The problem
A monolingual prospecting model in a multilingual market.
Jacobs operated in one of the world's most linguistically diverse markets, yet its outbound engine ran English-first. Agents rotated through shared dialers, hand-translating scripts into ad-hoc Arabic, Hindi, and Russian.
Daily prospecting capped at a few hundred meaningful conversations. First-touch arrived days after a prospect's moment of interest. Large segments, South Asian, Filipino, Persian, and francophone communities, stayed underpenetrated. Lead costs inflated to compensate; senior closers burned hours chasing unqualified names.
The solution
Real-time multilingual AI outreach at scale.
Over a four-week build, ConvoScale stood up ten language personas: Arabic, English, Hindi, Urdu, Malayalam, Tagalog, Russian, Farsi, French, and Mandarin. Each with localized openers, qualification logic, and calendar rules.
Wired into Jacobs' existing CRM and WhatsApp/SMS stack, every call outcome triggered the next best action: instant re-attempts on busy/no-answer, language-matched text follow-ups, live transfers for hot prospects. Sub-second latency, 10,000+ simultaneous calls, 24/7 coverage including off-peak hours aligned to expat shift work.
The compounding effect
Median time from list import to first touch dropped from 2.1 days to under 15 minutes. No-shows fell from 32% to 19% thanks to language-specific confirmations. Senior agents got redeployed from cold outreach to curated closings. Within six weeks, Jacobs had a scalable, always-on growth loop: multilingual outreach filled tomorrow's calendar, off-hours AI nurtured today's inbound interest, and closers focused on the highest-probability deals.