The math nobody explains: why your rating moves the way it does
A Google star rating is a rolling average of every review still attached to your profile. That's it. There's no decay curve, no time-based forgiveness built into the number itself (Google's ranking algorithm may weight recency, but the visible star average does not). So if you're sitting at 30 reviews averaging 4.9 and you take three 1-star hits in a bad month, the math does the damage instantly. Recovery is the same math running in reverse, and it's slower, because you can't remove old reviews (short of a rare Terms of Service violation), you can only add enough new ones to outweigh them.
Here's the part that matters for planning: the fewer total reviews you have, the more each new one swings the average, for better or worse. A shop with 20 reviews sees a real number move with every 5 new 5-star reviews added. A shop with 350 reviews needs 40 or 50 new 5-star reviews to move the same tenth of a point. That's not a reason to stop asking, it's the reason volume becomes armor over time.
Run the arithmetic yourself before you panic over a rating drop. Multiply your current review count by your current average, that's your total star-points. Subtract the stars lost from the bad review(s), then figure out how many 5-star reviews you need to add before dividing back out to your target average. Most contractors have never actually done this math, they just watch the number and feel worse each time it ticks down. Knowing the real denominator turns a vague dread into an actual plan with a number attached to it.
- Low review count (under 50): fastest visible recovery, most volatile to the next bad review too.
- Mid review count (50 to 200): steady, predictable movement with a consistent weekly ask.
- High review count (200+): slow to move, but a single bad review barely dents the average.
We run the actual math per client before setting a timeline. Guessing at "6 months" without knowing your denominator is how contractors end up disappointed.
The realistic recovery timeline, month by month
Assuming you start actively asking every completed job for a review (the single biggest lever, covered in depth in our guide on getting more Google reviews), here's what a typical recovery arc looks like for a contractor with a moderate review count who took a rating hit:
| Timeframe | What's happening | What you'll see |
|---|---|---|
| Weeks 1-4 | Review requests go out on every closed job, response templates go live for the bad review(s) | Rating may still look flat or dip slightly as new reviews trickle in slowly |
| Month 2-3 | Weekly review volume becomes consistent, 5-star reviews start outpacing the old damage | First visible movement in the average, often a tenth of a point |
| Month 4-6 | Volume compounds, review count itself starts working in your favor for map pack trust | Rating stabilizes near pre-incident level or better |
| Month 7-9 | Enough fresh reviews exist that the old 1-star or 2-star reviews are a small fraction of the total | Full recovery for most contractors, rating resistant to the next isolated complaint |
That's a typical arc, not a guarantee. A roofer who just closed a big storm-restoration season and has 40 jobs a month to ask reviews from will move faster than a specialty electrician doing 6 service calls a month. Job volume sets your review-request ceiling, and your review-request ceiling sets your recovery speed.
What actually speeds up recovery (and what doesn't)
Contractors waste time on tactics that feel productive but don't move the average. Here's the split.
What moves the number:
- Asking every satisfied customer, on the spot or same-day, not weeks later when the job has faded from memory.
- A simple, repeatable ask (text link, QR code on the invoice, a two-line email) beats a clunky multi-step process every time.
- Responding to the bad review publicly, calmly, with specifics. It doesn't remove the star, but it stops the review from costing you the NEXT ten customers who read it and decide you handled it like a professional.
- Consistency. Ten reviews a month for nine months beats a 90-review dump in one push that trails off.
- Training every crew lead or office scheduler to make the ask part of the closeout, not an afterthought owned solely by the owner.
What doesn't move the number, or actively backfires:
- Offering discounts or incentives in exchange for a review. Google's guidelines prohibit incentivized reviews, and profiles built on gated or paid reviews are exactly the kind that get flagged and stripped.
- Filing removal requests on every negative review. Google only removes reviews that violate its policies (fake, off-topic, hate speech, conflict of interest). A real customer's honest bad experience is staying up, and burning weeks disputing it is weeks not spent generating new ones.
- Going silent after the bad review. Silence reads as guilt to the next homeowner scrolling your profile.
- Buying reviews. Beyond the ethics, Google's spam detection catches unnatural review patterns (velocity, geography, matching phrasing) and profiles get suppressed or the reviews get wiped in bulk, taking your real reviews' momentum with them.
- Review-gating, routing happy customers to Google and unhappy ones to a private form so only positive reviews ever post publicly. It's a policy violation on top of being bad practice, and it's a different question from recovery timelines. If you're weighing whether gating is even legal, that's covered in the reviews-generation guide, not here.
The pattern across every tactic that works: it's boring, repeatable, and has nothing to do with cleverness. Recovery rewards the contractor who asks the same way, every time, for months, over the one looking for a shortcut.
How your trade changes the timeline
Recovery speed isn't uniform across the trades. It tracks job frequency and average job size, because both control how many completed jobs you have to ask reviews from in a given month.
