A/B test social share buttons: help, hurt, or just there?
Social share buttons are one of the most over-used and under-measured conversion elements on the web. VWO's famous Taloon.com test saw conversions rise 11.9% after removing share buttons — visible share counts near zero acted as negative social proof. The right question isn't "add them" — it's "on which pages, with which platforms, at which visibility threshold?"
This test matters more for SaaS and B2B than most teams assume. Share buttons on pricing and feature pages are pure distraction. Share buttons on blog content can drive real referral traffic — but only if configured correctly.
The hypothesis
If we remove social share buttons from decision pages (pricing, product, checkout) but keep them on content pages (blog, guides), overall signups will increase — because we cut distraction on pages where the visitor is deciding, without losing referral traffic on pages where sharing is natural.
Test setup
Control: Share buttons present on all pages (current state)
Variant A: Share buttons removed from pricing, product, comparison, and checkout pages; kept on blog and resource pages
Configuration rules:
Hide share counts when below 50 — low counts hurt trust more than help
Test 3–4 platforms (LinkedIn, X, Email, Copy Link), not 8–10 — too many causes choice paralysis
Default position: floating left rail or inline-above-content on blog posts
Primary and guardrail metrics
Primary: Signup or trial conversion rate on tested pages
Secondary: Shares per 1,000 pageviews (blog only)
Guardrail 1: Organic blog referral traffic from social
Guardrail 2: Average engagement time on blog
Guardrail 3: Scroll depth on blog pages
Sample size and duration
You need 10,000+ sessions per variant per page type to detect a 5% relative lift with statistical confidence. 4–6 weeks for a medium-traffic SaaS.
Variations to try
Platform mix: Everything vs just Copy Link + Email. B2B often wins with minimalism.
Share count visibility: Always hidden vs only shown if above 50
Position: Top vs floating vs end-of-post
Style: Icon-only vs icon + "Share" label
Trigger: Always visible vs appearing after 30% scroll depth — less visual noise, same share rate
Common mistakes
Showing share counts below 50. Looks like nobody cares — negative social proof.
Including 8+ platforms. Choice paralysis reduces shares, not expands them.
Putting share buttons on pricing or checkout pages. Pure distraction from the primary CTA.
Not measuring the full funnel. Shares are usually near zero anyway; what you care about is whether removing them lifts conversion.
Using 3rd-party scripts (AddThis, ShareThis) that add 500ms+ to page load. The speed cost alone can outweigh any sharing gain.
