How to Analyze Exported Reviews in Excel/Google Sheets (With Templates)
A simple workflow to turn raw review CSVs into clear charts and insights—no data science required.
TL;DR
Got a CSV of reviews from G2, Capterra, Trustpilot, Shopify, Google Play, Amazon, etc.?
This guide shows a practical, no-code workflow in Excel/Sheets to get:
- Sentiment breakdown (positive / neutral / negative)
- Top complaints & themes
- Simple competitor comparison
- Basic trend over time
You can absolutely do this yourself, but you'll hit limitations as volume grows. At the end I'll also show when it makes sense to switch to AI Analytics instead of fighting with formulas.
1. Start with a clean export
You should have one row = one review. Minimum useful columns:
- rating (1–5)
- date
- review_text (and/or pros/cons)
- reviewer (optional)
- product or brand (if you're comparing multiple)
- source (G2, Capterra, Trustpilot, etc.)
If you're using Reviews Extractor extensions, you already have this format.
2. Import properly (avoid broken text)
Excel
- 1Open Excel
- 2Go to Data → From Text/CSV
- 3Select your .csv
- 4In the import dialog: Set File Origin to UTF-8, Delimiter = Comma, Ensure the review_text column is set as Text, not "General"
This prevents weird characters and keeps long reviews from breaking.
Google Sheets
- 1Create a new sheet
- 2Go to File → Import → Upload
- 3Choose your CSV
- 4Select "Insert new sheet(s)" and make sure the separator is Comma
3. Create sentiment buckets from star ratings
Add a new column: sentiment_bucket.
In Excel/Sheets, use something like:
=IF(A2<=2,"Negative",IF(A2=3,"Neutral","Positive"))(Assuming A is the rating column.)
Now you can:
- See how many reviews are negative vs positive
- Filter to only negatives when searching for pain points
4. Build a basic sentiment pivot
Create a Pivot Table:
- • Rows: sentiment_bucket
- • Values: count of review_text
- • (Optional) Columns: product (if comparing competitors)
Now you get overall sentiment breakdown or sentiment per competitor/product.
Add a simple bar chart to make it visually obvious.
[screenshot spot: pivot table with Positive/Neutral/Negative counts]
5. Find top complaint keywords (fast & dirty)
Full NLP is what your AI dashboard does, but for a quick spreadsheet-only view:
Add a helper column lower_text:
=LOWER(C2)(Assuming C is review_text.)
Create "keyword columns" for the things you care about, e.g.:
- mentions_support
- mentions_price
- mentions_bug
- mentions_integration
- mentions_ui
In each column, use formulas like:
=IF(ISNUMBER(SEARCH("support", $D2)),1,0)=IF(OR(ISNUMBER(SEARCH("price",$D2)),ISNUMBER(SEARCH("pricing",$D2))),1,0)=IF(OR(ISNUMBER(SEARCH("bug",$D2)),ISNUMBER(SEARCH("crash",$D2)),ISNUMBER(SEARCH("error",$D2))),1,0)Then build a pivot:
- Rows: keyword columns (or just create one table per keyword)
- Values: sum of each keyword column
This gives you a rough "heatmap" of what people complain about most.
Limitations (this is where your AI is better):
- Doesn't understand context or sarcasm
- Misses synonyms unless you manually add them
- Doesn't know if the mention is positive or negative
But as a first pass, this works surprisingly well.
6. Analyze trends over time
To see if things are getting better or worse:
Ensure date is imported as a proper date, then add a helper column month:
=EOMONTH(B2,0)(Assuming B is date.)
Make a pivot:
- Rows: month
- Columns: sentiment_bucket
- Values: count of review_text
Add a line chart / area chart to visualize:
- Negative vs positive over time
- Spikes after releases / pricing changes
You can do the same per product or source.
7. Compare competitors side-by-side
If you exported multiple tools:
Make sure you have a product column, then create a pivot:
- Rows: product
- Columns: sentiment_bucket
- Values: count of review_text
Optional: normalize by total reviews per product to get percentages:
- Add a total_reviews column using
=COUNTIF(product_range, product_cell) - Create calculated field negative_share = negative_count / total_reviews
Now you can literally see: "Tool A has 35% negative reviews" vs "Tool B has only 12% negatives but more complaints about pricing"
Great slide fodder for decks and internal discussions.
8. Turn findings into decisions
Once you have:
- Top complaint keywords
- Sentiment trends
- Competitor comparison
You can start asking:
What should we fix first?
Onboarding? Support? Performance? Pricing messaging?
What should we build?
Integrations? Mobile app? Specific workflows?
How should we position?
"Simpler than X", "Faster support than Y", "Built for [segment] who are ignored by Z"
You can also plug the themes into a prioritization framework (like the FAST framework you already use: Frequency, Affected segment, Severity, Time-to-fix).
9. When spreadsheets stop being enough
Spreadsheets are fine for:
- One-off analyses
- 200–1,000 reviews
- A few keywords and simple pivots
But they struggle when you want:
- Real theme detection without manually defining keywords
- Positive vs negative mentions of the same feature
- Dozens of competitors & sources (G2 + Capterra + Trustpilot + Google Play…)
- Monthly automation without re-cleaning everything
That's where your AI Analytics shines:
Upload CSVs from any source (or directly from your extensions), and models do:
- Document-level sentiment
- Aspect-based sentiment (support, price, features, UX, etc.)
- Automatic clustering of themes
- Feature-request mining
- Competitive signals ("switched from…", "cheaper than…")
Dashboard gives:
- • Top pain points
- • Top opportunities
- • Trend charts
- • Executive summaries with backing quotes
Skip the formulas next time
Keep using Excel/Sheets if you enjoy tinkering—but if you want automatic pain-point detection, instant competitor comparison, and AI-generated summaries for PMs, growth & leadership, you can upload the same CSV to AI Analytics and get a full insight report in minutes.