Cutting Edge is an established salon whose website was not doing any of the work a website should do. I rebuilt it, I still maintain it, and it is live. You can go and use it as a customer would in the next minute. This is the whole build written up: what I found, what I built, what I deliberately left out, and the one thing I would do differently.
The old home page opened with a banner, then dropped straight into two paragraphs of small justified text about the salon’s philosophy. The clearest thing it asked you to do was download a document. There was no obvious way to book, no opening hours in view, and nothing at all that answered the questions people actually ring up to ask.
The navigation carried seven items and not one of them said Book. The whole thing was laid out for a desktop monitor, and almost nobody looking for a hairdresser is sitting at one.
None of that is a criticism of the salon. They are excellent at cutting hair. Nobody in that building was ever going to notice that their front door had quietly stopped opening.
I cannot tell you. Nobody was counting, and I am not going to invent a number for a case study.
What I can tell you is where the cost lands in a salon. Every question the site fails to answer arrives instead as a phone call, and that call arrives while somebody is mid-colour with both hands full. Opening hours. Whether they do balayage. Whether a bride can be seen on a Sunday. Whether you need a patch test first, and how far ahead. All of it comes down the phone, one call at a time, into a room where the only person who can answer it is holding a brush.
That is the real bill. It is paid in interruptions and in the calls that ring out while everyone is busy, and it never appears on any invoice, which is exactly why it runs for years without anyone deciding to fix it.
The salon did not have a marketing problem. It had a front door that answered no questions and offered no way in.
So the brief I set myself was narrow. Build the pages people were already ringing up to ask about, put the way in on every screen, and stop there.
That meant deciding what the site was actually for. It is not a shop and it is not a booking engine. It is the thing that answers the question and then hands you the fastest possible route to a human. Everything below follows from that one decision, including the parts I refused.
Three things were on the table. I turned down all three, and I would turn them down again.
A salon diary lives with the stylists and changes by the hour. A booking widget that cannot see the real diary sells slots the salon cannot honour, and every one of those becomes a phone call and an apology.
Colour work needs a patch test booked days ahead. A consultation often decides what the appointment even is. The salon says it plainly on their own booking page: they would rather do it by phone. I agreed with them, and built for how they work instead of against it.
A CMS would have meant a login, a password reset, an update treadmill and a plugin surface, all so that the owner could change her opening hours perhaps four times a year.
She did not want a publishing tool. She wanted her hours to be right. Handing somebody a system they will use twice a year is how you guarantee it is broken and out of date by the third time.
I build assistants for a living, and I did not put one on this site. At the time there was nothing for it to do that the Book Now button at the top of every screen did not already do faster and better.
Fitting one anyway would have been me buying a portfolio piece with somebody else’s money. If they ever want one, the honest trigger for it is missed calls, not novelty.
Hand-written HTML, CSS and JavaScript. No framework, no page builder, no CMS, no plugins, no database and no login. It is a folder of static files on a server, which means there is no update treadmill, nothing the salon has to sign in to, and nothing that breaks because a plugin vendor the salon has never heard of shipped a release on a Tuesday.
Three things on it are not mine: the web fonts, one animation library, and the map embedded on the home and contact pages. Each of those is a request to somebody else’s server. Today I would self-host the fonts and the library and swap the embedded map for a static image and a link, which would leave the site calling out to nobody. It is a small thing, and it is the sort of small thing a page like this normally leaves out.
The pages cover what people ring up to ask. What the salon does. What the colour and treatment work involves. Bridal. The gallery. The reviews. Where it is, when it is open, and how to get in. And an FAQ carrying the two policies that cause the most awkward conversations in any salon: cancellations, and the patch test rule before colour.
The navigation came down from seven items to six, plus a Book Now button that was not there before. Every one of the six is a page somebody has rung up about. And the clearest action on the home page is now booking an appointment, where it used to be a download.
Nothing about how the salon works had to change. No new login, no new system, no training day, no migration weekend. The site went live, the phone still rang, and the difference was that the calls arriving had already had their first three questions answered somewhere else.
Maintenance runs through me. A new service, new photographs, a change to the hours: they message me, I edit it, it is live. That is a deliberate trade and it is worth naming out loud. It weighs a tool I would have had to teach them against a message they were going to send me anyway.
What they hold is a folder of plain HTML and CSS. There is no licence, no subscription and no builder account standing between the salon and its own website. If they ever want to take it elsewhere, or hand it to somebody else, they can, and nothing of mine has to be unpicked first. That is the standing arrangement on everything I build, and on a five-year view it is worth more than any feature I could have added instead.
This is where case studies usually start lying, so here is exactly what I know and exactly what I do not.
The site is live. The salon uses it. Their own footer credits me, so you can confirm this is my work in about thirty seconds without asking me for anything.
Beyond that, no baseline existed. Nobody was counting enquiries before the rebuild, and I did not install anything to count them with afterwards. That one is mine, not theirs. Measurement now goes in before a build rather than after it, and of everything on this page that is the change this project actually made to how I work.
Their Google rating and the reviews behind it are theirs. They were earned in a chair, over years, by people I have never met. I built the page that displays them. I did not build the reputation, and I am not going to drop it into my own proof slot and let you read it as mine.
Awaiting sign-off from the client
The owner’s own words, her name, her photograph and any numbers she has go in this space once she has read them back and signed them off. Until then it stays empty on purpose.
I am not writing a quote on her behalf and calling it a testimonial, and a figure I cannot show you the source of is worth less than an honest blank. In the meantime the site is live and the salon is real. Go and look at both.
I built them a front door and left everything behind it manual.
If I took the same job today, missed-call recovery would go in on day one. A call that rings out gets a text back inside a minute, carrying the salon’s name and a promise to ring back, before the caller has finished dialling the next salon on their list. It is a couple of days of work and it catches the exact thing this business loses: the person who could not get through at half past ten on a Saturday.
That is not a criticism of the website. It is an admission that a website was only ever half the job, and that I sold half a job. It is also the reason every other page on this site leads with systems and automation, and puts websites after them.
If your front door answers no questions, and every one of them arrives as a phone call while you are mid-job, that is the same problem in a different trade. Twenty minutes is enough to work out which half to fix first.
20 minutes. You leave with a clear view of what to build first, whether you hire me or not.
Ask ODUEX
Daniel’s assistant · answers, then routes
I’m the assistant Daniel built for ODUEX. Ask what he builds, how the work runs, or tell me what is going wrong in your business and I’ll tell you where he’d start.
Daniel’s assistant, not Daniel. It can be wrong, and anything that matters he confirms himself. Leave an email and the whole conversation goes to him.