Strategic Friction
Add friction that helps a user understand why the product is for them; cut friction that doesn't
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 4
- Confidence
- 90%
The naive growth instinct is to strip every step and minimize time-to-value. Amol argues the opposite: at Mercury, Masterclass, Calm, and Anthropic, deliberately adding the right friction consistently raised conversion and funnel completion. The test is whether a step helps the user understand why the product is for them. Questions about who the user is and what they want create friction — but that data drives activation, life-cycle marketing, and even look-alike ad targeting down-funnel.
Origin
A pattern Amol observed repeatedly across his growth roles (Masterclass, Mercury, Anthropic) and points to in peer products (Calm's onboarding quiz); a counter to the common 'cut all steps' growth dogma.
Core principles
- 01Friction is not inherently bad — only friction that fails to add value is
- 02Steps that help a user grasp 'why this product is for me' raise conversion despite adding effort
- 03The user data captured by friction pays off far beyond signup: activation, life-cycle, and ad-layer look-alike targeting
- 04High-volume products should test friction rather than assume; the direction of the effect is empirical
- 05Reducing cognitive load (splitting a six-field screen into two) is itself a form of tuned friction
How to run it
- 1
Classify every step as value-adding or annoying
Audit your flow and label each step: does it help the user understand why the product fits them, or is it dead friction that adds nothing? Only the second category is a candidate for cutting.
Watch out Don't default to 'cut everything to minimize time-to-value' — that fails most of the time.
- 2
Add questions that profile the user
Insert steps that ask who the user is and what their interest areas are, then use the answers to recommend the right product, feature, or content. This makes the product feel personally relevant.
Pro tip Splitting one dense form screen into two reduces cognitive load and can lift completion.
- 3
Test it against your own funnel, don't assume
Because the effect is business-specific, run the majority of these changes as experiments — especially at high volume — and confirm whether added friction lifts your completion before rolling it out.
Pro tip If the data shows a long flow performing well, defend it against 'too much friction' critics with the numbers.
Watch out A tactic that worked at another company can still fail on yours; verify before believing.
- 4
Reuse the captured profile down-funnel
Feed the who-is-this-user data into activation recommendations, life-cycle messaging, and even ad-layer look-alike targeting for users who dropped off. The profile is a 'juice that keeps on giving.'
In the wild
Anthropic's signup flow asks users who they are and what their interest areas are, then recommends products and features accordingly. Observers repeatedly tell Amol the flow has too much friction and is too long.
→ The team has the data showing the long flow performs well and is happy with it, using the captured profile to route users to the right products.
Common mistakes
Cutting all steps to 'just solve time-to-value'
The simplest instinct — strip every step and dump the user into the product — usually underperforms because it removes the moments that help the user understand the product is for them.
Adding friction that doesn't add value
Annoying friction that captures nothing useful and helps the user understand nothing should be removed; only value-adding friction earns its place.
Is it for you?
Best for
Growth and product teams with multiple products or personas and enough traffic to A/B test onboarding, purchase, and activation flows
Not ideal for
Products with a single obvious use case, very low volume that can't support testing, or flows where added steps genuinely teach the user nothing
From the transcript
“adding friction and adding the right steps uh leads to high-end conversion and and higher funnel completion.”
“cut friction when it doesn't add to the experience of helping a user understand why the product is for them.”
“it helps you with activation, but then it helps you with life cycle.”
From the episode
Head of Growth (Anthropic): “Claude is growing itself at this point”
Amol Avasare