Invisible's Software-Managed Margins

Diving deeper into

Invisible at $134M in revenue

Document
generating significantly higher margins than traditional BPOs by continuously automating repetitive steps while maintaining humans for nuanced judgment
Analyzed 4 sources

This margin profile shows Invisible is not selling raw labor, it is selling a software managed workflow where human time is used only on the hardest judgment calls. Traditional BPOs often staff whole teams against a client process. Invisible breaks work into small steps, automates the repeatable pieces, routes exceptions to specialists, and charges on outcomes or task complexity. That lets the same labor pool produce more billable output per worker and supports 11% EBITDA margin at $134M of 2024 revenue.

  • In practice, the product looks more like an operations engine than an outsourced team. A client sends work by dashboard or email, Invisible decomposes it into subtasks, software bots handle rote steps, and trained agents complete reviews, rankings, or approvals. The client sees one finished result and a tracked unit cost, not a seat based labor contract.
  • The labor economics are closer to a curated marketplace with automation than a classic call center. For RLHF work, Invisible has charged roughly $30 to $45 per hour while paying raters about $15 to $20 per hour, with its internal platform handling routing and quality control across a 3,000 plus person workforce in 35 plus countries.
  • A useful comparison is Upwork versus staffing firms. In labor businesses, technology changes margin by lifting revenue per internal employee and cutting matching, contracting, and payment overhead. Invisible applies that same logic inside managed services, then adds premium pricing for specialized AI training and other workflows where human judgment still matters.

The next step is moving this model into regulated enterprise workflows where full automation is not enough. If Invisible keeps turning human review into a thin quality layer on top of software driven execution, it can keep expanding from AI labeling into finance, insurance, and other back office categories that were historically owned by BPOs.