How to Mine Competitor Reviews for Insights Using a Reviews API
Reviews tell you what customers reward and what they punish — and your competitors' reviews tell you the same about them.
This guide walks through a spreadsheet-based workflow that collects reviews for several businesses in bulk through a reviews API, then breaks the text into recurring phrases so patterns surface quickly. You will finish with a shortlist of strengths to match and weaknesses to answer, plus concrete places to use them: your website copy, paid search ads and day-to-day operations.
Why competitor reviews are worth mining
Reviews sit at the intersection of search visibility and reputation, so they influence both how a business is found and how it is judged once found. Reading a handful by hand gives you anecdotes; reading hundreds across several competitors gives you themes.
The useful split is directional. Praise shows what a rival is doing that customers notice and value — the standard you need to meet. Complaints show where expectations are being missed — the gap you can answer honestly. Treat both as evidence about customer priorities rather than as a verdict on any single business.
What the workflow needs before you start
The approach uses a Google Sheets workbook driven by AppScript, a JavaScript-based language, calling the reviews API to gather reviews for multiple businesses in bulk. Before touching the sheet, confirm four things:
- A reviews API account with an API key and API secret available from the API Access page in your account settings.
- A copy of the workbook you can edit, kept separate from any shared master.
- A list of the businesses you want to compare, with the directory you want reviews pulled from for each.
- A realistic expectation about timing: review collection runs as a batch job, so results arrive shortly after the script finishes rather than instantly.
Step 1: add your API credentials
Open the Config tab and place your API key in cell C5 and your API secret in cell C7. These two cells are the only credential inputs the workbook needs; everything else builds on them.
Keep the sheet private while credentials are in it. If you share the workbook, clear those cells first and re-enter the values in your own copy.
Step 2: pull in the business directories
From the workbook's Reviews menu, choose Get Directories. The workbook queries the API for the directories it can analyse reviews from, and the list populates on the RAW – Config tab. You only need to do this once per session; the directory list then feeds the drop-down used in the next step.
Step 3: enter the businesses to compare
On the Businesses tab, fill in the white input columns (D to I) with the details for each business whose reviews you want. Column G is the directory field and offers a drop-down built from the directories you just pulled, so each row points at a specific source.
If you are unsure which competitors deserve attention, a local search audit can surface the businesses you are competing with in local results. As a worked example, a set of burger restaurants in New York City would be entered row by row, each with its own directory selection.
Step 4: collect the reviews
Choose Get Reviews from the workbook's Reviews menu. The Job ID, Batch ID and Status columns fill in as the request is queued.
The script will often finish running while Status still reads In Progress. That is expected behaviour, not a failure: the batch API processes the request in the background, and after a minute or two the status changes to complete and the RAW – Reviews tab fills with the collected reviews. Wait for the status change before judging the run; only treat it as a problem if the status does not resolve.
Step 5: read the positive and negative themes
Two analysis tabs do the interpretive work. Insights – Positive looks only at reviews scoring 5 or higher; Insights – Negative looks only at reviews scoring 3 or lower. Each runs an n-gram analysis and lists the most frequent bi, tri and quad grams — two-, three- and four-word sequences — so repeated phrasing rises to the top.
In the burger example, the positive tab surfaced three recurring ideas: the burgers being the best, a relatively clean location, and meals at a good price. The negative tab surfaced two: long wait times and overcrowding. Those five phrases are the raw material for everything that follows.
Turning themes into website copy
Complaints are the easiest themes to answer on-site. If a rival is repeatedly criticised for something you handle well, say so plainly in your own copy — a short callout that names the concern and states your position on it. In the example, a competing burger chain could highlight a short wait for food, because wait time is the recurring complaint about the rival.
Keep the claim true and specific to your operation. Copy that promises something you cannot deliver simply moves the complaint to your own review profile.
Using themes in paid search ads
A conquesting campaign targets searches around a competitor's brand and uses the negative themes as headline material. The example ad lines play directly on crowding and speed — asking whether the reader is tired of crowds, or promising a burger in under 30 seconds.