High-frequency trades (HVAC service calls, plumbing repairs, small electrical, landscaping/lawn) close jobs daily or weekly. A consistent ask process here can generate 15 to 30 new reviews a month, which is enough volume to move a rating in as little as 60 to 90 days for a mid-sized review base. The risk on the other side of that coin: high job volume also means more chances for a bad day to become a bad review, so the same trades that recover fastest also churn through incidents fastest. The system has to run continuously, not just during a recovery push.
Project-based trades (remodeling, roofing replacement, window and siding installs, custom builds) close fewer, bigger jobs. Five or ten completed installs a month means five or ten review opportunities. Recovery here realistically runs the full 4 to 9 month range, sometimes longer if the job count is genuinely low. The upside: project-based trades tend to get longer, more detailed reviews that carry more weight with a homeowner reading them before a $30,000 decision, so quality partially offsets slower volume. A remodeler with 15 reviews that each run three paragraphs and mention specific rooms, timelines, and crew names reads as more credible to a homeowner than a competitor with 60 one-line reviews, even at a slightly lower star average. We see this play out in our reputation work with remodelers specifically: depth beats speed when the purchase decision itself is slow and high-stakes.
Seasonal trades (roofing after storm season, snow removal, pool builders) see recovery cluster around their busy months. A roofer who took a hit in the off-season may sit flat for two or three months, then recover fast once storm season restarts the job pipeline and the review requests with it. Planning the recovery push to line up with your busy season, not fighting the calendar, is part of a real strategy here, not an accident. If your bad review lands in your slow season, don't panic about the flat months, use them to get the response templates and the ask process built so you're ready to move the instant volume returns.
Whatever your trade, the lever is the same: more completed jobs asked for a review, asked consistently, asked close to job completion. Window and siding companies in particular tend to underuse this, treating the install completion as the end of the customer relationship instead of the best possible moment to ask.
When a single bad review is doing outsized damage
Sometimes the problem isn't the average, it's that one specific 1-star review sits at the top of your profile (Google's default sort favors relevance, which often surfaces detailed reviews, positive or negative) and it's the first thing every homeowner reads before they call anyone else. In that case the fix isn't purely mathematical, it's about what surrounds it.
Respond to it. A calm, factual, non-defensive public response, ideally within a day or two of it posting, does two things: it shows the next 50 people reading your profile that you handled a bad outcome like a professional, and it sometimes prompts the original reviewer to update or remove their review once the issue is actually resolved. That happens more than people expect, but it's not guaranteed and shouldn't be the plan. What kills a response is defensiveness, arguing the facts in public, or a response that reads like it was written by a lawyer instead of the owner. Name what happened, name what you did about it, invite them to call you directly. That's the whole formula.
Get more reviews around it, fast. A single bad review sitting alone at the top of a 12-review profile is a crater. The same review sitting in position four of a profile that gained eight new 5-star reviews in the following month reads as an outlier, which is what it probably was. Volume doesn't erase the review, it recontextualizes it, and recontextualizing is usually all a homeowner needs to keep reading instead of closing the tab.
Check if it's fake or fraudulent. If the review comes from someone who was never a customer, references a job you never did, or is clearly a competitor or personal dispute unrelated to the work, it may qualify for removal under Google's policies. That's a real path, but it's the exception, not the default plan, and it should never be the first thing you try before also asking for genuine new reviews. Flag it through Google's official reporting process, document why it violates policy specifically, and keep asking real customers for reviews while that request sits in queue. Removal decisions can take weeks, your recovery timeline shouldn't wait on them.
One more thing worth saying plainly: not every bad review deserves a fight. If a crew genuinely missed a deadline, left a mess, or underdelivered, the honest move is to own it in the response and fix the underlying operational issue so it doesn't repeat. A star rating that recovers because the work got better is a real recovery. A star rating that recovers because you buried one bad review under volume while the same mistake keeps happening is a countdown to the next one.
Why we treat this as a system, not a fire drill
Contractors call us after the bad review already happened, wanting the fastest fix. The fastest fix is honest: there isn't one, because the math doesn't allow it. What we build instead is the system that makes the NEXT bad review a rounding error instead of a crisis: a standing review-request habit tied to every closed job, response templates ready before you need them, monitoring so you know about a new 1-star within hours instead of finding it three weeks later when a lead mentions it, and schema markup so your review count and rating show up correctly to both Google and the AI answer engines now quoting local businesses by reputation.
That last part matters more every month. When a homeowner asks an AI assistant "who's the best roofer near me," the engine is pulling from review volume, review recency, and review content, not just the star number. A contractor with a thin, stale review profile is invisible to that query even with a decent average. Reputation isn't a one-time repair job, it's infrastructure, same as your website or your map pack listing. Fix the immediate math, then leave the system running so you're never back here explaining a dip to a new customer.
The contractors who ask us this question twice are the ones who treated the first recovery as a one-time push: three weeks of hard asking, the number climbs back, everyone stops. Six months later a bad job happens again, the review lands, and there's no system left running to absorb it. The ones who never ask us this question again are the ones who kept the request habit going after the crisis passed, so the rating stayed high enough that one bad review barely registered. Same math, different discipline.