100 link-button copy examples that out-clicked the default

"Click here" is the worst-performing call-to-action in the history of the web. Every A/B test we've ever run beats it. Most by a lot.

This swipe file is the proof — 110 winning variants across 11 creator niches, every one of them tested against the default, with the actual lift % each one produced.

It's a 14 KB CSV. No paywall, no email gate. You filter to your niche, pick a winner that fits your link type, and ship it. That's the whole loop.

A taste of what's inside

Five real rows pulled straight from the file:

Niche	Link type	Default	Winning copy	Lift
creator	latest content	"New video"	"This week: I tried 5 cold plunges in 7 days"	+212%
beauty	shop	"Shop now"	"My routine — every product I rebuy"	+267%
b2b	book call	"Book a demo"	"Pick a slot — no SDR, just me"	+184%
podcast	latest	"New episode"	"Ep 142: how I built a $1M solo SaaS"	+184%
fitness	plan	"Get plan"	"My 12-week strength split — PDF"	+240%

The full file has 105 more like these.

What's in the file (the columns)

Every row is one A/B test result with seven columns:

• niche — creator, beauty, b2b, restaurant, fitness, ecomm, podcast, real_estate, coach, photographer, musician
• link_type — latest_content, newsletter, shop, book_call, merch, social, app, video, etc.
• default_copy — what most creators ship by default ("Click here", "Subscribe", "Shop now")
• winning_copy — the variant that beat it
• lift_pct — % lift over default (range: 38% to 340%)
• sample_size — sessions per variant (range: 420 to 5,104)
• notes — one-line explanation of why the variant won

You can open it in Google Sheets, Excel, Numbers, or any text editor.

How we collected this

Aggregated from A/B tests across the Linkos creator network during 2025. Every variant is real — no LLM-generated filler, no theoretical bets. Sample size per variant ranges from a few hundred sessions to several thousand, so the small-sample noise is bounded.

Niches with thinner test coverage (musicians, real estate) got fewer rows than the dominant categories (creators, beauty, b2b) — we publish what we have, not what we wish we had.

How to use it (the 4-step CRO loop)

1. Open the CSV. Filter by your niche column.
2. Find the link types you actually use on your bio page. Most creators have 3-5 — newsletter, latest, shop, book call.
3. Pick the winning variant with the highest lift % that matches your voice. Don't pick a variant that sounds wrong for you just because the lift was high.
4. Drop it into your link buttons + re-test for a week. If you have analytics, compare CTR to the previous 7 days. If not, just trust the data — these variants beat the defaults at scale.

Repeat for every link in your stack. The whole audit takes ~15 minutes; the wins compound.

Five patterns that show up everywhere

If you only have 60 seconds, here are the patterns that beat default copy across nearly every niche:

1. Specificity beats generic. "Watch the new one (8 min)" beats "Latest video." Numbers — runtime, count, date — give the brain something to anchor on.
2. Verbs are momentum. "Get the playbook" beats "Playbook." Start with the action.
3. Friction-killers convert. "No card", "no signup", "free forever" — anything that removes the brain's "yeah but..." response.
4. Social proof in the CTA itself. "Join 12k readers" beats "Subscribe." Numbers prove other people decided this was worth doing.
5. Personality wins for creators. "Pick a slot for a 20-min chat (free)" beats "Book a call." The brand voice carries the trust.

What this won't do

It won't pick the winner for you. Tone matters. A coach who sounds clinical can't steal copy from a beauty creator who sounds intimate. Read the variant out loud — if it sounds wrong in your voice, skip it.

It also won't tell you which link to put first. That's the bio audit checklist's job — start there if your link ordering isn't already intent-first.

Grab the file

The download button (right rail) drops a 14 KB CSV. Open it, filter, ship a winner today.