Atlas FM is one of the UK's largest facilities management companies. They employ around 15,000 people -- cleaners, security staff, maintenance engineers, and support workers -- spread across thousands of sites nationwide. Their challenge wasn't technical complexity. It was human connection.
This is the story of how we built an employee communications platform that connected a workforce that had never been connected before.
The Challenge
Imagine you're a cleaner working the early morning shift at an office building in Manchester. You've worked for Atlas FM for three years. You're good at your job. But you've never met anyone from head office. You don't have a company email address. You don't have a desk. You might not even have regular access to a computer.
Now multiply that by 15,000.
That was Atlas FM's reality. The vast majority of their workforce -- the frontline people who actually deliver the service -- were essentially invisible to the organization. Communication was one-directional at best: notices pinned to break room walls, messages passed down through site supervisors, occasional newsletters that may or may not reach the people they were intended for.
There was no way for a cleaner in Manchester to know what was happening at head office in London. No way for a security guard in Birmingham to share an idea with the operations team. No way for the CEO to speak directly to the people who represent the company to its clients every single day.
The result was a disconnect that affected everything: engagement, retention, culture, and ultimately service quality. People who feel invisible tend to disengage. And in facilities management, disengaged staff means lower standards, higher turnover, and clients who start looking elsewhere.
Our Approach
When Atlas FM came to us, the brief was straightforward on paper: build an app that lets us communicate with our frontline workforce. But the real challenge was much more nuanced than that.
Discovery Workshops
We started with extensive discovery. Not just with the leadership team -- with the actual frontline workers. We visited sites across the UK, talked to cleaners, security staff, and maintenance engineers, and asked a simple question: what would make you feel more connected to Atlas FM?
The answers were revealing. People didn't want another corporate intranet. They didn't want a tool that felt like work. What they wanted was surprisingly human: they wanted to know what was happening in the company. They wanted to share things they were proud of. They wanted to feel like they belonged to something bigger than their individual site.
That insight changed everything about how we approached the build.
Community Before Functionality
Based on what we heard, we made a deliberate decision to prioritize community features over business functionality. The conventional approach would have been to build a top-down communications tool -- announcements from leadership, policy documents, HR forms. Important, but not engaging.
Instead, we designed the app around a social feed. Think of it as a private social network for Atlas FM employees. People could post updates, share photos of their work, celebrate team achievements, and interact with colleagues across the country. The business communications were there too -- but they lived alongside the community content rather than replacing it.
This wasn't just a design choice. It was a strategic bet: if we could get people to open the app because they genuinely wanted to, the business communications would reach them naturally. If we built a boring corporate tool, adoption would be a constant battle.
Progressive Rollout
You don't launch an app to 15,000 people on day one. We rolled it out in carefully managed phases:
- Phase 1: Pilot with 500 employees across 20 sites. Intensive feedback gathering, rapid iteration on UX issues, and validation of the community-first approach.
- Phase 2: Expansion to 3,000 employees. Regional rollout with dedicated champions at each site who could help colleagues get set up and demonstrate the value.
- Phase 3: Full rollout to all 15,000 employees. By this point, word of mouth was doing most of the heavy lifting -- people were asking for access before we reached their site.
Each phase included training materials, support resources, and a feedback loop that fed directly into the next sprint's priorities. The app that launched in Phase 3 was significantly better than the one we piloted in Phase 1, because we'd learned from thousands of real users along the way.
What We Built
The finished platform combined social engagement with practical business functionality:
Social Feed
The heart of the app. A scrollable feed where any employee could post updates, photos, and achievements. Moderated but not censored -- we worked with Atlas FM to create guidelines that encouraged authentic communication without creating risk. The feed included likes, comments, and sharing, making it feel familiar to anyone who'd used social media.
Business System Links
Integrated access to the systems employees actually needed: shift schedules, payslips, HR requests, training modules, and policy documents. Rather than building these from scratch, we connected to Atlas FM's existing systems via APIs, presenting them through a unified, mobile-friendly interface.
Push Notifications
Targeted notifications that could reach specific groups -- all security staff, all employees in a particular region, or the entire workforce. This replaced the old system of cascade communications through site supervisors, giving leadership a direct line to any segment of the workforce in seconds.
Team and Site Content
Dedicated spaces for individual sites and teams. A cleaning team at a hospital could have their own channel where they shared updates, coordinated shift swaps, and celebrated good inspection results. Regional managers could broadcast to their sites without flooding the national feed.
The Impact
The numbers told one story: adoption rates that exceeded every target Atlas FM had set, engagement metrics that rivaled consumer social apps, and a dramatic reduction in the communication lag between head office and frontline staff.
But the real impact was harder to quantify. For the first time, Atlas FM's leadership could hear directly from the people delivering their service. A cleaner could post a photo of a particularly well-maintained space and receive recognition from the CEO. A maintenance engineer could flag a safety concern and know it would reach the right person within hours, not weeks.
The platform began evolving into something broader than a communications app. Atlas FM started using it as the foundation for a wider digital workplace -- a single entry point for everything an employee needed, from checking their next shift to accessing training materials to celebrating a colleague's work anniversary.
The feedback from frontline staff was consistently positive. People felt seen. They felt connected. And that translated into measurable improvements in engagement scores and retention rates.
Tech Stack
The technology choices were driven by the constraints of the audience:
- Progressive Web App (PWA) -- Cross-platform by default. Frontline workers use a mix of Android and iOS devices, often older models with limited storage. A PWA gave us native-like performance without requiring an app store download or consuming significant device storage.
- Vue.js -- Frontend framework. Fast, lightweight, and well-suited to the kind of real-time feed experience we needed to build.
- Node.js -- Backend services. Event-driven architecture to handle the real-time aspects of the social feed and notification system.
- Cloud-hosted, multi-region -- Deployed across multiple UK regions for low latency regardless of where employees were located. Auto-scaling to handle the spikes that came with company-wide announcements.
- Integration APIs -- RESTful APIs connecting to Atlas FM's existing HR, scheduling, and payroll systems. We built an integration layer that abstracted the complexity of their legacy systems behind a clean, modern API.
Key Takeaway
The most important lesson from Atlas FM is one that applies to any enterprise software project: design for the actual end users, not the procurement team.
If we'd built what the brief asked for -- a corporate communications tool -- we would have delivered something that ticked every box on the requirements document and sat unused on 15,000 phones. Instead, we spent time understanding what the real users actually wanted, and we built something they chose to open every day.
The features that drove adoption weren't the business systems or the policy documents. They were the social feed, the ability to share a photo of good work, the feeling of being part of something. The business functionality rode on the back of genuine engagement -- not the other way around.
This is a pattern we see repeatedly in enterprise software: the projects that succeed are the ones that treat end users as the primary audience, not an afterthought. It sounds obvious. In practice, it's surprisingly rare.