App Store ratings: why 4.5 stars isn't good enough anymore

How we got here

Rating inflation happened gradually. Apple introduced SKStoreReviewController in iOS 10.3, which made it trivially easy to prompt for a review without leaving the app. Before that, reviewing an app meant going to the App Store, finding the app, scrolling down, tapping the stars, and writing something. Only frustrated people bothered. That selection bias kept ratings artificially low.

Once in-app prompts became standard, happy users started leaving ratings too. The average went up across the board. Then Apple added a "rate this app" shortcut after deleting an app, which added more casual ratings. Google Play followed a similar pattern with their in-app review API.

The result: if your competitors are all 4.7+, your 4.5 doesn't communicate "good app, minor issues." It communicates "noticeably worse than the alternatives." Context matters more than absolute numbers, and the context has shifted.

Ratings affect two things: ranking and trust

Apple's search algorithm weighs several factors, and ratings are one of them. The exact weight isn't public, but from what I've seen (and from what most ASO practitioners agree on), rating score and rating volume both matter. Between two apps targeting the same keyword, the one with a higher rating and more reviews tends to rank higher, all else being equal.

But ranking is only half of it. The bigger effect is on conversion. When someone searches for "habit tracker" and sees five results, they're scanning icons, titles, and star ratings in about two seconds. A 4.3 next to three 4.7s gets skipped. The user doesn't even tap through to your listing.

I've seen numbers suggesting that the difference between 4.0 and 4.5 stars can mean a 30-50% drop in conversion rate. Between 4.5 and 4.7, the effect is smaller but still measurable: roughly 10-20%. Those numbers vary by category and market, but the direction is consistent.

What rating you actually need

The answer depends on your category. In some categories (Games, Social Networking), ratings are all over the place. You'll see popular apps at 4.2 doing fine because users expect more divisive experiences. In Productivity or Health & Fitness, 4.5 is the minimum for being taken seriously, and 4.7+ is where the top apps sit.

A practical way to find your target: search for your main keyword on the App Store. Look at the ratings of the top 10 results. Your rating needs to be within 0.2 points of the median, or you're at a disadvantage. If everyone above you has 4.7 or 4.8, aim for 4.6 at minimum.

Interestingly, 4.9 and 5.0 can actually look suspicious. If an app has a 5.0 rating with 47 reviews, some users assume the reviews are fake. The sweet spot is 4.6-4.8 with a healthy review count. It communicates "real app used by real people who mostly like it."

Volume matters as much as the number

A 4.8 rating with 12 reviews means something very different than a 4.8 with 12,000 reviews. Users know this intuitively. App Store algorithms know this explicitly.

For search ranking purposes, review volume acts as a confidence signal. Apple and Google can't rely heavily on a rating from a tiny sample size, so apps with more reviews get more weight in search results.

For users, volume is a proxy for popularity. An app with 50,000 ratings feels "safe" in a way that an app with 50 ratings doesn't, even if the actual score is the same. You can't manufacture volume overnight, but you can stop accidentally suppressing it.

One thing to watch: Apple lets you reset your rating when you ship a new version. This sounds tempting if you've fixed major issues, but you lose all your accumulated volume. I'd only reset if your current rating is below 3.5 and you've genuinely fixed the problems. Otherwise, the volume loss hurts more than the score improvement helps.

When to ask for a rating (and when to shut up about it)

The timing of your review prompt is the single biggest lever you have over your rating. Get it right and your average goes up. Get it wrong and you're collecting 1-star reviews from people you interrupted at the worst possible moment.

The principle is simple: prompt people when they're happy. The execution is harder because "happy" looks different in every app. Here are patterns that work across categories:

After a completion event. The user just finished a workout, completed a task, reached a goal, exported something successfully. They feel accomplished. A review prompt at this moment catches positive momentum.

