The situation
festivalnetworks.com started out selling connectivity. Mobile internet kit rented to festival organisers so their crew could actually work on site when the public networks fell over at 9pm on a Friday. Good business. Steady business.
Phil Davidson, the founder, could see a bigger gap. The public going to festivals was being underserved in a completely different way. They did not need a better SIM. They needed one place to find the festival they wanted, sort out who they were going with, and deal with the travel and accommodation around it. Right now that information sits across ten tabs and a WhatsApp group, and nobody had pulled it together.
Phil wanted to pivot the company into being exactly that hub. The idea was the easy bit. Building something that could actually do it was the hard bit.
What we did
We started with the business, not the technology. A lot of whiteboard time with Phil, trying to work out what festivalnetworks was really meant to become. Who was going to use it, what had to be in place on day one, where he was willing to compromise and where he flatly was not.
Then we went off and did the homework. Existing festival listing sites, community platforms that had worked elsewhere and others that had not, booking flows that did the job and booking flows that actively got in the way. We came back with a plan Phil could sign off on without having to read a ninety-page specification.
And then we built the thing. festivalnetworks.com as it runs today was built from scratch by us, with Phil alongside throughout. Content discovery, community features, travel booking, and the integrations behind it that keep the whole thing stitched together. Where AI genuinely does something plain software cannot, it is in there. Where it is not needed, we did not put it in just because we could.
The result
The site is live and growing. A business that eighteen months ago was renting mobile SIMs at festivals is now the festival community platform, and the software that made that pivot possible is the software we built.
What Phil told us when we were done was that we had taken his business somewhere he could not have got it on his own. That is the bit we care about.
Frequently asked questions
How long did the platform pivot take from idea to launch?
From the first whiteboard session with Phil to the live festivalnetworks.com running in front of real users took several months of focused work. A lot of that time was research and product design rather than code. We did not start building the platform until Phil could sign off on a plan in plain English, because the cost of getting the shape wrong on a pivot like this is much higher than the cost of a fortnight of homework.
Where does AI actually do something in the platform, and where does it not?
AI is in the platform where it earns its keep, mostly around content discovery, surfacing the right festival to the right user, and removing repetitive admin from the back office. We deliberately left it out of places where plain software does the job better, like booking flows and payment handling. The brief was to use AI where it genuinely adds value, not as a marketing label, and that line has been kept.
Could a similar pivot work for a business that is not in the events space?
Yes. The pattern is what matters. Phil had a steady core business, a clear view of an underserved customer, and the courage to commit to building for that customer rather than defending the existing one. We see the same shape in family businesses across construction, manufacturing and professional services. The technology question is the easier half. The harder half is being honest about what the business is becoming and why.
How do you charge for a build like this?
Engagements like this are scoped per project rather than priced from a brochure. The variables are the size of the build, how much existing data and infrastructure we can reuse, and whether you want ongoing care after launch. We always start with a free AI Opportunity Report or a short scoping call so the cost conversation happens with real numbers rather than guesses, and you can decide whether to commit at all.