UX friction is anything in a page or flow that slows people down, confuses them, or makes them work harder than they expected on the way to a goal, such as extra form fields, forced account creation, or unclear next steps. In conversion work, friction is a cost the visitor pays, and reducing the unnecessary part of it is one of the most reliable ways to raise a conversion rate.
Friction is not the same as anxiety. Friction is effort; anxiety is worry (about price, privacy, or whether the product works).
A page can be easy to use and still fail because visitors don't trust it, which is where trust signals come in.
Friction in UX: where it comes from
MarketingExperiments, the research arm of MECLABS, defines friction as "the psychological resistance to a given element in the sales process" and splits it into two parts: length and difficulty.
- Length: the number of fields, steps, clicks, or screens between wanting something and getting it.
- Difficulty: how hard each step feels, such as confusing labels, a disruptive eye path, slow loading, error messages, or asking for information people don't have at hand.
The two don't always move together. In one MarketingExperiments test, a shorter one-page accordion checkout converted 29% worse than the longer three-step version, because the new layout felt harder to follow. Fewer steps is not automatically less friction.
Conversion friction in A/B testing
In the MECLABS conversion sequence heuristic, written as C = 4m + 3v + 2(i − f) − 2a, friction (f) is a cost that carries a minus sign and offsets any incentive (i).
MarketingExperiments describes it as the "aggravation factor" and stresses that you can only remove unnecessary friction: every conversion asks the visitor to do something.
The formula is a thinking tool, not a measured model, so the weights shouldn't be read as real effect sizes.
Common friction tests in our database include:
- Removing, merging, or deferring form fields.
- Hiding distracting inputs such as coupon boxes in the cart.
- Offering guest checkout or express payment buttons.
- Giving people a way around a step they can't complete right now.
Examples from real tests
VWO reports that Qualicorp removed repeated form fields from a quote form and got +19.31% completed quote requests.
Tinkoff Bank added a "Don't remember passport details?" option so applicants could skip a hard step and finish later, which produced +35.8% applications.
That is a difficulty fix, not a length fix.
In ecommerce, Sq1 removed the coupon and gift card boxes from the Bionic Gloves cart and saw +24.7% revenue.
A visible coupon box sends some shoppers off to search for codes, which adds an extra task in the middle of checkout.
Friction tests also lose. MarketingSherpa documented a lead form where adding a guide choice plus one extra field caused a significant drop in leads.
Some changes barely move: GANNI's test that made the guest checkout option clearer came back inconclusive, and DRIP Agency's removal of optional checkout fields at Koro gave a modest +0.9% revenue per user.
DRIP AgencyWhen friction is worth keeping
Some friction does a job. A qualifying question on a lead form can cut junk leads, and a confirmation step can prevent costly mistakes.
If you remove fields, judge the result on lead quality or revenue, not just form submissions.
Related terms
- Information scent: weak scent adds friction because people have to hunt for the next step.
- Trust signals: these address anxiety, the cost that often gets confused with friction.
- For deeper guides, see form field tests and checkout flow optimization.