Subtract Before You Add: What Buffer's Conversion Data Actually Proves

Buffer's Head of Growth recently shared something that most conversion teams won't tell you: the changes that moved the needle most weren't the ones that added something. They were the ones that removed something.

Cut an entire page from the signup flow. 17% lift. Single test.

Simplified the homepage. Simplified the pricing page. Reduced decisions. Won.

Replaced a visible CAPTCHA with an invisible one. Won. Changed "Get started now" to "Get started for free." Won.

Added a persistent banner. Lost. Added social proof to the signup page. Lost. Defaulted to a higher pricing tier. Lost.

The pattern is consistent enough to be a principle: removing friction consistently outperforms adding persuasion.

Why this is harder to act on than it sounds

The instinct in conversion work — and in B2B website design more broadly — is additive. More proof. More context. More reasons to trust. More CTAs. More copy that anticipates objections and handles them before they come up.

That instinct isn't irrational. It comes from a real insight: people need reasons to say yes. The problem is that it gets applied at the wrong moment.

By the time someone is in your signup flow, they've mostly decided. They are not evaluating. They are executing. Every element you add at that point is not persuasion — it's interference. The social proof, the banner, the nudge toward a higher tier — all of it is asking a person who is already moving to stop and reconsider.

Most of them don't reconsider in your favour. They just stop.

The design implication

The question most B2B website teams should be asking is not "what else can we add to this page to improve conversion?" It's "what does someone have to do, read, or decide that they shouldn't have to?"

Every step in a flow, every field in a form, every piece of copy that explains something the user didn't ask about — each one has a cost. Most teams don't measure that cost because they're focused on what the addition might gain, not what it takes from the person trying to complete the action.

Buffer's data makes the cost visible. Routing people through a pricing page before signup cost them 57% of visitors. That's not a small leak. That's the majority of their potential conversions, lost to a decision someone made about information architecture.

Subtract before you add

The principle that comes out of this — subtract before you add — is deceptively simple and operationally difficult.

Subtraction requires confidence. You have to believe that the person arriving at your page is capable of making a decision without being managed through every objection in sequence. You have to trust that the work done earlier — in the positioning, the messaging, the reputation — has done enough that the conversion moment doesn't have to carry everything.

When that upstream work is solid, simplifying the path is easy. When it isn't, teams compensate by adding — more explanation, more proof, more copy — and the funnel gets heavier without getting better.

The conversion problem is often not a conversion problem. It's a trust problem that gets diagnosed as a content problem and treated with more content.

Subtract first. See what was actually needed.