What is Zipang? The Indonesian BPO platform explained (2026)

Zipang is a BPO-first talent acquisition and management platform. The product is a managed pod of Indonesian remote professionals, screened through a 5-gate funnel, supervised by Zipang managers, and measured against KPIs that the client agrees to in writing before the program starts. The model is outcome-based: clients pay per seat, per shift, or per task bundle, and the replacement guarantee is written into the contract.

The platform runs as a single-country, single-entity operation under PT Lima Cakar Bumi (Indonesia). All client contracts are executed through the local Indonesian entity, all talent is engaged through the local entity, and all statutory obligations (PPh 21, BPJS Kesehatan, BPJS Ketenagakerjaan, THR, year-end filings) are handled in-house. This is the structural difference between Zipang and a freelancer marketplace: Zipang is the employer of record for the talent, not a directory or matching layer.

Zipang is operated by PT Lima Cakar Bumi, an Indonesian BPO and shared-services company founded in 2015 by Yoseph Gratika. The company has been in continuous operation since then, with headquarters in Jakarta and delivery presence in Bandung, Surabaya, Yogyakarta, and Medan. The legal entity is registered as a Perseroan Terbatas (PT) under Indonesian law, with the NIB (Business Identification Number), NPWP (tax ID), and BPJS registrations public and verifiable.

PT Lima Cakar Bumi runs Zipang as its primary BPO production brand. The company also operates internal shared-services capacity that supports Zipang's managed pods, payroll administration, KPI engineering, training cohort management, and supervisor coverage. For a foreign client, the practical implication is that Zipang is not a startup or a marketplace; it is an established Indonesian BPO operator with a 10+ year track record, statutory registrations, and a published production history.

The BPO-first model is the structural difference between Zipang and the alternative ways a foreign employer can hire Indonesian remote talent. On a freelancer marketplace (Upwork, OnlineJobs.ph, Fiverr), the client contracts an individual worker, holds the worker's profile as the trust signal, and is responsible for screening, payment, tax, IP, and replacement. On a curated-individual platform (Toptal), the client pays a premium hourly rate for a senior individual expert, again with the bilateral contract as the trust signal. On a recruiter, the client pays a placement fee and receives a shortlist of profiles, with the placement as the unit of value.

Zipang's model is none of those. The unit of value is the pod: a managed group of Indonesian remote professionals with a published KPI, a manager, a supervisor, a QA function, and a 30/90-day replacement guarantee. The trust signal is not the worker's profile or the placement; it is the published production data of similar programs (432 deployed, 3.4M tasks/month, 90%+ accuracy, 88%+ retention). The client does not re-run hiring when an individual underperforms; the operator does.

Zipang publishes production numbers that are first-party and verifiable during a vendor diligence call. The largest program to date deployed 432 professionals for a 100+ hypermarket retail AI program in France; 208 are in active production at any time, processing 3.4M production tasks per month at 90%+ sustained accuracy with microsecond-level KPI tracking. A second named program, a Transperfect–Dataforce interpolation annotation engagement, runs 20 full-time annotators out of 60 trained, with sustained accuracy climbing from 78% in week 1 to 92% in week 12 via peer review and a three-layer supervisor calibration cycle.

The third benchmark is a 24/7 content-moderation program for a global social platform, with 50+ Bahasa and English moderators running 4 follow-the-sun shift rotations, a mental-health program, and 88%+ 12-month retention. The numbers matter because they are what an employer verifies: not marketing claims, but production data from named clients and named program shapes. They are also the data behind the 30-day replacement window: if a hire in a similar program does not meet KPI, the operator pulls from the same funnel that produced these results.

Zipang runs named programs across a handful of publicly disclosed client relationships. ByteDance / TikTok Indonesia engaged Zipang for KOL training and activation: 4,000+ KOLs were trained, 240 activated, and a Tencent / PUBG Mobile campaign drove 120M views in 10 days from 300 KOLs. Transperfect–Dataforce runs a sustained interpolation annotation program through Zipang. A 100+ hypermarket retail network in France runs a multi-year shelf-edge computer vision annotation program, and a global social platform runs the content-moderation program referenced above.

Most of the client roster is not publicly disclosed (NDA-bound, and many of Zipang's enterprise clients prefer the operator to keep program shapes private). What is public is the production data, the screening methodology, and the verifiable case studies. For HR and operations leaders, the practical implication is that Zipang's KPIs are anchored to live programs, not promised on a pitch deck.

Zipang's talent pipeline is built around a 5-gate screening funnel, applied in three stages. Stage 1 (gates 1–2) is mass filtering: CV relevance scan and async English plus SOP-comprehension test. Stage 2 (gates 3–4) is role-specific validation: a role-specific quiz and a structured video interview. Stage 3 (gate 5) is the paid trial task scored against the client's actual production KPI. Aggregate end-to-end pass rate is 6–12% of applicants, depending on the program.

From approved shortlist to first productive shift, Zipang publishes 7–14 days for most roles, with 3–6 weeks for larger teams (50+) running phased onboarding. The 6–12% pass rate is unusually transparent for the category, most BPO operators do not publish pass rates, and the 7–14 day ramp is anchored to the same funnel that produced the 432 deployed, 90%+ accuracy, 88%+ retention benchmarks. For a client, the operational meaning is that the timing is data-driven rather than self-reported.

Zipang is not a freelancer marketplace. There is no public profile directory, no hourly billing, no platform-mediated dispute resolution, and no self-attested trust signal. The contract is between the client and PT Lima Cakar Bumi, not between the client and an individual worker.

Zipang is not a CV library or a resume aggregator. Candidates are not made available for the client to shop through; the client sees shortlists that have already passed the 5-gate funnel, and personal data is hidden until the client formally shortlists.

Zipang is not a recruiter or a placement agency. There is no placement fee, no one-off handoff, and no 'good luck' once the candidate is placed. The engagement is operational: Zipang manages the pod, the KPI, the supervisor, and the replacement guarantee for the life of the program.

Engaging Zipang as a foreign employer starts with a free registration at /employers, which opens an NDA-gated conversation with the Zipang employer team. The standard path is: (1) submit a brief describing the workload, volume, SLA, and KPI expectations; (2) receive a shortlist of pod structures with proposed ramp windows and pricing; (3) run a 1–3 seat pilot on the 5-gate funnel, with the trial task scored against the agreed KPI; (4) if the pilot meets KPI, scale to 5–50 seats over 4–12 weeks; (5) ongoing program management through the client portal with weekly KPI reports, monthly reviews, and a 30/90-day replacement guarantee.

For most programs, the timeline from registration to first productive shift is 14–28 days. For larger 50+ seat ramps, the timeline is 3–6 weeks of phased onboarding. There is no upfront cost to register or to receive a shortlist, and the trial task is paid (the candidate is paid for the trial, not the client paying Zipang to test). For a 50+ seat program, expect an MSA, a per-seat or per-shift pricing schedule, and a published KPI dashboard from day one.