Each NPS band doubles referral rate. Moving users from 6→8, 8→9, or 9→10 compounds exponentially across millions of customers.
"If you move NPS from 30 to 50, you increase the viral coefficient of your entire customer base."
70% of Wise's 1M new users/quarter come via word of mouth
Wise NPS: 70 — higher than the iPhone or Google Search
Banks average −30 NPS. The bar was low; the opportunity was huge.
Framework
The 3 Pillars That Drive Word of Mouth
70%
growth via WOM
1M
new users/quarter
0.35%
avg fee (was 6-7%)
How Wise found the pillars
NPS comments were emailed to the whole company every week for 3–4 years. Customers kept repeating the same things: make it faster, cheaper, easier. The pillars weren't chosen — they emerged from listening.
At 5.9% vs competitor's 6%, users use Wise but don't recommend it
At 8–10× cheaper, they tell everyone they know
Advocacy only kicks in at a threshold — marginally better never triggers it
NPS comments guided the product roadmap for years — not A/B tests
One list: things that make customers happy — not two lists (customers vs revenue)
The 10× Playbook
Give Users an Experience They Didn't Know Was Possible
Most teams stop when the product "works." Some polish conversion rates. But word of mouth only comes when you blow users' socks off — doing something that didn't previously exist.
The Singapore EKYC Story
Regulators required face-to-face verification for every customer. Wise sent a team, physically met 20,000 people on a waitlist — expensive, unscalable, painful. But customers complained to the government. After a year of lobbying Wise secured the world's first EKYC licence: selfie + photo ID = verified. That's 10× better. That's advocacy.
The Central Bank Play
The cheapest banking partner isn't a cheaper bank — it's the central bank. Wise now holds accounts at the Bank of England, National Bank of Singapore, and Reserve Bank of Australia. Each took years of lobbying. The payoff: structural cost advantage no competitor can replicate.
Cost Allocation Model
Every cost — a support call, FX risk exposure, partner fee — is allocated to the exact transaction that generated it. Margin added on top. The costliest 20% of customers get a price increase; everyone else gets cheaper. Teams then systematically engineer those costs down, opening cheaper price bands over time.
Ask: what is the theoretical minimum cost / maximum speed for this route?
Don't iterate toward 10× — jump to the end state and work backwards
"You can't split test your way to love" — build conviction, then execute big
Teams are rewarded for tackling the seemingly impossible, not quick wins
Users believed they saved money with Wise but didn't trust the number. One PM discovered how to make the value undeniable.
The Graph Insight — 3× sharing
A PM sketched a comparison chart: your bank's hidden fees vs Wise. Iterated it at a coffee shop until people said "I'm never using my bank again." Added it to the success screen with a share button. Referral rate tripled.
Product marketing isn't a separate team — it lives inside the success moment
Give a PM 3 weeks to talk to 5 customers and pick one big thing to build — not run 10 tests
Wise monitors ~70 bank accounts worldwide to log competitor prices daily
The principle
"Closing the delta between what you've done and what's perceived to be done is product marketing within the product."
Contrarian
Word of Mouth Myths
✗Run experiments to find your growth leverINSTEAD →✓ Build conviction through qualitative insight first. "You can't split test your way to love." Talk to customers, pick one big bet, then ship it confidently.
✗Trust is a landing page problemINSTEAD →✓ People trust their friends, not landing pages. Word of mouth is the strongest trust signal that exists — and it's free. Build so good that friends become your sales team.
✗Mission emails don't acquire customersINSTEAD →✓ Wise's mission email — no CTA, no sign-up button — was forwarded so widely it outperformed every other campaign. Authentic purpose travels further than any ad.
✗Marginally cheaper products still go viralINSTEAD →✓ At 5.9% vs 6.0%, users use you and stay silent. At 8–10× cheaper, they evangelise. There is a threshold for advocacy — incrementally better never reaches it.