How to Use Muse to Analyze Customer Reviews

1. The short answer

One early tester told Meta that Muse worked through 87,000 customer reviews from his website and built a full strengths, weaknesses, opportunities, and threats analysis in minutes, work he said would have taken him months by hand. You can run the same workflow: give Muse your reviews and ask for the patterns, the repeat complaints, and the praise worth amplifying. Available in the US and Canada right now.

Here is why this matters more than it sounds. Reviews are the only place customers tell you the truth in their own words. But nobody with a business to run has time to read hundreds of them properly, so the reading never happens and the patterns stay hidden. This guide shows you how to feed Muse your reviews, which questions surface the most value, and how to turn the analysis into action.

Muse is invite only right now. Need an invite code? Click here to grab one free from the community pool. Waiting on a launch in your country? Join the waitlist and we will send you a fresh code on launch day.

2. What you need

• Muse access. You need an account with access first.
• Your reviews in one place. An export from your store or review platform, a connected tool that holds them, or pasted text. Whatever is easiest to gather is fine.
• A goal for the analysis. Fix complaints, find marketing angles, brief staff. The goal shapes the questions, so pick one before you start.

3. The method, step by step

The frustrations, because you have felt most of them. You have hundreds of reviews and no time to read them, so they pile up unread. One harsh review ruins your morning even though fifty happy customers said nothing. You suspect the same complaint keeps appearing but you cannot prove it, so you cannot justify fixing it. Customers praise things about your product that never appear in your marketing. You know something should change but you cannot tell which complaint to fix first. And replying to reviews takes forever, and half your replies sound defensive.

The method turns the pile into decisions. Read everything once through Muse, count what repeats, fix the top complaint, amplify the top praise, and draft the replies without the defensiveness.

1. Gather the reviews. Export them from your store or review platform, or connect the tool that holds them to Muse. Pasted text works too. Do not filter them first. The one star and the five star reviews both carry signal.
2. Ask for the breakdown. Say: analyze these reviews. What do customers love, and what frustrates them, in their own words. Ask for their phrases, not summaries. Their words become your marketing copy later.
3. Go deeper on patterns. Ask: which complaints repeat, and which products get the most praise. Frequency matters more than any single review. Ten people mentioning slow shipping is a fact. One person disliking the color is an opinion.
4. Ask for the full picture. Say: build me a strengths, weaknesses, opportunities, and threats analysis from these reviews, with quotes as evidence. The quotes matter. They keep the analysis honest and give you material you can show your team.
5. Turn it into drafts. Say: draft replies to the top three complaints, or turn the top praises into homepage copy. Approve everything before it goes public. The drafts give you calm, professional replies without the hour of staring at the screen.
6. Rerun it regularly. Run the analysis monthly or quarterly. Trends across runs tell you whether the fixes worked. A complaint that fades after you fix shipping is proof. One that keeps growing is your next priority.

Copy this prompt

You are my customer feedback analyst. Here is how we work.
My reviews: [paste them below, or say: read the export I uploaded, or: read the reviews in my connected store]
My goal for this analysis: [pick one: find what to fix first, find marketing angles in customer praise, brief my staff on repeat issues]
Rules: you analyze and draft. You never post a reply or change anything public without my explicit approval. Quote real review text as evidence and never invent quotes. If the data is too thin for a claim, say so.

Do this in order.
1. Read every review. Summarize in plain words what customers love most and what frustrates them most, using their own phrases where you can.
2. Count the patterns: which complaints repeat and how often, and which products get the most praise. Frequency matters more than any single review.
3. Build a strengths, weaknesses, opportunities, and threats analysis from the reviews, with 2 or 3 real quotes supporting each point.
4. Tell me the single complaint worth fixing first and why, plus the single praise worth amplifying in my marketing and how you would phrase it.
5. Draft replies to the 3 most common complaints in a calm, human voice. Show me the drafts. Do not post them.

4. Turning analysis into action

Analysis only pays when it changes something. Three moves cover most of the value.

Fix the top complaint and say so publicly. Pick the complaint with the highest frequency, fix it for real, then tell customers. A reply that says we heard you about shipping times and here is what changed does more for trust than ten five star reviews. Then watch the next analysis run to confirm the complaint fades.

Amplify the top praise in your marketing. Customers describe your product better than you do, because they describe the feeling, not the features. Take their exact phrases and put them on your homepage, your ads, your packaging. This is the cheapest good copy you will ever get.

Brief your team on the patterns. The same issues recur because nobody named them out loud. Share the top three complaints and the top three praises with everyone who touches the product or the customer. A team that knows what customers actually say stops guessing.

Run the analysis monthly or quarterly. Trends across runs tell you whether the fixes worked, which is the difference between a one time exercise and a real feedback loop.

5. What to watch for

Treat the analysis as a strong first pass, not a verdict. Spot check surprising claims against the raw reviews, especially before changing a product or policy over them. If Muse says everyone hates the packaging, read ten of those reviews yourself before you redesign anything.

Keep approval on for any public reply Muse drafts. A calm, professional draft still goes out under your name, so give it your eyes first. And remember the evidence rule: real quotes only. If a claim has no quote behind it, treat it as a hypothesis, not a finding.

This is US and Canada only for now. Everywhere else, the waitlist is the way in.

Get started

Muse has not launched in your country yet? Join the waitlist and we will send you a fresh code on launch day.

Frequently asked questions

How many reviews can it handle?

It depends on your plan's usage limits. The example Meta shared involved 87,000 reviews processed in minutes.

Does it work with Google reviews?

It works with whatever reviews you can give it: exports, connected tools, or pasted text.

Will it reply to reviews for me?

It can draft replies. Posting anything publicly needs your approval.

Is the analysis trustworthy?

Treat it as a first pass and spot check against the raw reviews, especially for surprising claims.

What if many of my reviews are incentivized or fake?

The analysis is only as honest as the reviews you feed it. If you suspect fake reviews, say so in the prompt and ask Muse to give more weight to detailed, specific reviews, which are harder to fake than one line praise.

Can it compare my reviews with a competitor's?

Only from public review text you provide yourself. It cannot pull private data or browse on your behalf, so gather the competitor reviews first and paste or upload them alongside yours.

How often should I rerun the analysis?

Monthly or quarterly. A single run tells you where you stand. Repeated runs tell you whether your fixes worked, which is where the real value compounds.

My reviews are not in English. Does it still work?

Muse reads many languages. Ask for the summary in the language you prefer, and spot check the key claims against the original reviews as you would in English.

I am outside the US and Canada. What now?

Muse has not launched in your country yet. Join the waitlist and we will send you a fresh code on launch day.

Built with Muse is an independent community project. Not affiliated with or endorsed by Meta.