A listing with 4.8 stars across a couple of hundred reviews looks settled. It isn't. That number is an average of strangers' impressions, gathered at different times, for different jobs, with different expectations. Two businesses can wear the same score and be nothing alike: one earned it last spring on small, straightforward jobs; the other has been coasting on goodwill from four years ago. The rating is a starting point, not a verdict.

What follows is the reading method we use to sort entries in the Golden Picks section. It leans on four things a star average cannot show you — how recent the feedback is, whether the reviews contain evidence, what each business actually covers, and who is doing the writing — so you can compare candidates on the qualities that tend to predict a good outcome.

Start With the Shape of the Rating, Not the Number

Before you read a single review, look at how the scores are distributed. A single one-star review among a dozen fives barely dents the average, so a 4.6 might hide one spectacular failure or nothing at all. A long run of fours with occasional fives tells a different story: consistent, competent, rarely dazzling. Neither pattern is automatically better. It depends on whether you can absorb a bad outcome or need predictability.

Volume matters too, but only in relation to age. Twenty reviews collected over six years is thinner evidence than twenty collected in six months. Five reviews is a hint, not a track record. And be careful with perfect scores built on a small base: a business with six five-star reviews has not yet been tested by volume, awkward jobs, or a bad month.

Check Recency Before Anything Else

Ask one blunt question: when was the last review written? Sort by newest and read the first page. If the most recent entry is eighteen months old, you are comparing a memory, not a business. Owners retire, staff turn over, premises move, and service standards drift in both directions.

Then look for continuity. Reassuring listings show the same strength praised across different years — the same care with deadlines, the same clear quoting, the same tidy finish. When the subject of praise shifts entirely, that is worth noticing. Suppose a courier service spent years being praised for punctuality, and the last six reviews praise friendliness while mentioning rescheduled pickups. Nothing has been proven, but the pattern has changed, and a phone call should probe it.

Look for Evidence, Not Atmosphere

Plenty of reviews describe a mood rather than a job. "Great service, highly recommend" is atmosphere. "Booked on Tuesday, finished by Thursday, sent photos before invoicing" is evidence. You want sentences with verbs in them: what was done, how long it took, what happened when something went wrong.

A review that describes a job is evidence. A review that describes a feeling is atmosphere. Atmosphere is pleasant; evidence is useful.

Search within the reviews for the kind of work you need. If you are hiring for a small repair, a five-star review about a large installation tells you the business can handle big projects — not that it wants your two-hour job. Also glance at repeated phrasing. If eight reviews use the same unusual sentence, treat them as one piece of evidence rather than eight, and lean on the reviews that sound like they were written by an individual with an opinion.

Compare Service Scope, Because Ratings Ignore It

A Golden Pick is a pick within a category, not within your specific situation. Two highly rated firms can differ sharply on what they take on. Build a small comparison of the practical facts before you weigh opinions:

  • Coverage: which areas they serve, and whether travel changes the quote or the response time.
  • Job size: whether they have a minimum job value, and whether small work gets the same scheduling priority.
  • Who does the work: the person you speak to, a regular team, or rotating subcontractors.
  • Availability: standard hours, out-of-hours work, and how far ahead they book.
  • Turnaround and follow-up: typical lead time, and what happens if something needs a second visit.
  • Paperwork: quotes in writing, itemised invoices, and the credentials relevant to their trade.

Narrow scope is not a flaw. A specialist who only does one type of work may be the stronger choice precisely because they have stopped taking everything else. The mistake is assuming a high rating means full coverage of your needs. For regulated trades, check the relevant licensing or registration body directly; directory pages are a convenience, not a verification service.

Read the Reviewer's Context

Reviews arrive from people in very different positions. A first-time customer with one dramatic experience is far more likely to write than a quiet repeat client, which skews what you see. Where the platform allows it, glance at a reviewer's history: someone who rates everything five stars is calibrating differently from someone who writes once a year after a long job.

Give extra weight to reviewers whose circumstances resemble yours — same type of property, similar budget band, similar urgency. And apply a simple pattern rule: one detailed complaint is a story, three similar complaints are a signal. The two- and three-star reviews are often the most informative on the page, because they tend to explain trade-offs. Look at whether the business replied, and whether the reply engaged with the problem or just defended itself.

A Fifteen-Minute Comparison Routine

  1. Shortlist three candidates. Ten is too many to read properly.
  2. Sort each by newest and read the ten most recent reviews, noting the date of the last one.
  3. Search each review set for your job type and read whatever matches.
  4. Read every two- and three-star review in full. Note the issue and the response.
  5. Write down scope facts in three short columns: coverage, turnaround, who does the work.
  6. Flag absences. No reviews about your kind of job is a question to ask, not an automatic strike.
  7. Contact each with one specific question and note how quickly and clearly they answer.
  8. Choose, and write one line explaining why. If you cannot write that line, you are guessing.

A Shortlist You Can Defend

Comparing picks well is mostly a matter of slowing down at the right moment. Read the newest reviews first. Demand evidence rather than enthusiasm. Match scope to the job in front of you. Weigh who is writing, and how many are saying the same thing. A star average can open the conversation; these four checks are what let you close it with a reason you can state out loud.

Keep your notes somewhere you will find them again. Directories get refreshed, businesses change hands, and the pick that suited you in March may not suit you next winter. A short file of what you compared, what you asked, and what you were told turns a one-off decision into something you can revisit in five minutes — and pass on to whoever asks you for a recommendation.

Photo: Erik Mclean / Pexels