We analysed 24,900 active employees across 137 hospitality operators and 1,400+ locations. The most engaged employees leave 52% less often than the least engaged - a gap that holds up after we control for location, manager, pay, and tenure. This page goes deep on what we found, why it matters, and what to do with it.
Hospitality has a turnover problem that other industries don't. While most US industries see 12-15% annual turnover, hospitality runs at 74-80%. Quick service restaurants regularly exceed 100% - meaning the staff turn over completely in less than a year, sometimes twice.
The standard playbook for fixing this is more pay, better benefits, faster onboarding. All useful. None of them are the lever this report is about. We wanted to know if a different lever - making the everyday experience of working at a hospitality business genuinely good - moves the same number.
The short answer: yes, by a meaningful margin. The longer answer is the rest of this page. If you'd rather skip ahead, the calculator lets you plug in your numbers, and the playbook section tells you what to actually do.
For broader context on why hospitality retention has become an existential issue rather than just a cost line, our piece on why employee value proposition matters more than ever is a good companion read.
Replacing a frontline hospitality employee costs $3,000 to $5,000 in recruitment, training, and lost productivity. Plug your numbers in below.
For reference: among employees who actively use a workforce platform, the least-engaged third in our UK and Nordic dataset sit at 4.8% per quarter and the most-engaged third sit at 2.3%. Industry-wide annual turnover (US-weighted, QSR-heavy) is much higher at 70-80%; see the FAQ for why the two numbers aren't directly comparable.
Replacement cost band: $3,000-$5,000 per frontline hire (Cornell Center for Hospitality Research). Calculator uses the midpoint.
The savings figure is the gap between current turnover and what your numbers would look like if everyone was a power user of an engagement platform. In practice no team gets to 100% power-user adoption, so treat it as the upper bound. The middle scenario (-15%) is closer to what most operators see in their first 12 months on a platform.
Across 24,900 active employees at 137 paying customers, the most engaged third leaves at 2.3% per quarter. The least engaged third leaves at 4.8%. That's a 52% gap.
That p-value matters. It means the chance this result is a fluke is less than one in 10,000. The effect holds whether we look at the simplest comparison (all users pooled) or the strictest one (same site, same manager, same length of employment). Tightening the controls actually makes the effect stronger, not weaker. That's a strong signal that we're measuring something real.
What "engaged" means here. We measure engagement on two continuous scales: distinct days the employee opened the app in the last 90 days, and chat messages they sent in the last 90 days. We don't use self-reported survey data. This is system-recorded behaviour, not what someone wrote on a Tuesday afternoon engagement survey.
We pulled three system-confirmed signals to identify leavers: admin deactivation, platform removal, and HR-system termination. We required all three to be event-validated rather than relying on shortcut fields, because the shortcut fields miss thousands of real departures.
We then compared employees within the same physical location (same manager, same pay, same local job market) to isolate the effect of platform engagement from any company-level differences. We layered in tenure as a separate control to make sure the result wasn't just new hires churning fast.
Each location gets its own baseline. The engagement effect actually gets stronger with tighter controls, not weaker.
Adding tenure as a control reduces the effect from 30% to 23% per 10 login days. Most of the signal is engagement itself.
Across seven engagement bins, every step shows fewer departures. A consistent monotonic gradient.
Logins can mean checking a rota. Chat means actual workplace communication. The retention signal is sharper there.
At a typical location, an employee logging in roughly once a week has a 13.6% probability of leaving that quarter. Daily users sit at 1.5%. Same location, same manager, same pay - the only difference is how often they engage with the platform.
Location fixed-effects logistic regression. n = 24,900 active users across 1,466 locations. p < 0.0001. Shaded band omitted for readability; full confidence intervals are in the report.
Each percentage on the curve is the absolute predicted probability of an employee leaving that quarter, at the median location in our dataset. At 5 login days (about once a week), roughly 1 in 7 employees leave that quarter. By 50 login days (about daily), the rate drops to roughly 1 in 50.
The model accounts for everything shared at a location - manager, pay, local job market, working conditions. So the curve reflects engagement itself, not the difference between a well-run place and a badly-run one.
The skeptic's question: is this just a fluke at the extremes? People who never log in are different from people who use the app every day, sure - but does the relationship hold in between? Yes, monotonically. Each engagement bin shows fewer leavers than the one before it.
n = 24,900 active users. Bins chosen so the curve is monotonically decreasing. The 1-7 day baseline absorbs a small early-tenure spike.
