Customer Due Diligence: Know Why Customers Stay or Leave Before You Buy
How investors can read a target's public reviews before the data room opens, shown on 9,186 reviews of 33 Planet Fitness clubs, and what to request once access comes.
To do customer due diligence on an acquisition target, read its public reviews as data: tag every review for what went wrong and whether the customer says they will leave, compare each location with locations like it, and rank problems by what they do to the rating rather than by how often they come up. Done before the data room opens, that tells you why customers stay or leave, whether the trouble sits in a few locations or everywhere, and whether it is an operating problem a new owner can fix. For a deal team that then wants the same read on the target's own tickets, surveys and revenue, Dimension Labs is the tool built for it: it reads every customer channel, joins it to operational data and turns the result into the memo, deck and dashboards the deal team and operating partners share.
The worked example uses public data only: our published analysis of 9,186 Google reviews of 33 Planet Fitness clubs in metro Chicago, a listed company whose stock fell 31% in a day in May 2026 after net member joins came in light. It is an illustration of the method on a business anyone can check, not investment advice.
From public reviews only
Four numbers a deal team could have had before the data room opened
Drawn from our published analysis of 9,186 public Google reviews of 33 Planet Fitness clubs in metro Chicago, January 2025 to July 2026.
What should customer due diligence answer?
Most deals need four answers about customers: why they leave, which problems actually cost the business, whether those problems are spread across every location or concentrated in some, and whether things are getting better or worse. A fifth question decides the value-creation plan: is the trouble in processes a new owner controls, such as staffing and billing, or in price and product?
Public reviews answer more of this than most teams expect, for any business with many locations or a consumer app, because every review carries a date, a place and a rating. The table shows what to measure for each question and what it found for Planet Fitness.
Method
Five diligence questions, and what answered each one in the Planet Fitness reviews
What to measure in a target's public reviews, with the answer the published analysis found.
| Diligence question | What to measure in the reviews | What the Planet Fitness reviews showed |
|---|---|---|
| Why do customers leave? | Share of reviews raising each issue that say the customer will cancel, not come back or warn others | Billing or cancellation: 50%, against 1.4% overall |
| Which problems actually hurt? | Each issue's effect on the chance of a one- or two-star review, with location and period held fixed | Front desk +36 points; broken equipment +5 |
| Is it everywhere or in some locations? | Unhappy share by location, compared with similar locations | Urban clubs 3.46 stars; outer-suburban 4.84 |
| Is the rating telling the truth? | Reviews per month by star band; bursts of five-star reviews | About 4,903 extra five-star reviews in nine months, unhappy count flat |
| Fixable operations or a price problem? | Whether the top causes are processes a new owner controls, or price and product | Operational: 26 reviews complain about price, 125 praise it |
Reviews do not give you churn rates or revenue at risk. They tell you what to ask for, and which numbers in the management presentation to test first. The section on the data room below covers the step that turns the read into customers and dollars.
What do customer reviews say about a company before we invest?
In the Planet Fitness case, the public filings and the reviews tell two halves of one story. After the May 2026 miss, management put it down to higher attrition and its easy-cancel messaging, and the company does not report why members leave. The members' reviews fill in that missing half.
The first thing they show is that the loudest complaint is the wrong one to budget for. Broken equipment is the most-named failure, in 122 reviews. Measured by its effect on a one- or two-star review, with club, period, reviewer type and review length held fixed, it comes last at about 5 points. A rude or absent front desk comes first at 36 points (range 26 to 52), with billing failures and cleanliness at about 25 each. A team that counts complaints plans a capital budget for machines. A team that measures effects looks at front-desk staffing and management.
From the report: Planet Fitness: a causal analysis
The second is where talk of leaving comes from. Half of the members who raise a billing or cancellation problem say they will leave or warn others, against 1.4% of all reviews and 13% for front-desk complaints. About one in six billing complaints describes a charge after cancelling or a fee the member did not expect. That is the part of the member experience closest to the attrition management described, and it is a process a buyer can audit in the billing system during confirmatory diligence.
From the report: Planet Fitness: a causal analysis
How do I use reviews to spot churn risk in a portfolio company?
Watch what customers say about leaving, by location type, over time. The star average is the weakest signal of the three, because it moves slowly and can be pushed around. Between February and October 2025, about 4,903 five-star reviews arrived above the normal rate at the Chicago clubs while the number of unhappy reviews did not change. A rating that improved over that stretch would have said nothing about members' experience.
Location type matters as much as the brand. Urban clubs averaged 3.46 stars with a 34% one-star share; outer-suburban clubs averaged 4.84 with a 3% one-star share. For a buyer, that is two operating models inside one P&L, and a roll-out plan that treats every location the same will be wrong for one of them. The same check on a retailer: across 24 Home Depot stores in metro Atlanta, the share of unhappy reviews ran from 14.7% to 32.8% by store (the top figure from a store with under 70 reviews), and reviews about not finding help on the floor were 16.5 times as likely to be one or two stars as reviews with no complaint. See the Atlanta store analysis.
