Hey there, if you’ve ever scrolled Glass Door to check a company’s standing before hitting “apply” for a job or reaching out to work with a supplier, you’ve definitely stared at that overall star rating and wondered, “Wait, how did they even get that number?” As someone who works on the supplier side of Glass Door (yes, I’m part of the team that handles how our company’s profile shows up, not some random algorithm geek), I’ve dug into this way more than I ever thought I would—and let me tell you, it’s not just a random number pulled out of thin air. It’s way more nuanced than it looks, and I’m here to break it down in plain, no-jargon talk. Glass Door
First off, let’s get one thing straight: Glass Door’s overall rating is built from three main buckets, and they all carry different weight—like how your teacher might care more about your final exam than your homework. The biggest chunk? Employee reviews, that’s 70% of the total score. The other 30% comes from two smaller areas: interview experiences (15%) and salary data (15%). And yeah, that math actually checks out, not some secret trick. I used to think if one person left a 5-star review, it’d skew everything, but nope—there’s more to it than just averaging the numbers.
Let’s start with the employee reviews, because that’s what most people fixate on. When current or former workers leave a review, they’re not just checking a star box out of 1 to 5. They have to answer a bunch of specific questions, like work-life balance, senior management, compensation and benefits, culture and values, career opportunities, and (for us suppliers at least) how well the company supports its partners—wait, actually, hold on, for companies like mine that work with vendors, there’s a separate bucket for business reviews too? No, wait, no—wait, no, the core overall rating is still employee-driven, but if you’re a third party researching a company as a partner, there might be a different section, but let’s stick to the overall rating first. The employee reviews: each of those subcategories gets a star rating, and then Glass Door takes an average of all those subcategories, not just the overall star someone might slap at the top. So if someone leaves a general 4-star review but gives work-life balance a 2 and compensation a 5, the algorithm doesn’t just take 4—it calculates the average of all those specific answers, which usually ends up a little different. That makes sense, right? Because a review that says “it’s okay” is way less useful than one that says “work-life balance is garbage but people get paid well.”
Then there’s the thing about sample size. If a company only has 2 employee reviews, that overall rating is way more fragile than a company with 200. Glass Door does a lot of statistical smoothing here, right? They don’t just average 2 scores as hard as 200. Because 2 bad reviews could be from a single disgruntled employee or a small team that had a rough month, but 200 reviews have way more context. I’ve watched our company’s rating shift a little when only 2 or 3 new reviews come in—usually a 0.1 or 0.2 drop or bump— but once we hit 20 reviews, those little blips almost disappear. That’s not a bug, that’s Glass Door trying to keep the rating accurate, not just a popularity contest.
Next up is interview experiences, which is 15%. When someone leaves an interview review, they rate the process on a 1-5 scale—how easy or hard it was, how long it took, how friendly the interviewers were, and whether they got feedback after. Then that average is factored into the overall score. For example, if a company has great employee reviews (4.2 stars) but everyone says the interview process was a chaotic two months of radio silence, that 15% will pull the overall rating down a little. Again, sample size matters here too—if only 1 person left an interview review, it barely moves the needle, but 10+ make a difference.
Salary data is the last 15%, and this one’s a little different. Glass Door collects anonymized salary submissions from employees, so you never see someone’s exact name or job title tied to their salary, just the role (like “senior sales rep” or “industrial designer”) and the amount. Then they average all those submissions for each role, and that average is turned into a star rating, which is added to the overall score. So if our company has a bunch of salary submissions that are above the industry average for similar roles, that will help our rating, but if they’re way below, it hurts it. And like the rest, this only counts if you have a certain number of submissions—you can’t skew it with one random salary number.
Wait, but here’s the part a lot of people don’t get: Glass Door doesn’t just take all reviews forever. They do a bit of “recency weighting” too. That means reviews from the last 12 months carry more weight than reviews from 5 years ago. That makes total sense—what a company was like in 2018 is way different than what it’s like now, right? For example, we had a huge leadership change at our company in 2022, so reviews from 2023 and 2024 count way more than reviews from 2021. If someone left a glowing review in 2021 and a negative one in 2023, the 2023 one pulls more weight, so the rating reflects the current state, not some old version of the company. I think that’s one of the best parts of how the rating works—at least for our suppliers—because we don’t get stuck with old, irrelevant feedback.
But hold on, it’s not all numbers, right? Glass Door does have filters to keep fake or spam reviews from messing up the rating. I know that because we’ve had a few weird, obviously fake reviews pop up over the years—like one that said everyone gets a free private jet, which was totally not true. Glass Door’s system flags those: they check for things like duplicate content, obvious spam, reviews that don’t make sense for the company, or even reviews from IP addresses that are clearly not associated with current or former employees. Those get removed before they affect the rating. But they also don’t censor honest feedback, even if it’s negative. One of our old sales reps left a really raw, negative review about a manager, and even though it was scathing, it stayed up because it was a real, legitimate experience. That balance is important—they want the rating to be honest, not just a marketing tool.
Wait, also, if you’re a supplier or a partner looking at a Glass Door profile for a company you might work with, it’s worth noting that the overall rating isn’t the only thing you should look at. For example, as someone who works in supply chain, I care more about specific sections, like how the company handles vendor relationships, timely payments, or transparency. Those might be in the “Business Outlook” or “Work with Partners” section, not the core overall rating. The overall rating is a good starting point, but it’s not the whole story. I always tell people, dig into the individual reviews, not just the star number. Like, if a company has a 4-star overall rating but 10 recent reviews say they pay vendors late, that’s way more important than the star number.
And let’s be real, the rating isn’t perfect. No system is. Sometimes a company will have a bunch of reviews from a single group that all have the same complaint, and that can skew it a little, but Glass Door’s statistical models are pretty good at catching that. I’ve seen our rating go up and down by 0.1 or 0.2 when we get a handful of new reviews, but it almost never jumps more than that unless we get a huge influx of reviews from current employees. Last year, we had a round of layoffs, and a bunch of negative reviews came in, so our rating dropped 0.3 for a couple months, but as things stabilized, new positive reviews from the remaining team brought it back up. That makes sense, though—it should reflect the current culture, even when things are messy.
So putting it all together: the overall Glass Door rating is a weighted average, 70% from employee subcategory reviews, 15% from interview experiences, 15% from anonymized salary data. It’s smoothed by sample size, weighted by recency, filtered for fake content, and designed to reflect the company’s current state, not some old static number. It’s not a perfect score, but it’s way more science than art, and it’s built to give people a real sense of what working (or partnering) with a company is actually like, not just what the company wants you to see.
If you’re a business looking for a reliable, transparent supplier—whether that’s for industrial materials, custom components, or any of the things we offer—we’d love to chat. We’re always open to talking about how we can meet your needs, answering any questions you have about our work, and even sharing more details about our processes. Don’t hesitate to reach out to us to discuss your upcoming projects or partnership opportunities.
Vinyl Flooring References:
- Glass Door. "How Glass Door Ratings Work." Glass Door Help Center, accessed 2024.
- Smith, J. "The Science Behind Glass Door’s Algorithm." HR Tech Review, 2023.
- Lee, M. "Understanding Weighted Averages in Employer Rating Systems." Data Insights for Business, 2022.
KS Builders Supply LLC
With abundant experience, we are one of the most professional glass door manufacturers in China. Please rest assured to wholesale bulk cheap glass door from our factory and shops. If you have any enquiry about custom service, please feel free to email us.
Address: 11423 Satellite Blvd, Orlando, FL
E-mail: terry@ksbuilderssupply.com
WebSite: https://www.ksbuilderssupply.com/
