56% of Multilingual Shoppers Abandon Purchases Over Poor Localisation

What happened

56% of multilingual consumers say they have walked away from a purchase because of poor website localisation, according to a Liferay survey covered by CustomerThink.

The research also found that these consumers feel confident they can tell when website content has been machine-translated. Shoppers are not passively accepting rough translations. They are actively judging them, and that judgement feeds into their decisions.

The reporting frames localised digital experiences as something that can influence trust, loyalty and which brands people choose, not only whether they complete a single transaction.

Why it matters

For organisations selling across language markets, localisation is part of the customer experience itself. Language quality is one of the first signals a shopper uses to judge whether a brand has taken them seriously. When copy reads as machine-translated, the customer may reasonably conclude that care and local knowledge are missing, even if the product and price are competitive.

There is a behavioural dimension here too. People use ease of reading and clarity of language as shortcuts for credibility. Awkward phrasing creates friction, and friction at a moment of decision often gets read as risk. That makes localisation a trust question as much as a translation question, and it sits squarely within how digital experiences are designed and governed.

The Renascence take

Most organisations still treat localisation as a line item measured by the number of languages supported. That is the wrong yardstick. The measure that matters is whether a shopper can complete a task in their own language without pausing to question the brand.

Machine translation is not the problem in itself; unreviewed machine translation at high-stakes moments is. Pricing, delivery terms, returns, payment and trust signals are where doubt turns into abandonment, so they deserve human review first. Map the journey in each market, find the points where a customer hesitates, and fix those before expanding language coverage elsewhere. Localisation done this way is a service-design discipline, not a translation project.