A Jeweler Told Me Code Never Sold a Ring
Jewelers carry schema markup on 77.4% of their sites, the strongest showing this trade puts up on any check I run. The objection I get about it, taken seriously, and the thing I think is really going on.
“A line of code has never sold a ring.”
The jeweler said it kindly, which made it harder to argue with, and he’d earned the right to say it however he liked.
Thirty one years behind the same counter. He has watched a man propose in his shop, twice. He has never once watched anyone buy anything because of a tag buried in a web page.
Neither have I.
Rings sell under a good light, on a hand, with someone’s mother standing there saying it suits her. Nothing I check on a website has ever competed with that, and nothing ever will.
So let me agree with the whole objection first. Then I’ll show you where it quietly stops being true.
The job the code is actually doing
It isn’t selling. It happens much earlier than the selling, and it’s a far smaller job than the one he was refusing to pay for.
A woman two towns over types the name of a stone and the word repair into her phone at eleven at night. She has never heard of your shop. Between her question and your front door sits a machine deciding which businesses are even part of the conversation.
That machine doesn’t read your page the way she would.
When I check a site for schema markup, it arrives on my screen as a block of curly brackets near the bottom of the page code. Your name, your street, your telephone number, your opening hours, one fact to a line. I read a few hundred of them in one week this summer, and by the second evening I’d stopped seeing the brackets at all.
None of it is written for a customer. It’s written so that the machine never has to guess.
How else would anything out there know you take repairs in, if the only place you mention it is a sentence halfway down a page about the family history?
I sampled 2,676 jewelry businesses across 24 cities for this, and audited every website they listed that would open for me. Your trade hands those facts over more reliably than almost every trade I check. 77.4% of the jewelry sites I opened carry the labels, above the 70.5% I get across everything, and this is the only place where jewelers finish that far ahead of the field.
Plastic surgeons do it better. So do vets, funeral homes, and the hotel sites I run through the same check. Accountants come dead last at it, which I enjoy more than a grown man ought to.
Now the part he didn’t enjoy
You didn’t do it.
37.6% of the jewelry sites I audited run on Shopify, a long way past the 5.5% Shopify holds across my whole sample. A store platform writes those labels for you the day you switch it on, and it carries on writing them whether anyone is watching or not.
Nobody in your shop typed a word of them.
I don’t say that to take anything away from you. I say it because a job nobody was ever handed is a job nobody ever checks, and in ten years of opening other people’s websites I have not once found the exception. Manufacturers win the title tag outright and not one owner I asked had typed theirs either, so it isn’t your trade, it’s the software.
Which is fine, right up until the facts inside them stop being true.
The hours changed after the second Christmas. The number changed when you finally gave up the landline. You moved the bench into the back room and started closing on Mondays, and you told the regulars, and the regulars told each other.
The labels went on saying the old thing. In perfect order, with total confidence, to every machine that asked.
Nobody proofreads code. It doesn’t look like writing, so it never gets read like writing, and a wrong Thursday closing time can sit in there for four years without a single person noticing it.
Have you ever seen yours? Would you know who to ask?
A woman I know drove twenty minutes on a Sunday in December to collect a repair, because everything she could find on her phone said the shop opened at noon in the fortnight before Christmas. It hadn’t opened on a Sunday in two years.
She stood in the doorway with her hands full, phoned her sister about it, and drove home.
The shop never found out that any of that happened. A fact nobody owns doesn’t fail loudly, it fails once at a time, to one person at a time, and that person has no way of telling you about it afterwards.
When did anybody last compare what your website claims your hours are against the hours you actually keep?
What the labels can’t do
There’s a second half to his objection, and honestly it’s the stronger half.
The labels can’t say anything.
A label can hold your opening hours. It can’t say that you’ll restring pearls while the customer waits, or that you turned a repair down last month because the stone wasn’t worth what the new setting would have cost, or what you charge to look at a watch before you agree to touch it.
That has to be written by a person who knows the trade, and this trade hardly writes at all.
I’ve stood in a lot of these shops now, and the answer a jeweler gives across the counter is always longer, warmer and more useful than anything I find written on the website afterwards. Which of those answers would you write down first?
So a machine knows your address perfectly and has almost nothing of yours to quote when a real question turns up. 28.4% of these sites publish anything of their own, where 37% do across my whole file.
Every one of those unwritten pages already exists out loud, in your shop, in your voice. You gave one this morning to a man who wanted to know whether his father’s watch was worth servicing, and you gave it away for nothing because that’s what decent people in this trade do.
How much did that answer cost him? How many more of him are out there tonight, asking a phone instead of asking you?
Being listed and being chosen are two separate problems, and the code only ever solves the first one. It gets your name into the room where she’s comparing four shops she has never heard of, and then it stands there with nothing to add while she reads whatever the four of you wrote about yourselves.
Most of what she finds on those four pages was written in a hurry by whoever built them, years ago, about a shop none of them had ever stood inside. Which is either terrible news or the cheapest opportunity you have had in a decade, depending on how fast you want to move on it.
Add my ten checks up and jewelers land on 73 out of 100, held up at one end by code that someone else installed on your behalf.
When I take on a jeweler’s website, the labels are the part I barely touch. The day goes on opening hours, repair prices, and the four sentences you’ve already said a thousand times at the counter. If you’d rather spend a Saturday on it yourself, I put those jobs in the order I’d do them, and the labels come last on that list too.
Go and read what your shop says about itself
Tonight, when the case is locked and the alarm is on.
Take your homepage address and drop it into Google’s own rich results test, then read the list it hands back to you. That list is what your website has been telling machines about your business, every single day, since the week it went live, and reading it once is the sort of job I keep on the list I work through before a site launches.
Check three of them against reality. The opening hours. The telephone number. The address, unit number included, if you have ever moved the shop.
I run mine through it about twice a year, and I’ve caught my own site out more than once.
I’d put money on one of the three being wrong, and on it having been wrong for a good while.
What would it be worth to you to find that out this evening, in a quiet shop with a coffee, instead of finding it out next December when a woman with her hands full is standing outside a locked door?
Then come back and tell me a line of code never mattered.
