Independent
Software Developer
HR Manager
Team Leader
Accountant
Copy Writer
Customer Service
Virtual Assistance
The Third Path
On 3 August, Yellow.ai — a Bengaluru AI startup with about $30 million in revenue — announced it was going public through a SPAC. The plan: buy the call centers, then automate them.
For a month this column has walked through two possible answers to what AI does to outsourcing. One is the incumbent’s — Teleperformance’s pledge to make its 500,000 employees AI-enabled by 2027, IKEA’s earlier reskill, Genpact’s outcomes talk. Reskill the humans; keep the customer relationship. The other is the challenger’s — Sierra at $15.8 billion, Decagon at $4.5 billion, the short interest against the incumbents. Build a new operator; take the customer relationship by force.
On 3 August, Yellow.ai proposed a third. Merge with Bluerock Acquisition Corp. Raise up to $200 million in the SPAC trust. Take the whole package public at a pro forma equity value of $550 million. Then use the war chest and the listing to buy the incumbents and refit them as AI-native. It is the first publicly-declared version of a thesis General Catalyst has been running with $1.5 billion privately since 2023.
The two paths so far
The reskill path is slow and cash-negative. TP is accepting 0–2% revenue growth this year, plus €70–90 million in one-time reskilling costs, for the privilege of not firing 500,000 people. It works, but only if the market gives it three or four years to prove it — and the market so far has been pricing the sector as if the seats have already gone.
The displace path is capital-plentiful and customer-poor. Sierra sells about $150 million a year and is priced at $15.8 billion. Its problem is not funding; it is that it does not yet own the enterprise customer contracts the incumbents sit on top of. AI wins the operating margin. Distribution — SLAs, procurement paperwork, industry certifications — is what wins the enterprise account.
What Yellow.ai is actually buying
The interesting thing about Yellow.ai’s play is not the AI platform. It is the theory of what is scarce. AI adoption is cheap; the code is commoditizing every quarter. Customer contracts with named enterprises are not.
Yellow.ai plans to spend its SPAC cash — with a founding partnership that combines BPO operating leadership and PE roll-up execution — on complementary BPO operators, then use its own platform to convert them post-close. That is not a start-up move. It is a private-equity move dressed as a technology one. General Catalyst’s $1.5 billion Titan vehicle has been running the exact playbook with RFA and other acquisitions. 2025 already saw Capgemini pay $3.3 billion for WNS, TELUS take its Digital arm private, and Blackstone take TaskUs private. Yellow.ai just made the same trade public and cheap.
The math the market is missing
ISG data shows BPO annual contract value fell 14% in 2025, the lowest since 2020, as enterprises stalled on new outsourcing commitments to see what AI actually does to costs. That drop is Yellow.ai’s actual thesis — buyer indecision is creating discount asset prices, and a roll-up acquirer inherits the customer relationship at the discount instead of trying to win it back.
Two honest caveats matter. First, SPAC redemption is real: cash in a Bluerock trust only fully arrives if the SPAC’s holders do not redeem their shares, and 2025 redemption rates were running 60–90%. The $200 million is a ceiling, not a floor. Second, 9× revenue is a rich price on Yellow.ai’s ~$30 million top line, and the trade only works if roll-up execution moves faster than the runway the actual cash creates. Neither risk is disqualifying. Both are quotable when the first acquisition target haggles on price.
What this means for you
For BPO buyers, the calculation just gained a variable. Three paths, three risk profiles: an operator reskilling in place, a challenger building from scratch, and now a roll-up acquirer who buys the operator you already signed with and refits it inside your contract term.
That third possibility is the one buyers rarely stress-test. The right renewal question is no longer only “will your operator have AI.” It is: “will your operator still be your operator in eighteen months, and if it changes hands to a well-capitalized AI-native, does my contract come with it?”
The question for your business
Is your BPO provider the acquirer, the acquired, or the replaced?