This is the part that rules out "fluke at the extremes" as an explanation. If the relationship were just a quirk - say, people who log in once-a-quarter behaving very differently from people who log in once-a-month - we'd expect a flat middle and a sharp drop only at the ends. We don't see that. We see a clean, consistent gradient that gets steeper as engagement deepens.
Logins can mean anything: checking a shift, clocking in, glancing at the news feed. Chat messages mean your team is actually talking to each other. Asking a manager a question. Confirming a swap. Saying happy birthday. Sharing a photo of the new menu.
That's a different kind of engagement, and it shows the sharpest retention curve in our entire dataset.
n = 24,900 active users. 90-day window. Direct messages and group chat combined.
One important nuance: chat volume is not the goal. The goal is connection. Chat is just the most visible artefact of connection in a platform's data. Three messages a week between someone and their manager probably matters more than 50 between two people who'd be friends regardless.
That said, the dose-response is steep enough that volume by itself is a fair leading indicator. If you're a multi-site operator and you can see your sites' chat volumes, the low-volume sites are probably your retention risk sites - independent of pay or hours.
The data tells us engagement and retention move together. It doesn't tell us why. Here's our best read on the underlying mechanisms - some of which we can already partially test, others that the next phase of the study will address.
A hospitality job pays roughly the same wherever you go. What differentiates a place worth staying at is whether you feel known. Our chat dose-response shows employees who exchange messages with their team are progressively less likely to leave. Talking to your colleagues - even about which keg to tap - means you have relationships at work. People don't quit relationships easily.
Pay matters; it's a hygiene factor. But pay alone explains roughly nothing in our data once we control for location. Two employees at the same site, doing the same job for the same wage, have very different likelihoods of leaving based on how engaged they are.
When new starters don't know what's happening, who to ask, or what's expected, they fill the gap with assumptions. Most of those assumptions are negative. A platform that surfaces what's going on at the company - new openings, recognition posts, the team's win this week - reduces that gap.
People who feel informed feel included. People who feel included stay. The feed in our data correlates with retention almost as strongly as chat does, even though it's mostly passive consumption.
Up to 22% of all hospitality turnover happens in the first six weeks. That's the window where engagement either takes hold or doesn't. Structured preboarding (sending content before day one) is associated with roughly 40% better 90-day retention. A platform that gets used during onboarding builds the habit that compounds across the rest of the employment.
Read more on this dynamic in our pieces on preboarding and structured onboarding - both directly address the first-45-days problem.
A "great shift today" message in a public team chat costs the manager 5 seconds. To the person who got it, it's a memory they take home. Multiply that across thousands of small interactions across thousands of locations, and you get the dose-response we observed: each step deeper in chat engagement correlates with measurably fewer departures.
Recognition isn't an annual awards ceremony. It's the constant micro-feedback that tells people their work is seen. A workplace platform makes that micro-feedback visible, fast, and shareable.
Beautiful platforms with low adoption don't move retention. Average-looking platforms with high adoption do. Our entire dataset is on engaged users only - employees who at least open the app. The retention gap we measure is the difference between using a tool a little and using it a lot.
The product that gets opened four times a day is the one that wins the retention race. That's why the things people open it for - chat, feed, schedule, pay info - matter more than the things they don't (annual review modules, quarterly survey features). Hospitality teams need tools that disappear into the workflow, not new workflows themselves.
If you read this report and do nothing, the data was for entertainment. Here are the five most actionable things we'd recommend, ordered by leverage.
Preboarding cuts 90-day churn by ~40%. Send the welcome content, the role explainer, and a "meet the team" message before they walk in. Day one is too late - by then they've already formed a first impression of how organised you are.
How to do it well: our preboarding guide walks through what to send and when.
People talk to people they actually work with. Auto-syncing chat channels from your rota does the configuration work for you - new hires arrive into the right group, leavers fall out automatically, and managers don't have to maintain chat groups manually.
Generic group chats with 200 people in them go quiet. Site-specific chats with 12 colleagues hum.
Posts about a specific site's wins, hires, and celebrations get more engagement than HQ-level updates. Train managers on what good looks like: short posts, photos, names. Send weekly prompts to make this a habit.
One realistic target: each site posts something three times a week. That's 12 posts per month per site. Multiply by your number of sites and your feed becomes a place worth opening daily.
For thirteen more concrete tactics that hospitality operators use to keep staff longer, see our piece on retention secrets that top restaurants and bars swear by.