Two practical checks before you run this: make sure the claim in the headline is one your business can stand behind, and follow the ad platform's rules on competitor targeting and trademark use in your market.
Improving your own operation
The most durable use of the data is internal. Positive themes show what customers of a rival value, which is a reasonable proxy for what your own customers will value. If a competitor is praised for consistent cleanliness and fair pricing, those become operational targets: a cleaning schedule you actually keep, and pricing you can defend as fair.
This is the step that compounds. Copy and ads borrow a competitor's weakness for a campaign; process changes remove your own weaknesses permanently.
A quick pre-flight checklist
Run through this before each analysis cycle:
- Credentials entered in Config cells C5 and C7, and the sheet kept private.
- Directories pulled via Get Directories and visible on RAW – Config.
- Business rows complete in columns D to I, with a directory chosen in column G for every row.
- Get Reviews run, and Status allowed to move from In Progress to complete before reading results.
- Both Insights tabs reviewed, with the top phrases written down and matched to an action.
- Each planned claim checked for accuracy before it reaches a website or an ad.
Where to take it next
Three applications — site copy, paid ads and operational fixes — are a starting set, not a limit. The same phrase lists can inform FAQ content, sales objection handling, onboarding scripts or staff training. The constraint is not the data; it is whether each insight can be turned into something true and useful for your own customers.
| Area | Field or control | What it holds | Why it matters |
|---|---|---|---|
| Config tab | Cell C5 | API key | Authenticates requests to the reviews API |
| Config tab | Cell C7 | API secret | Second half of the credential pair |
| Reviews menu | Get Directories | Directory list | Populates RAW – Config and the column G drop-down |
| Businesses tab | Columns D to I | Business details | One row per business to be compared |
| Businesses tab | Column G | Directory | Drop-down of directories pulled in the previous step |
| Reviews menu | Get Reviews | Batch request | Queues review collection for every business row |
| Businesses tab | Job ID, Batch ID, Status | Request tracking | Status moves from In Progress to complete |
| RAW – Reviews tab | Collected reviews | Raw review data | Source for both analysis tabs |
| Insights – Positive tab | Bi, tri and quad grams | Frequent phrases | Only reviews scoring 5 or higher |
| Insights – Negative tab | Bi, tri and quad grams | Frequent phrases | Only reviews scoring 3 or lower |
Field names and cell references describe the workbook as documented in July 2020; check the current template before relying on exact cell positions.
| Review signal | Observed theme | Type | Possible action |
|---|---|---|---|
| Positive | The burgers are the best | Strength to match | Ensure product quality is consistently the headline of your own offer |
| Positive | Their location is relatively clean | Strength to match | Set and keep a visible cleaning routine |
| Positive | Their meals are a good price | Strength to match | Review pricing so value is defensible |
| Negative | There are long wait times | Weakness to answer | State your actual service speed in site copy and ad headlines |
| Negative | It was overcrowded with a lot of people | Weakness to answer | Address crowding in ad copy where your venue genuinely differs |
Themes come from one illustrative example set of reviews; they are not a general finding about any chain and should not be reused as claims about a named business.
Frequently asked questions
Do I need to wait before reading the reviews?
Yes. The script can finish while the Status column still shows In Progress. That reflects the batch API working in the background; after a minute or two the status changes to complete and the RAW – Reviews tab fills in.
Which reviews land in each insights tab?
Insights – Positive uses only reviews that scored 5 or higher. Insights – Negative uses only reviews that scored 3 or lower. Reviews in between are collected but do not feed either analysis.
What does the n-gram analysis actually produce?
A ranked list of frequently repeated word sequences of two, three and four words. Reading the top entries shows the phrases customers use most when describing a business, which is faster than reading every review individually.
Can I use competitor review themes in ads?
The workflow supports it, including conquesting campaigns built around a rival's recurring complaints. Check that any claim in the headline is true for your business and that the ad platform's rules on competitor targeting and trademarks are met in your market.