Recorded data
Exact model, observed purchase price, supply price and yield, product system, source or research-pack reference, observation date where stored, and confidence or verification status where available.
Methodology and data policy
See exactly how
the estimates are built.
Real Receipt separates recorded facts, user inputs and modelling assumptions. Unsupported costs stay excluded rather than being filled with confident-looking guesses.
Included costs are visible Assumptions are labelled Rankings are not sponsored
Last reviewed: 2 August 2026
87printer models in the current dataset
40coffee machines in the current dataset
0paid placements in the rankings
Core policy
Facts, inputs and assumptions are different things.
A useful comparison must show which numbers came from research, which came from the user and which are modelling choices.
User inputs
Print volume, number of printers, ownership period, cups per day, bean price, milk habits and other values entered directly into a calculator.
Modelled assumptions
Values such as coffee-machine energy per cup or care-product replacement cadence. These are explicitly labelled and should be tested with the user's own assumptions.
Printer methodology
Purchase price plus routine replacement supplies.
Printer rankings currently use only cost components with a consistent production rule across the dataset. That narrower scope is deliberate.
- 01Purchase cost
Recorded exact-model purchase price multiplied by the selected quantity.
- 02Routine supplies
Printed pages are divided by recorded supply yield and multiplied by replacement price.
- 03Colour-page treatment
Colour pages include black usage plus the complete cyan, magenta and yellow set, or one matching tri-colour cartridge.
Printer formula
Included total = purchase cost + black-supply cost + colour-supply cost.
Cost per printed page divides that included total by all pages in the selected ownership period. It therefore includes the printer purchase as well as supplies.
Suitability and ranking
The recommended view places models with a verified recommended monthly volume that covers the user's workload before models with unknown or exceeded limits. Within the same suitability status, the lower included total ranks first.
The “Best for” groups use the recorded upper recommended monthly volume: home and light use up to 500 pages, home office 501–1,500, small business 1,501–5,000, high-volume business above 5,000, plus a separate unknown-volume group.
Printer costs currently excluded
- Energy and standby electricity
- Paper
- Starter cartridges, toner or bundled ink
- Imaging drums, printheads and maintenance kits without a verified replacement rule
- Repairs, maintenance, finance and resale
These costs remain excluded until the exact-model part, price, replacement rule, source and observation date can be applied consistently.
Coffee-machine methodology
Machine, coffee, milk, modelled energy and care products.
Coffee-machine results combine recorded product data with editable user inputs and clearly labelled modelling assumptions.
- 01Coffee or pods
Bean and ground-coffee machines use dose × user-entered coffee price. Pod machines use a recorded per-drink benchmark.
- 02External milk
Added only for machines whose model requires separately purchased milk under the stored research rule.
- 03Energy and care
Energy per cup and some care-product replacement cadences are modelling assumptions and are labelled as such.
Coffee-machine formula
Estimated total = machine price + coffee or pods + external milk + modelled energy + care products.
Cost per cup divides the estimated total by the number of cups entered over the selected ownership period.
What the confidence labels mean
Confidence describes the completeness of the stored research record, not a promise about product quality, reliability or future price. A high-confidence record can still contain openly labelled modelling assumptions.
What the comparison cannot establish
- Coffee taste, extraction quality or milk texture
- Long-term reliability or repair probability
- Live stock, discounts or future capsule and bean prices
- Whether a specific machine suits a user's preferred workflow
Sources and freshness
Exact-model evidence comes before broad estimates.
Prices and specifications can change. Every result should be treated as a dated comparison snapshot.
Source hierarchy
- Official exact-model manufacturer specifications and accessory pages
- Official brand stores and exact-model product pages
- Major UK retailers with an exact model match and visible price
- Specialist retailers where official availability is incomplete
- Clearly labelled modelling assumptions only when a measured value is unavailable
Current limitation: printer provenance display
Coffee-machine records already store an official URL, research date and confidence field. Printer data has import audits and dated research packs, but full source URL, observation timestamp and confidence are not yet exposed for every numeric field in every result card.
Until that migration is complete, printer prices should be treated as research snapshots and verified again before purchase.
Price selection
New imports use the lowest clearly visible UK price in the research evidence for the exact model when availability and model identity can be confirmed. Delivery, membership pricing, subscriptions and temporary bundles are excluded unless the research record states otherwise.
AI and privacy
The AI explains stored results; it does not create product facts.
The printer comparison chat is optional. The coffee calculator currently has no AI chat.
Server-rebuilt figures
The printer chat sends the current scenario and selected filters. The server rebuilds the comparison from the internal dataset instead of trusting browser-supplied totals.
Strict limits
The AI is instructed to use only listed results and to refuse unsupported claims about quality, reliability, availability or specifications not present in the data.
Calculator privacy
Normal calculator changes run in the browser. Printer inputs and messages are sent only when the optional comparison chat is used.
Corrections policy
Conflicts stay visible until they are resolved.
Real Receipt should become more accurate as the dataset grows, not merely larger.
Suspected errors, stale prices and source conflicts should be reported to hello@realreceipt.co.uk. A record may be corrected, marked unknown or removed from ranking when the evidence is incomplete.
We do not silently replace missing data with an industry average. Material calculation or import changes should be documented in the repository audit records.
Use the assumptions that match you