After repeated use. Someone who's opened your app 10 times is more likely to rate it well than someone on their second session. They've already decided they like it enough to keep coming back. Wait at least a week and several sessions before prompting.

After a "wow" moment. If your app has a feature that tends to delight people (an insight, a visualization, a transformation), prompt after they've seen it. You want the review to reflect their best experience with your app, not their average one.

Never prompt:

• Right after a crash, error, or failed action. This seems obvious, but I've seen apps that prompt on every Nth session regardless of what happened in that session.
• During onboarding. The user doesn't know if they like your app yet. You're asking them to evaluate something they haven't used.
• Immediately after a paywall dismissal. They just said no to paying. They're not in a generous mood.
• When they're in the middle of something. If the user is typing, editing, or mid-workflow, a popup is an interruption, and interruptions breed resentment.

What Apple and Google actually let you do

Apple restricts how you can ask for reviews. You must use SKStoreReviewController (or the SwiftUI equivalent, requestReview). You get three prompts per 365-day period per app per device. Apple decides whether to actually show the prompt, so calling the API doesn't guarantee the dialog appears.

You cannot incentivize reviews (no "rate us for a free month"), and you can't use custom UI to mimic the system prompt. If you pre-screen ("Are you enjoying the app? Yes/No") and only route "Yes" answers to the review prompt, Apple's review team may flag your app. This used to be common; Apple now explicitly discourages it.

Google's In-App Review API has similar constraints. You can't control when the prompt appears (Google decides based on frequency limits), and you can't incentivize. But Google is less restrictive about when you call the API. You won't get rejected for calling it after a positive interaction.

The three-per-year limit on Apple is the reason timing matters so much. You have three shots. Wasting one on a user who just hit a bug is a shot you can't get back for months. Some developers track positive signals internally and only fire the prompt when multiple positive conditions are met simultaneously: loyal user (10+ sessions), recent success event, no recent crashes.

What to do about bad ratings you already have

If you're sitting at 4.2 with 500 reviews, you need a plan. The math of ratings means you need a lot of 5-star reviews to move the average. Going from 4.2 to 4.5 with 500 existing reviews requires roughly 200 new 5-star ratings (assuming no more bad ones come in). That's months of work.

First, fix whatever is causing the bad reviews. Read your 1-star and 2-star reviews and look for patterns. If 40% of them mention a specific bug or missing feature, that's your priority. A 1-star review about a bug you've fixed is wasted damage. Fix it, then tell users you fixed it in your update notes.

Second, respond to negative reviews. On both platforms, you can reply publicly to reviews. A thoughtful response ("Thanks for reporting this. We fixed the sync issue in version 2.3.1.") does two things: it sometimes gets the reviewer to update their rating, and it signals to future readers that someone is actually maintaining this app.

Third, update your review prompt timing using the principles above. If your prompt has been firing at random intervals, switching to post-success-event timing can shift your incoming rating mix from 3.8 average to 4.5+ average on new reviews.

Fourth, use update notes well. Write human release notes that acknowledge what you fixed. "Fixed the crash on startup that some users were experiencing" tells the reviewer who left a 1-star for that crash that you listened. Some will go back and update their review. Many won't, but even a few helps.

The outsized impact of 1-star reviews

One 1-star review has more mathematical impact than one 5-star review. If you have a 4.7 with 100 reviews and get a single 1-star, you drop to 4.66. Getting a 5-star only moves you to 4.70. The damage-to-repair ratio is roughly 12:1 for a 4.7-rated app. One angry user requires about twelve happy ones to offset.

This means preventing 1-star reviews is more valuable than generating 5-star ones. And most 1-star reviews come from preventable situations:

• Bugs and crashes. Ship stable updates. Use a crash reporter. If you see a spike in crashes after an update, fix it within days, not weeks.
• Broken onboarding. If users can't figure out what to do in the first two minutes, some of them will leave a 1-star before uninstalling.
• Bad paywall timing. Hitting someone with a paywall before they've seen value is one of the fastest ways to earn a 1-star.
• Data loss. If your app stores user data and it disappears after an update, that's a guaranteed wave of 1-star reviews. Test data migration paths thoroughly.
• Subscription frustration. Users who feel tricked into a subscription or can't figure out how to cancel will leave angry reviews. Make cancellation clear.

