124apps All articles
Opinion & Curation

The Stowaways in the Share Button: A 25-App Audit of Links, Tags and Tag-Alongs

124apps
The Stowaways in the Share Button: A 25-App Audit of Links, Tags and Tag-Alongs

Sharing a link should be one of the simplest things a phone does. You find a thing, you press the little box with the arrow, you pick a person. Done. But over the past fortnight I've been copying the same handful of links out of 25 everyday UK apps and pasting them into a notes file, and the results are quietly fascinating. Pressing share, it turns out, is a bit like handing your car keys to a valet you've never met. Some bring it back spotless. Others leave a receipt in the cup holder, a half-eaten pasty on the dash and a mysterious code in the glovebox that you're fairly sure wasn't yours.

This isn't a scandal. Nobody's being defrauded by a ?utm_source= at the end of a Rightmove listing. But if you share links a lot — in group chats, on the family WhatsApp, in a work Slack — the mess adds up, and occasionally it does something genuinely annoying.

How the test worked

Two phones, one iPhone running the current iOS and one mid-range Android, both on UK networks. For each of the 25 apps I did the same thing: opened a typical piece of content (a news story, a product page, a listing, a track, a train journey), hit share, and sent the link to WhatsApp, to Messages and to email — then copied it out and read it character by character in a notes app.

I was looking for four things: length, tracking parameters, whether the link pointed straight at the destination or at a redirect in between, and what the preview looked like on the other end. I also noted, separately, which app the share sheet offered to send to first, because that turned out to be its own small comedy.

What was actually attached

The clutter fell into three families.

The trackers. Long strings beginning utm_source, utm_medium, utm_campaign, plus the ad-network fingerprints — fbclid, gclid, and the short si= that Spotify hands out with everything. These don't break anything. They just follow you around and make a neat 60-character address into a 240-character monster.

The referral tags. Retail and resale were the worst offenders. Marketplace listings, second-hand clothing, a couple of DIY and homeware shops — several arrived with a friend-code or affiliate parameter welded on, presumably so the app could claim credit if a stranger bought something. Fair enough commercially; slightly odd when you're just sending your sister a link to a shed.

The wrappers. Some apps don't give you the address at all — they give you a short link on their own domain, which redirects through two more before landing. It works when you tap it. It is useless when you try to read it out over the phone.

Roughly nine of the 25 handed over something close to clean. Around eleven added at least one parameter we hadn't asked for. The remaining handful did something more elaborate — a wrapper, a referral tag, or a link that quietly opened in the app rather than the browser for anyone who had it installed.

The clean ones were mostly the boring ones

The apps that behaved tended to be the ones with the least to gain. News and radio apps were generally tidy — a plain address, a sensible preview, no tail. Government and utility apps were similarly unfussy, which I suppose is what happens when nobody's paid to think about growth. A couple of the transport apps handed over clean links too, though one insisted on routing through its own shortener, which at least resolved reliably.

Broadcasters were a mixed bag. Two gave me clean addresses; one attached a campaign parameter that made the link twice as long as the article it pointed to.

The read-it-aloud test

Here's a test that doesn't need any technical knowledge: try reading the link out loud to somebody. I did this with my mum, on a Sunday, over a cup of tea.

The clean links took about eight seconds: 'bbc dot co dot uk, slash news, slash…' Fine. The wrapped ones took half a minute of spelling out random capitals, and she still had to be talked through it twice. Anything with three parameters was hopeless. If a link can't survive being spoken, it's not really a link anymore — it's a barcode.

Previews the size of a billboard

The other half of the story is the preview. Several apps attached a thumbnail so large that in WhatsApp it filled the entire screen before a single word of my message appeared. One screenshot share from a shopping app arrived with a promotional banner, a price, a 'shop now' button and my own profile picture attached — I'd sent a picture of a kettle and accidentally sent a small advert as well.

On a decent connection this is harmless. On one bar of signal on a train through Cumbria, it's the difference between a message that sends and a message that sits there spinning while the guard comes round.

The default recipient problem

Less discussed, and more likely to actually embarrass you: the share sheet remembers. Both platforms now put a row of suggested people at the top, based on who you've shared with most. The result is that the first button on the screen is often a specific human being, not a neutral destination. I very nearly sent a work document to a group chat called 'Sunday League' purely because the muscle memory was there.

It's a genuinely useful feature that's one inattentive tap away from being a genuinely awkward one. If you've ever wondered why somebody sent you a link with no context at all, this is probably why.

What you can do about it

The good news is that stripping the luggage takes seconds, and you don't have to give up the share button.

The verdict

None of this is a reason to stop sharing. The share button is one of the genuinely good ideas in modern software, and most of the 25 apps we tested are only mildly guilty of over-packing.

But it is a reason to glance at the tail before you press send. A link is a piece of writing, and the way it arrives says something about the app that sent it. Nine of our 25 understood that. The rest could stand to travel a bit lighter.

All Articles

Related Articles

The Lock Screen Shopfront: Thirty UK Apps, Nine Days of Push and the Opt-Outs That Didn’t Stick

The Lock Screen Shopfront: Thirty UK Apps, Nine Days of Push and the Opt-Outs That Didn’t Stick

Minor Improvements, Major Headache: Six Months Inside 25 UK App Updates

Minor Improvements, Major Headache: Six Months Inside 25 UK App Updates

Lost Phone, Cold Sweat: A Two-Factor Recovery Audit of Twenty UK Apps

Lost Phone, Cold Sweat: A Two-Factor Recovery Audit of Twenty UK Apps