How We Test – Happy English UK

The six-stage testing process

Stage 1 — Shortlist (week 1)

For every category we cover (e.g. reading pens for ESL learners), we start by collecting every credible candidate from four sources: Amazon best-sellers, Reddit/Discord communities for English learners, education-supplier catalogues (e.g. World of Books, Twinkl), and direct manufacturer ranges. We then cut the shortlist to 5–8 products that are realistically purchasable in the UK, US, EU and Canada.

Stage 2 — Buy at retail

We purchase every product on the shortlist with our own funds, at retail price, from the same channel a normal reader would use (typically Amazon, sometimes the manufacturer). We do not request review units, accept manufacturer loans, or take “freebies” — this rule is absolute.

Stage 3 — Use it as a real learner would

Each product is used for at least 2 weeks of daily, real-world study by a member of our team or a panel of volunteer learners (mix of CEFR A2 → C1, ages 16 → 55). We don’t read the manual cover-to-cover and we don’t tweak settings to make the product look its best — we use it the way a busy learner actually would.

Stage 4 — Score against a fixed rubric

Each category has its own rubric. For example, reading pens are scored on:

Scan accuracy on common ESL graded readers (Penguin Readers, Oxford Bookworms) — we record what % of words it gets first try.
Pronunciation quality — judged blind by a native UK English speaker.
Dictionary depth — does it return definitions, examples, collocations?
Battery life — measured continuous-scan minutes vs. the manufacturer claim.
Build & ergonomics — weight, grip, durability after 2 weeks of daily use.
Companion app / offline mode — works without Wi-Fi? Syncs to a learning app?

Books, apps, dictionaries and flashcards each have analogous rubrics. We publish the rubric for any category on request.

Stage 5 — Cross-reference owner feedback

After our hands-on testing, we read 50+ verified reviews on Amazon, Trustpilot and the manufacturer’s site to spot long-term issues we couldn’t catch in two weeks (battery degradation after 6 months, app pulled from the store, etc.). Where verified-owner feedback contradicts our short-term experience, we say so.

Stage 6 — Write & rank

We assemble the round-up with a single clear winner per use-case (Best Overall, Best Value, Best for Beginners, etc.) and write each entry with a one-paragraph “should I skip this?” honest take. If nothing in a category is genuinely good, we say so and explain what to wait for.

What we will not do

Accept paid placements or “sponsored slots” in our best-of lists.
Hide negative findings to keep an affiliate relationship.
Republish manufacturer copy or stock images and call it a review.
Leave outdated picks at the top of an article — we re-test annually and timestamp every page.
Use AI to write a review without a human editor verifying every fact, spec and rating.

How AI fits into our workflow

We’re transparent about this: some draft outlines, comparison tables and FAQ sections on our site are generated with AI assistance to speed up structure work. However, every published recommendation is reviewed and edited by a human who has used the products, and every spec, rating and price is independently verified before publication. AI never makes the final call on which product wins.

Mistakes & corrections

If you spot an error — a wrong spec, a discontinued product still being recommended, a price that’s three months out of date — please tell us via the contact form. We respond within 5 working days and update the article with a visible “Updated on …” timestamp.

Re-testing schedule

Top 10 round-ups — re-tested every 12 months, or sooner if a major new product launches in the category.
Single-product reviews — re-checked for accuracy every 6 months.
Buyer’s guides — refreshed annually with current pricing and replacement picks.

How we test English-learning gear.