Our downgrade rage scanner specifically looks for apps where recent updates caused a ratings drop. It's a pattern that shows up more often than you'd expect: a perfectly good app ships a bad update, the rating tanks, and users leave for competitors.

Ratings work differently in different countries

Rating behavior varies by market in ways that catch people off guard. In Japan, users tend to rate more conservatively. A 4.3 in the Japanese App Store might be equivalent to a 4.6 in the US in terms of user satisfaction. Korean users rate slightly more generously than Japanese users but still more conservatively than American users.

This matters if you're localizing for Asian markets. A 4.5 rating in Japan might put you above most competitors, while the same rating in the US puts you below average. Check your category in each specific market before panicking about your number.

Also worth noting: Apple shows ratings per-country by default. Your US rating and your German rating are calculated separately. If you have a small user base in a country, a few bad reviews there can tank your local rating even if your global average is fine. Monitor ratings by country in App Store Connect, especially for markets where you're actively trying to grow.

The review text matters too, not just the stars

Users read reviews, especially the most recent ones and the most critical ones. Apple shows a "Most Helpful" review and the most recent reviews prominently. If your most helpful review is a 2-star complaint from eight months ago about a bug you've fixed, it's still influencing decisions today.

Reply to it. Mention the fix, the version number, and invite them to try again. Even if they never update their review, future readers will see your response and think "okay, this developer is paying attention."

Some developers worry that responding to negative reviews draws attention to them. In practice, the opposite is true. An unanswered 2-star review looks like nobody's home. A replied-to 2-star review looks like an active developer who cares. The research on review reading patterns suggests that potential users actually seek out negative reviews to evaluate worst-case scenarios. Your response is part of their evaluation.

Why you shouldn't try to game it

I should address this because the temptation is real. Services exist that sell ratings and reviews. Some developers create burner accounts and review their own apps. Others use "review exchange" communities where developers rate each other's apps.

Apple and Google detect all of these with increasing accuracy. Apple has pulled apps entirely for review manipulation. Google has removed millions of fake reviews in bulk purges. If you get caught, you lose your reviews, potentially your app, and possibly your developer account.

Beyond the risk: fake reviews don't solve the underlying problem. If your app has a legitimate 4.0 rating, buying reviews to reach 4.7 just means users download something that doesn't match their expectations. They'll leave real 1-star reviews that drag you right back down, and now you've also wasted money.

The only sustainable path is making the app better and prompting at the right time.

A practical playbook for improving your rating

Here's what I'd do if I had a 4.3 and wanted to get to 4.6 within three months:

Week 1: Audit. Read every 1-star and 2-star review from the past six months. Categorize them. Usually three or four issues account for 80% of complaints. Write them down.

Week 2-3: Fix the top two issues. Don't try to fix everything. Focus on the problems that generate the most anger. Ship a focused update.

Week 3: Update your review prompt logic. Move it to post-success events. Add conditions: only prompt users with 7+ days of usage, 5+ sessions, no crashes in the last 3 sessions. This filters your prompt audience to people most likely to rate well.

Week 3-4: Respond to every negative review from the last three months. Keep responses short, specific, and non-defensive. "This bug was fixed in v2.4. Sorry about the trouble." Don't argue.

Week 4+: Write useful update notes with every release. Mention specific fixes. Users who had those problems will sometimes go back and change their review. Even if they don't, new users reading the update history can see you're actively improving the app.

Month 2-3: Monitor incoming ratings weekly. Track the ratio of 5-star to 1-star reviews. If you're getting 3:1 or better (three 5-stars for every 1-star), your average will climb. If the ratio is worse, you have more product work to do.

