CityvettedWhy we weight recent reviews more than old ones
A 4.9 from 2019 describes a company that may not exist anymore. How recency-weighting keeps our rankings honest.
A business is not a fixed object. It is a group of people, and people leave. The lead electrician who earned a wall of glowing reviews in 2021 may have started his own company in 2024, taking his judgment and his careful habits with him. The reviews he generated are still sitting there, still counting toward a star average, still telling you about a version of the company that no longer exists. This is the core reason we weight recent reviews more heavily than old ones, and it is one of the most consequential choices in how we rank.
- Businesses change: staff turns over, owners sell, standards drift. Old reviews describe a company that may be gone.
- We give reviews from the last 18 months full weight and taper older ones, so the score reflects who you would actually hire today.
- We never delete old reviews. We just stop letting 2019 outvote last month.
Why an average lies to you
Imagine two roofing companies, both sitting at a 4.4 star average. The first earned its rating steadily, with strong reviews arriving every month right up to last week. The second earned a flood of praise four years ago and has collected mostly two-star complaints since. A plain average treats them as equals. They are not equals. One is on the way up and one is on the way down, and the only signal that separates them is when the reviews landed.
Recency weighting is how we read that timeline. A review from last month counts close to full. A review from three years ago still counts, but less, because it is older evidence about a moving target. The result is a score that tracks the present rather than embalming the past.
Cityvetted methodology, 2026.
The honest tradeoff
There is a real cost to this approach, and I would rather name it than hide it. Recency weighting can be unfair to a great company having one rough quarter. A burst pipe of bad luck, a single difficult job that went public, a short-staffed summer: any of these can dent a recent window. We try to soften that with consistency checks, looking at whether a dip is a blip or a trend. A company with five strong years and one shaky month reads very differently from one in steady decline, and our score is built to tell those apart.
"We are not ranking who was good. We are ranking who is good, and those are not always the same business."
to Daniel Ruiz, Senior EditorHow to read recency yourself
You do not need our math to use this idea. When you are looking at any local business, the single most useful move is to sort the reviews by newest first and read the last ten. It takes four minutes and it tells you more than the star average ever will.
- Sort newest first.
Ignore the headline average for a moment. You want the most recent voices.
- Read the last ten reviews.
Look for a pattern, not a single bad day. One angry outlier is noise; three in a row is a signal.
- Check the gaps.
If the newest review is from fourteen months ago, the business may be winding down or may have stopped serving your area.
- Watch the owner replies.
Recent, specific, calm responses tell you the company is still paying attention.
That last point matters more than it looks. A business that answered complaints thoughtfully in 2020 and goes silent in 2026 has usually changed inside, even if the star average has not caught up yet. The reviews are a lagging indicator. Recency weighting is our attempt to make the score a little less late.
We tuned the fade curve again this year after readers flagged a few businesses that had clearly turned around but were still being dragged by old complaints. If you spot one, tell us. We would rather be corrected than confident and wrong.