BPO
Spanish-language customer experience for a United States omnichannel retailer
At a glance
- Industry
- E-commerce and retail
- Client geography
- United States, national footprint
- Client size
- Enterprise, approximately USD 1.9B annual revenue
- Service line
- BPO, customer experience across voice, chat, email and social
- Primary language
- Spanish for the US Hispanic market, with English overflow
- Delivery site
- Bogotá
- Engagement duration
- 19 months, ongoing
- Team size
- 108 agents, 9 team leaders, 3 QA analysts, 1 WFM analyst
Client profile
The client is a United States omnichannel retailer with approximately USD 1.9 billion in annual revenue, operating a national ecommerce platform alongside a physical store estate. Over the preceding five years Spanish-speaking customers had grown to roughly 20% of transacting customers, concentrated in Texas, Florida, California, Arizona and the New York metropolitan area.
The challenge
The client's customer satisfaction data carried a persistent and uncomfortable gap. English-language CSAT sat consistently around 88%. Spanish-language CSAT sat at 70%. The gap was not explained by product mix, order value or geography. It was explained by service: Spanish-speaking customers were served by a thin bilingual layer inside a predominantly English operation, with long queues, frequent transfers, and a meaningful share of interactions where the customer switched to English because it was faster than waiting for Spanish support.
The commercial consequence showed in repeat purchase behaviour. The client's analytics team found Spanish-speaking customers who contacted support repurchased at a rate materially below English-speaking customers who contacted support, against a baseline where the two segments purchased similarly when they did not contact support at all. Support contact was destroying value in one segment and creating it in the other.
Prior remediation had failed twice. Recruiting bilingual agents into United States contact centres ran into wage competition and attrition above 60%. A subsequent offshore pilot produced Spanish that customers described as formally correct but unfamiliar, with recurring complaints about vocabulary that did not match Latin American usage.
Timing pressure came from a planned Spanish-language checkout and app localisation. Launching a localised purchase experience on top of a support function Spanish-speaking customers already rated poorly was a risk the client's leadership would not take.
Why Corpshore Colombia
Corpshore Colombia was selected on three grounds. Colombia supplies Spanish with substantial exposure to United States idiom and consumer culture through a large diaspora connection, which addressed the verbatim complaints from the earlier offshore pilot. Colombia Time matches United States Eastern for much of the year, so the Spanish desk runs on the same schedule as the client's English operation without night premiums or handover gaps. And Corpshore proposed measuring Spanish CSAT against English CSAT as the contractual headline metric, which targeted the gap the client actually cared about.
The client's VP of customer experience noted that Corpshore was the only bidder to ask for the segmented repeat-purchase data before proposing, and the only one to structure the proposal around closing a gap rather than delivering a standalone service level.
The engagement
The team launched with 32 agents in Bogotá and scaled to 108 by month nine. Composition: 92 Spanish-primary agents with B2 or higher English, 16 English-primary for overflow, nine team leaders, three QA analysts and one workforce management analyst. Coverage runs 07:00 to 23:00 US Eastern, seven days, on the same clock as the client's operation. Tooling is the client's Salesforce Service Cloud with a Spanish knowledge base built during onboarding.
Approach and methodology
Spanish as the primary, not the exception. The team was built as a Spanish operation with English overflow rather than a bilingual operation that also handles Spanish. Agents are hired for Spanish quality first. This inverted the logic of every prior attempt and is the change the client credits most.
Latin American register, explicitly defined. A written Spanish style guide covering vocabulary, register, formality and regional variation was authored during onboarding, reviewed against verbatim feedback from the failed offshore pilot. Terms that had appeared in complaints were proscribed and alternatives specified.
A Spanish knowledge base, not a translated one. The English knowledge base was rebuilt in Spanish rather than machinetranslated, with 320 articles authored over the first four months. Translated help content had been a recurring source of resolution failure.
Gap-based quality management. The QA scorecard measures Spanish interactions against the same rubric applied to the client's English operation, with fortnightly joint calibration between Corpshore QA and the client's United States quality team, so the two operations converge rather than drift.
Customer satisfaction by language of service
Results
Spanish-language CSAT rose from 70% to 92% by month 12, closing and slightly exceeding the English benchmark of 89%. First contact resolution on Spanish interactions rose from 52% to 85%.
The repeat-purchase gap that prompted the engagement closed substantially, with Spanish-speaking customers who contacted support repurchasing within a few percentage points of the English equivalent by month 12.
The Spanish checkout and app localisation launched on schedule in month 14, supported from day one by a desk already at parity. Cost per contact fell 57% against the client's blended United States bilingual cost.
Enduring value
The Spanish style guide and the 320-article knowledge base are client-owned assets that now support the localised checkout, the mobile app and email marketing. The joint calibration cadence has continued for the full engagement and has become the model the client applies to its other outsourced functions.
Key indicators
| Metric | Baseline | Month 12 | Change |
|---|---|---|---|
| Spanish-language CSAT | 70% | 92% | +22 pts |
| Gap to English CSAT | -18 pts | +3 pts | Closed |
| First contact resolution, Spanish | 52% | 85% | +33 pts |
| Average handle time | 9.6 min | 6.2 min | -35% |
| Transfer rate | 32% | 7% | -78% |
| Cost per contact | USD 7.60 | USD 3.27 | -57% |
| Annualised attrition | 61% (US bilingual) | 18% | -70% |
Where a metric disclosed only a change, the absolute figures are shown as a dash. Figures are client-reported or jointly measured.
We had an eighteen-point satisfaction gap between our English and Spanish customers and we had stopped believing it could close. It closed in a year. The difference was building a Spanish operation rather than adding Spanish to an English one.
Related topics