The timeline is realistic but not fast. Moving a rating 0.3 points with hundreds of existing reviews takes steady effort over weeks, not a one-time push.

Your rating is a lagging indicator of your retention

This is the part most guides skip. Your rating is a symptom, not the disease. Apps with high retention have high ratings because users who stick around are more likely to rate and more likely to rate well. Apps with poor retention get bad ratings because their users are frustrated enough to leave.

If your day-1 retention is below 30%, fixing your review prompt won't save your rating long-term. You're losing 70% of users before they could even become promoters. The review prompt optimization only works if you have enough happy users to prompt.

Work backwards: improve retention first, especially the first session experience. Once users are actually sticking around and getting value, the review prompt does its job. Without retention, you're trying to squeeze good ratings out of a shrinking pool of users.

The same logic applies to the connection between ratings and subscription churn. High churn apps tend to accumulate angry reviews from people who feel they didn't get their money's worth. Reduce churn, and your incoming review sentiment improves automatically.

Using competitor ratings as market intelligence

Your competitors' ratings tell you things their marketing won't. An app with a 4.1 rating and declining review sentiment is vulnerable. An app that dropped from 4.7 to 4.3 after a recent update just alienated a chunk of its user base.

When you read competitor listings, check their recent reviews specifically. Sort by most recent and read the 1-star and 2-star ones from the last three months. These are real-time signals about what users want that the competitor isn't providing.

Our scanners do something similar at scale. The downgrade rage scanner finds apps where ratings are dropping, which often means a recent update broke something users cared about. That's an opportunity if you can offer what the competitor just took away.

Mistakes that keep your rating stuck

Prompting on first launch. The user installed your app 30 seconds ago. They haven't done anything yet. Even if they dismiss the prompt, you've used one of your three chances on someone with no opinion.

Prompting after every update. Some apps trigger the review prompt after every version update. Users who've already rated get annoyed. Users who haven't rated get prompted at a random moment unrelated to their experience. Neither is ideal.

Ignoring the review response feature. Both platforms let you respond publicly. If you're not doing this, you're missing an easy way to demonstrate that your app is actively maintained.

Resetting ratings unnecessarily. Apple lets you reset your cumulative rating with a new version. Developers with a 4.3 and 2,000 reviews reset, hoping to start fresh. But now they have zero reviews, which looks worse than 4.3 with 2,000. Only reset if you're below 3.5 and have genuinely rebuilt the app.

Treating all markets the same. Your localization strategy should include thinking about ratings per-country. A bug that only affects users in Germany might tank your German rating while your US rating looks fine. You won't catch this unless you check country-level data.

Obsessing over individual reviews. One bad review with a fair point is useful feedback. One bad review from someone who expected your app to be something completely different is noise. Don't rewrite your roadmap for outliers.

Where ratings fit in the bigger picture

Ratings are one part of your App Store presence, alongside your screenshots, your keyword strategy, your category choice, and your distribution channels. Improving your rating from 4.3 to 4.6 might increase conversion by 15%, but if nobody is seeing your listing in the first place, that 15% improvement applies to a small number.

That said, ratings compound with everything else. Better keywords get you more impressions. Better screenshots convert more of those impressions to detail views. A better rating converts more detail views to downloads. Each improvement multiplies the others.

If you're doing A/B testing on your listing, keep in mind that your rating affects test results. A screenshot test while you have a 4.2 rating will produce different conversion numbers than the same test at 4.7. Fix your rating and then test, or at least be aware that the rating is a confounding variable.

Ultimately, the best thing you can do for your rating is build something people want to keep using. An app that solves a real problem, works reliably, and respects people's time and money will accumulate good reviews naturally. The prompting strategy, the response strategy, the timing, the market analysis are all important, but they're multipliers on a product that needs to be fundamentally worth rating well.