Which Problems Go With Bad Reviews at Home Depot Stores?
Using 2,354 public reviews of 24 Atlanta Home Depot stores, which problems go with one- and two-star ratings, and how to compare stores without ranking noise.
A retail chain with hundreds of stores has two questions about its reviews. Which store problems actually go with bad ratings, and which stores are really doing worse than the rest? The second is harder than it looks. A store's rating depends on who shops there and how big it is, and with a hundred or so reviews a store, part of any gap between two stores is chance.
Below, both questions are worked through on public Google reviews of one retailer, with the checks we use to keep the store comparison fair.
The data: 24 Home Depot stores in metro Atlanta
2,354 public Google reviews of 24 Home Depot stores in metro Atlanta, January 1, 2025 to June 12, 2026. Every store sits in the same metro, sells the same range under the same brand, and faces the same housing market, so the market is held roughly constant. What differs is how each store is run, and who walks in.
Of the 2,354 reviews, 1,447 are five stars and 481 are one star; 25.7% are one or two stars.
From the Atlanta Home Depot brief
One market, 24 stores, one problem at the top
Public Google reviews of 24 Home Depot stores in metro Atlanta, January 2025 to June 2026.
Which problems go with bad ratings
The brief tagged every review with the problem it describes, then asked two things of each problem: how much more likely a review that names it is to be one or two stars, compared with a review that names no complaint at all, and how often it comes up.
Not getting help on the floor (no one to ask, long waits for help, or rude and unhelpful staff) appears in 597 reviews, about a quarter of the set. Those reviews are 16.5 times as likely to be one or two stars as a review with no complaint. Problems with online orders and pickup (16.7 times, 136 reviews), returns (19.2 times, 57 reviews) and installation (20.3 times, 27 reviews) are about as bad or worse when they happen. Pricing and checkout complaints carry a lift of 9.0 times across 40 reviews.
Ranked on lift alone, installation comes first. Multiply each lift by the number of reviews and the order changes: not getting help scores 9,849, more than four times online orders at 2,278, then returns (1,097), installation (548) and pricing (359). A rare, severe problem and a common, slightly milder one need different responses: the first is a set of exceptions to escalate, the second is a staffing and coaching question for every store.
From the report: Atlanta Home Depot brief
The good reviews point at the same place. Shoppers praise a named employee 577 times and call out a named bad one 97 times, roughly six to one.
Check it is not the shopper or the store size
A chain-wide link can hide a mix effect. Pro contractors shop differently from DIY homeowners and may be harsher reviewers; a large store has more aisles to cover than a small one. If the stores with the worst problem rates also had the most Pro shoppers, the pattern could come from who shops there rather than how the store is run.
So the brief split the "can't get help" reviews both ways. They are one or two stars 76.9% of the time for DIY homeowners, 90.5% for Pro and contractor shoppers and 77.8% where the shopper type is not stated. By store size the rate stays between 76.4% and 80.1%. Few reviews say who the shopper is, so the DIY and Pro groups are small, and the brief does not publish their sizes or the cut-offs it used for store size. Reviews that mention not getting help stay mostly negative in every group, 76% to 91%, which is what you want to see before acting on a chain-wide finding. Pro shoppers who cannot get help are the most likely of all to leave a bad review.
From the report: Atlanta Home Depot brief
How to compare the stores fairly
The raw ranking runs from 14.7% one- and two-star reviews at 2525 Piedmont Rd NE to 32.8% at 1480 Satellite Blvd. Satellite Blvd has 67 reviews. With samples that small, the ranking is partly noise.
Putting a 95% interval around each store's rate shows how much. For a store with about 100 reviews and a rate near the average, the interval runs about 8.5 points either side. Satellite Blvd's runs from 22.8% to 44.7%. Only one store's interval sits clear of the 25.7% average: Piedmont, at 10.0% to 21.1%.
Dimension Labs analysis
With about 100 reviews a store, only one of the 24 stores is clearly apart from the average
Share of each store's reviews that are one or two stars (dot) with a 95% interval (line), worst to best. Grey: under about 70 reviews.
Dashed line: fleet average, 25.7%
A star-rating league table alone cannot tell most of these stores apart, so the brief used three other moves, and they are worth copying.
Flag thin stores instead of ranking them. Stores under about 70 reviews are marked "read with caution" and never used for a headline. That is why the first store to visit is 4325 New Snapfinger Woods Dr (31.5% one- and two-star across 127 reviews, with "can't get help" in 33.1% of its reviews) and not Satellite Blvd. Its interval still overlaps the average, so it is the first store to look at, not a proven outlier.
Pool stores to test the store-level link. In the quarter of stores with the most bad reviews, 29.5% of reviews cite not getting help; in the best quarter, 20.2%, a gap of about 9 points inside one city and one brand. Pooling six stores at a time gives the comparison enough reviews to mean something.
Find the store to copy. Piedmont is the one store that is clearly better than average. Eighty percent of its reviews are five stars, and its reviews name more praised employees than any other store's.
The restaurant version of this problem, where stores are grouped by the kind of location they sit in, is in how to compare restaurant locations using guest reviews. The two methods fit together: group stores by location type where traffic differs, and put an interval on every rate either way.
How to run this on your own stores
Hold the market constant. Start with stores in one metro, or compare each store only with stores in similar markets, so the economy is not doing the ranking.
Tag every problem in every review. Floor help, online orders, returns, installation and pricing each get their own field, and a review can carry more than one.
Score each problem on lift and on volume. Lift against reviews with no complaint shows how bad a problem is; multiplying by its count gives a combined score for ranking.
Split by shopper type and store size. Check the link holds inside each group before you fund a chain-wide fix.
Put an interval on every store's rate. Flag stores with too few reviews, and pool stores into quartiles to test whether a problem separates good stores from bad ones.
Pick a store to learn from. The best store's reviews describe what good looks like in your own operation.
Reviews come from shoppers who chose to post, often after a strong experience, so read the rates as comparisons between stores rather than the share of all visits that went badly. They also cannot see payroll: the brief is explicit that reviews show whether a shopper found help, not actual staffing levels. To go from association to cause, join the review tags to what the chain already holds, such as labor hours by store and week. Our post on which complaints cause bad reviews covers that step, and the full store table and exhibits are in the Atlanta Home Depot brief.
FAQ
How many reviews does a store need before you rank it?
More than most stores have. At about 100 reviews, a store's one- and two-star rate is uncertain by roughly 8 to 9 points either way, so rank on intervals, flag thin stores, and pool stores to test patterns.
Should a retailer fix the most common complaint or the most severe one?
Score both. Here installation problems had the highest lift, but not getting help on the floor was common enough to carry more than four times the weight of any other problem.
Does this apply to other retail chains?
Yes. The same steps work for any chain with many locations and public reviews: hold the market constant, tag every problem, check the link inside each shopper and store group, and compare stores with intervals. See retail for how it applies to store networks.
Dimension Labs reads every review, survey and ticket for every location, tags each one with the problems and people it describes, and joins those tags to store data such as sales, labor and location type to show what drives ratings and revenue. Its agent then builds the store rankings, reports and alerts on a schedule. See multi-location intelligence and review intelligence, or start with one file of reviews.