The gap between light and deep platform engagement, scaled to representative customer sizes:
| Operator size | Fewer leavers / year | Annual savings range |
|---|---|---|
| Small (50 users) | ~5 | $15,000 - $25,000 |
| Mid-size (150 users) | ~15 | $45,000 - $75,000 |
| Large (500 users) | ~50 | $150,000 - $250,000 |
To put that in concrete terms: a 150-user operator that closes the engagement gap could fund a full additional manager, a whole quarter's marketing budget, or a meaningful pay rise across the team - just from money that's currently being spent on replacing staff who didn't have to leave.
Full methodology, statistical tables, complete ROI breakdown, and early access to our 2026 longitudinal results. Delivered to your inbox.
The most-cited figure is 70-80% annual turnover, compared to 12-15% across most other industries. That number is heavily US-weighted and QSR-weighted (Bureau of Labor Statistics, National Restaurant Association). UK hospitality typically runs closer to 30-40% annual; Nordic hospitality lower still. Premium and mid-market operators run lower than fast-food chains. So "the industry average" depends a lot on geography and operator type. See the next FAQ for why our dataset's numbers look so much lower.
Three reasons stacked, and we want to be upfront about all of them.
One: selection. This study only includes employees who logged into the platform at least once in the quarter. We exclude zero-login accounts deliberately - if someone hasn't opened the app in 90 days, they've most likely already left, and including them would inflate the leaver counts in a way that doesn't tell us anything about engagement. That filter removes the highest-churn segment of the workforce, which keeps our denominators cleaner but means our headline turnover rates aren't directly comparable to industry totals.
Two: who's in our sample. Our 137 customers are predominantly UK and Nordic hospitality operators, skewing toward mid-market and premium brands - Pizza Pilgrims, Gail's, Dishoom, Honest Burgers, Wahaca, Ottolenghi, Farmer J, and similar. UK hospitality runs at roughly 30-40% annual turnover; Nordic markets are lower again; premium operators run below QSR averages. The "70-80%" figure is heavily weighted toward US fast-food chains, which is a structurally different workforce.
Three: how we count. Industry statistics often use total separations divided by average headcount across the year, which can produce numbers above 100% when workforces churn quickly. We use confirmed system-event leavers (admin deactivation, platform removal, HR system termination) divided by activated employees, annualised linearly. Cleaner but more conservative.
The bottom line: the 52% gap and 23%-per-10-login-days finding are relative effects measured within our sample. Same employees, same locations, same conditions - just different engagement levels. Those internal comparisons are what the methodology is designed to be valid for, and they hold up under every control we add. The absolute turnover percentages are not meant to be a benchmark against US industry totals.
The Cornell Center for Hospitality Research puts the cost at $3,000 to $5,000 per frontline hire when you include recruitment, onboarding, trainer wages, and lost productivity. For managers the figure rises to roughly $13,000-$14,000. Some studies put total turnover cost at 30-150% of annual salary depending on role complexity. Most operators underestimate the real cost because they don't account for the productivity dip during the new hire's ramp.
Engagement with a platform correlates strongly with retention. In our study of 24,900 employees across 137 hospitality operators, the most engaged third leave at 2.3% per quarter versus 4.8% for the least engaged - a 52% gap. The effect remains statistically significant (p < 0.0001) after controlling for location, manager, pay, and tenure. Whether this is a causal effect of the app or a marker of underlying engagement is something we're testing with longitudinal data through 2026, but the directional evidence is consistent and robust.
Logins can mean almost anything: checking a rota, glancing at a notification, clocking in. Chat messages are evidence of a relationship. They mean the employee has someone at work to talk to, asks questions of, gets recognised by, or simply hangs out with digitally. Relationships create reasons to stay that pay alone cannot create. The chat dose-response in our data is the steepest curve we see, dropping from 0% (non-chatters) to -55% (heavy chatters).
The first 90 days are when adoption habits get established. Most operators see meaningful engagement metrics inside 60 days of launch. Retention impact follows on a slight lag because turnover is measured in quarters, not weeks. A reasonable expectation: visible engagement gains in quarter 1, measurable retention shift in quarters 2-4, full ROI realisation in year one.
Sources. All Gravy Retention Impact Report (Q1 2026, n = 24,900 active employees, 137 paying customers, 1,400+ locations). Cornell Center for Hospitality Research on cost-per-hire. Industry benchmarks from BLS, NRA, DailyPay, and SHRM.
Methodology. Logistic regression on quarterly leaver outcome with location fixed effects and tenure covariate. p < 0.0001 across all model specifications.
© 2026 All Gravy. UK and Nordic hospitality data.