |
Part 4 · Category weights Reference for Clauses 4 & 5
The weights are on this page, versioned with the standard. Across every category the majority of weight sits on what you actually pay over time — entry, renewal and total cost of ownership — because that is where the surprises hide. The leading weight in each table is marked.
Shared hosting, single site, most-promoted starter plan.
Renewal multiplier and 3-year TCO carry 40% between them — hosting is where the promo-to-renewal jump bites hardest.
.com TLD at standard (non-promotional) registration price.
Renewal price and renewal markup carry 45% — a $0.99 first year that renews at $18.99 is the headline, not the promo.
1,000 subscribers on the lowest paid plan with automation.
Entry cost, price stability, hidden-cost exposure and scaling cost lead — the bill rarely stays where it starts.
$2.95/mo intro $11.99/mo renews +306% step
Illustrative, not a current price. A shared-hosting plan advertised at a promo rate that renews far higher is exactly the cost the weights above are built to surface — which is why renewal and total cost carry the most weight. Confirm any real price with the provider.
Illustrative exampleBars are drawn relative to the largest weight in each table — they compare weights within a category, not across them. The full weight set lives in the project's scoring utility; these are the published values, read at build time.
Part 5 · Correction loop Reference for Clause 7
A correction is a chain of custody. Anyone can file one with evidence; it is held for review; it closes only when a source-backed reading confirms it. The published figure does not move in between.
A vendor or a reader submits a figure backed by something we can check — a cart total, a price page, a screenshot. This opens a pending item against the record.
FiledThe claim is weighed against the source. While it is open it is marked needs-review; it does not change the published figure. We aim to review within 48 hours.
Under reviewIf a source-backed reading confirms the change, the figure is updated, re-dated, and the version log records it. If it cannot be confirmed, the record stands and the item is closed with a note.
Closed on fileThe rule: A report is weighed, not applied. A claim can never move the canonical figure on its own — only a source-backed reading closes a correction.
Part 6 · Limits Reference for Clause 6
The standard is as much about its limits as its claims. A figure is a dated reading — these are the things it deliberately does not promise.
Weights, the dimension and coverage counts, and the rules-updated date are real values derived from the project's scoring utility; tier, band and correction copy is documented method copy, not invented for this page. Nothing here is presented as live, verified-now, monitored, or guaranteed-current. Confirm the current price with the provider before you buy.