From vision to reality: What two years developing our own tech taught me
With margins under pressure and operational costs continuing to rise, recruitment businesses are understandably looking for smarter ways to work. AI and automation are opening up possibilities that simply weren’t available a few years ago, and increasingly, businesses are looking beyond off-the shelf software towards technology built around their own processes and challenges.
Over the last couple of years, I’ve been on that journey myself.
At Liquid Friday, we’ve invested heavily in developing our own technology, including Stride, our platform that brings together contractor onboarding, compliance, Right to Work checks, payroll visibility and workforce management in one place.
The ambition was to solve some very real problems for our business and the recruitment agencies we work with. To remove necessary admin, speed up processes and give people better visibility and control.
It’s been fascinating, frustrating, rewarding and at times, all-consuming. I’ve learnt an enormous amount along the way, often the hard way.
For anyone considering building their own technology, or adapting existing AI tools to fit their business, these are the lessons I’d share.
1. Before you build anything, make sure it doesn’t exist already
This is probably the most important lesson of all. Don’t build technology for the sake of building technology.
It’s easy to get caught up in the excitement of creating something that’s yours. But you need to ask yourself a fairly fundamental question: am I building this because nothing else can do what I need, or because I want to be able to say we’ve built it ourselves?
If something already exists that does the job, use it. If you need to get 12 people from London to Manchester, hire a minibus. Don’t spend six months in your garage trying to build one!
Because building the technology is only the beginning. You’ve then got maintenance, ongoing development, big fixing, troubleshooting, penetration testing, security, regulatory requirements and everything else that comes with owning a system.
All of that takes time, money and resources, long after the initial excitement of launching something new has worn off.
There’s absolutely a place for bespoke development, we’ve found that ourselves. But it needs to solve a problem that existing technology genuinely can’t, or deliver something substantially better for your business. Otherwise, you risk creating a very expensive distraction.
2. It’s always bigger than you think
You start with what seems like a relatively straightforward idea. We’ve got this problem. If we build something that does this particular thing, we’ll solve it. Except it’s almost never that simple.
One of the biggest things I’ve learnt is how interconnected business processes really are. You might be doing something to improve onboarding, but that process touches compliance, payroll, operations, customer service and potentially several other parts of the business.
And because businesses naturally have specialists in different areas, nobody necessarily has complete visibility of how everything fits together.
We experienced this first-hand with our own technology. Through the development of Stride and the improvements we’ve made to our registration process, we’ve reduced the time it takes a worker to register from 20 minutes to just five. That’s four times as fast, and a significant efficiency gain when you’re dealing with thousands of contractors.
But improving one part of the process had an unexpected knock-on effect.
By significantly speeding up registration, we also massively increased the volume of information flowing through our onboarding and compliance processes. Suddenly, other parts of the business needed to scale up to handle it.
I often think of it like making a pipe bigger so more can flow through it. That’s great, but where is that pipe going, and can whatever’s at the other end cope?
You can solve one problem brilliantly, only to move the bottleneck somewhere else.
3. Don’t let your pride get in the way of getting it right
This is a difficult one, particularly when you’re personally invested in a project. There have been periods over the last couple of years when I’ve worked 90 hour weeks, with plenty of late nights, early mornings and sacrifices along the way.
When you’ve put that much time, effort and energy into creating something, you become incredibly invested in making it work.
So, when someone comes along and says, “That’s not quite right”, or “Wouldn’t it be better if it did this?”, it’s not always easy to hear. They haven’t necessarily seen the months of work, the testing, the setbacks or the 4am finishes. But they’re often exactly the people you need to listen to.
One of the most important things I’ve learnt is that you can’t let the effort you’ve already invested dictate what happens next. If you’ve spent three months developing something and it isn’t right, it isn’t right. You have to be prepared to change it.
The objective isn’t to prove that your original idea was perfect. It’s to create something that actually works and makes the business better.
4. Listen to everyone, but don’t lose sight of your vision
There’s a balance to be struck here. You absolutely need input from the people who understand your business. For us that was our payroll specialists, compliance leads, operational managers and the people who will actually use the technology every day. They have knowledge you don’t, and that knowledge is essential.
You also need to recognise where your own expertise ends and get external support where it’s needed.
But there’s something else I’ve learnt – you can’t outsource your vision.
If you’re the person driving a technology project, you need to understand enough about what you’re building to make informed decisions, challenge advice and retain control over the direction it takes.
That doesn’t mean becoming an expert developer overnight. It means investing the time to understand how things work, what the limitations are and whether the solutions being proposed are actually right for your business.
Otherwise, you’re entirely reliant on other people telling you what’s possible, what’s necessary and what things should cost. There’s also a danger in trying to incorporate every suggestion you receive.
You can start out with a vision for a thoroughbred racehorse and end up with a camel because everyone wanted to add something different!
Take the advice, listen to the feedback and be prepared to change things, but don’t lose sight of what you set out to achieve.
5. You can’t learn passion
This is probably the most personal lesson I’ve taken from the experience. Developing our technology has been one of the most challenging things I’ve done, but also one of the most interesting.
I’ve genuinely enjoyed learning how things work, understanding new concepts, solving problems and developing skills I didn’t have two years ago.
There have been plenty of frustrations, but there’s also something really satisfying about working through a difficult problem and finally getting it right.
Sometimes I’ve found myself working late into the night not simply because something needs finishing, but because I’m enjoying figuring it out.
It almost becomes a hobby! But I think that’s important. If you’re going to take on a project like this, particularly if you’re driving it yourself, you need to have the curiosity and enthusiasm for the process.
Because there will be setbacks. There will be things you don’t understand, and there will be times you wonder why you started.
If you don’t enjoy learning, experimenting and solving problems, it’s going to be a very difficult journey. If you’ve only interested in the finished product, rather than what it takes to get there, that’s absolutely fine, but maybe developing it yourself isn’t the right approach.
Find people who do enjoy that process, or find another way to get the outcome you need.
Final thoughts…
Perhaps my final thought on this is that technology should solve problems, not create them
Looking back over the past two years with what we’ve done at Liquid Friday, I wouldn’t underestimate the scale of what we’ve taken on, but I also wouldn’t underestimate what we’ve gained from it.
We’ve created technology that makes a meaningful difference to how our business operates and how recruitment agencies work with us. Just as importantly, the process has forced us to understand our operations in far greater detail, challenge established ways of working and think differently about what might be possible.
So for recruitment leaders considering AI, automation or bespoke development, my advice would be to start with the business problem, not the technology. Be absolutely clear about what you’re trying to achieve. Look hard at what’s already available. Understand the full cost and operational implications of building something yourself.
If you do decide to build, surround yourself with people who know things you don’t, listen to them and be prepared to change course.
Above all, don’t lose sight of why you started. The measure of successful technology isn’t how clever it is, or whether you built it yourself, it’s whether it makes your business better.