From Friction to Flow: How Wise Engineered Product Discovery into its Culture
Part 3 of series: How the world’s best companies turn Empathy + Evidence into culture
When you study how the world’s strongest product organisations work, you start to see the same pattern.
Amazon shows what happens when intuition outruns validation — and how quickly a giant can learn from a very public miss. Netflix shows how curiosity becomes a disciplined habit. Wise — formerly TransferWise — reveals the next evolution: a culture where empathy and evidence aren’t rituals but architecture.
Empathy that starts with lived friction
Wise’s founding insight didn’t come from a focus group. It came from two bank statements.
As Reuters reported, founders Taavet Hinrikus and Kristo Käärmann were Estonians living in London, constantly shifting money between pounds and euros — and losing large amounts to hidden exchange-rate markups and fees. The Financial Times has similarly documented how this everyday friction became the spark for what would become a global money-movement platform.
From that frustration emerged a mission that Wise repeats consistently across product and investor materials: “money without borders — instant, convenient, transparent and eventually free.” Even the BBC describes Wise’s mission through the same four lenses: lower cost, greater speed, clear fees, and ease of use.
This is not branding. It’s a decision algorithm.
Inside Wise, every significant product bet must answer:
Which part of this mission does it improve — and by how much?
It’s empathy operationalised: lived pain, converted into measurable constraints.
Discovery embedded in structure, not stages
Wise is famous for its autonomous squad model. In a feature on how Wise builds product, Wired UK described its teams as “independent, cross-functional and outcome-owned,” with product, design, engineering and analytics working shoulder-to-shoulder.
Wise’s own careers pages reinforce the same picture. Roles such as Design Researcher – Onboarding describe researchers as “embedded members of the product team,” shaping hypotheses and validating behaviour directly, rather than handing off static insight decks.
Product squads are also supported by a distinctive partner: Product Compliance.
In an article from Finextra, the industry trade publication, Wise’s compliance organisation is described as “working as a partner to product engineering, ensuring regulatory requirements are understood early and built into solutions rather than bolted on.” Wise’s own job descriptions echo this: compliance analysts sit within teams, helping shape ideas before code is committed.
Put together, you get teams where discovery and delivery run in parallel — not as separate phases.
Research asks is this desirable?
Compliance asks is this safe and legal?
Engineering asks can it scale?
And they answer together.
This avoids one of product’s most painful anti-patterns: “build fast, fix later.”
Architecture that makes experimentation safe
Evidence is not just a mindset at Wise — it’s infrastructure.
Wise has documented its rollout strategy extensively on the Wise Engineering Blog. They use safe deployment techniques — including automated canary releases — to ship changes to small slices of traffic while monitoring both technical and business metrics.
Google and Netflix jointly highlighted Wise’s canary approach in a Spinnaker case study on progressive delivery, noting how Wise uses automated rollback conditions to protect customer trust while still shipping continuously.
Experiments themselves run on Wise’s internal framework, tw-experimentation, which they open-sourced. It supports A/B tests, causal inference and guardrail metrics — allowing teams to measure both positive impact and unintended side effects.
For any new idea, two questions matter:
Does it improve a real customer outcome aligned to the mission?
Does it avoid causing harm elsewhere in the system?
If an experiment violates either, the system shows it quickly.
Blameless learning as organisational memory
Not every experiment works. In fast-moving, highly regulated systems, some will fail — and Wise expects that.
Wise’s engineering post on blameless postmortems describes how teams document incidents without blame, focusing on what happened, why, and how to avoid recurrence. This mirrors industry best practice, popularised by reliability leaders like Charity Majors at Honeycomb, who argues that psychological safety is a prerequisite for high-velocity teams.
When teams share these learnings openly, individual failures turn into collective foresight. Insights become findable. Patterns become visible. Mistakes need not be repeated.
This is how learning compounds — not just within a team, but across an organisation.
From intuition to infrastructure
If Amazon warns us what happens when intuition is left unchecked, and Netflix shows how curiosity becomes measurable, Wise demonstrates what happens next:
Empathy is encoded in mission, derived from real lived friction.
Discovery is built into structure, with embedded researchers and compliance partners informing decisions early.
Evidence is engineered into architecture, via experimentation frameworks and safe rollout systems.
Learning becomes organisational memory, reinforced through blameless postmortems.
Wise doesn’t “do” product discovery. It scales it.
This is what it looks like when empathy and evidence stop being rituals — and become the operating system.