In a portfolio company you own, run the same read every month and compare each location with its own group. A rising share of leaving language in one region is an early warning months before it shows in same-store sales. How to analyze Google reviews across 50 or more locations covers grouping and small-sample adjustments, and what customers will do next, and why covers turning reasons into estimates once you have the company's own outcomes.
What can public reviews not tell you, and what should we ask for in the data room?
Reviews come from customers who chose to write, so they show the shape of the problem, not how many customers left or what they were worth. That is exactly why they are a good first step: they tell you which exports to request and which management claims to test, so the data room time goes on the questions that matter.
Ask for four things: the support tickets, chats and call notes for the last 18 months; cancellation reasons and billing disputes; survey verbatims; and a customer or member table with join date, cancel date, plan and location. Tag the target's own conversations with the same fields you used on the reviews, join them to who actually cancelled, and you can state each reason's effect in customers and revenue, with a range. In Dimension Labs, the join to CRM and revenue data and effects with confidence intervals come with the Prove plan ($2,999 a month) and up.
That joined record of what customers said, what they did, what the business changed and what was happening around it is the measured core of a customer world model. After close it becomes how the operating team checks its first decisions, a price change, new hours, a closure, against what these same customers did after past changes; how to test a business decision before you make it walks through one.
How do I run customer due diligence on public reviews, step by step?
Collect every review for every location. Google, the app stores, Trustpilot and the BBB, for at least 18 months, with a location ID, date and star rating on every row. Pull competitors' reviews the same way if the thesis depends on taking share.
Write the fields before you read. The problems that matter in the industry (staff, billing, cleanliness, wait time, product quality), leaving language, competitors named and what customers praise, each with a one-line definition applied the same way to every review.
Tag every review, not a sample, and check it. Read a random sample of tagged reviews by hand before you trust any count.
Rank problems by effect, not count. Measure each problem's effect on a one- or two-star review with location and period held fixed, as in the Planet Fitness ranking above.
Compare locations with their peers. Group by location type first, then rank within each group. Treat small locations with care: the Atlanta analysis flags stores under about 70 reviews to read with caution.
Read the time series. Look for five-star bursts, a change in the unhappy count and reactions around dated events such as a price change or a closure.
Write the memo and the data room list. Three to five findings with numbers and real quotes, the locations to visit, and the exports that would confirm or kill each finding.
What does Dimension Labs add to a diligence team?
Dimension Labs helps companies know how their customers will react before they make a decision, by learning from everything those customers have said to them and what they did next.
For a deal team and the operating partners after close, that comes down to four things. It is fully omni-channel: public reviews before signing, then the target's tickets, chats, calls, emails and surveys once you have access, read in full rather than sampled. It integrates customer channels with operational data, so the member table, billing records and location data sit next to what customers said (the revenue join is on Prove and up). It enriches every interaction with Dimensions written for the target's industry, such as a billing dispute, a charge after cancelling or a named competitor, so the deal team, the operating partners and the company's own managers work from the same fields. And it adds a visual analytics layer: the diligence findings become a deck, a location dashboard, a PDF or executive report and a shareable link, on a schedule after close.
On cost and return, Travelers, a Fortune 100 insurer, reports 600% ROI within 18 months and $3.13M in total savings from unifying its customer interactions in Dimension Labs. In its Data Science Business Partner's words: "Dimension Labs gives us that 50,000-foot view we needed to figure out what's going on out there." See the Travelers story.
FAQ
How do I do customer due diligence on an acquisition target?
Before the data room, read every public review of the target as data: tag each one for its problems and leaving language, compare locations with similar ones, and rank problems by their effect on bad reviews. Then request the target's tickets, cancellation reasons, survey verbatims and customer table, and measure each reason against who actually left.
What is the best tool for customer due diligence?
For reading every review of a multi-location or consumer business, then the target's own customer channels joined to its operational data, with the memo and dashboards built for you, Dimension Labs. Expert calls and customer interviews add context on a few accounts; review management platforms suit running reviews after close rather than analyzing them for a deal.
What do customer reviews say about a company before we invest?
Why customers leave, which problems hurt the rating most, which locations are unlike the rest and whether the rating has been inflated. In the Planet Fitness reviews, billing trouble led talk of leaving at 50%, and the most-named complaint had the smallest effect.
Can reviews show churn risk in a portfolio company?
They show where it is building. Track the share of reviews with leaving language by location group each month. To turn that into a churn rate and revenue at risk, join the company's own conversations to its customer outcomes.
How many reviews do I need per location?
Enough that one bad week does not decide the rank. Our Atlanta analysis flags stores under about 70 reviews to read with caution. Pool small locations with their group and show ranges rather than single rankings.
Have a target's reviews, or a portfolio company's tickets and surveys? Start with one file and see the first read on your own data.