Bernd Rennebeck

Developer, Father, and All-Around Geek.

[Bernd Rennebeck @brennebeck]

Sources:

- Reuters interview on TCS’s forward-deployed engineering plan: https://t.co/QOG20pUigc
- Official TCS Q1 FY27 results and workforce figures: https://t.co/RiGcTqaWUD
- TCS earnings press-conference coverage, including the 10%-15% productivity pass-through and outcome-pricing comments: https://t.co/s8YBKV9id6
- SP2C/EY French outsourced customer-relations barometer release: https://t.co/5AQUwRlbLa
- Critical French trade coverage: https://t.co/vJpcQLWfWa

The French report does not separate the effects of AI, weak demand, conventional offshoring, and other automation. Its revenue figure covers services for the French market, including delivery abroad. Its 36,478 FTE figure covers jobs located in France, not the global workforces of providers such as Teleperformance or Concentrix. TCS has not disclosed a rollout timetable or whether most FDE roles will come from retraining or external hiring.

[Bernd Rennebeck @brennebeck]

France’s outsourced customer-service market finished 2025 with revenue down 1.4%. Across the qualified market, France-based employment fell 3.6% to 36,478 FTE. AI solutions accounted for only 8.7% of reported value.

Those numbers are not apples-to-apples. The revenue estimate includes services for the French market delivered from outside France. The 36,478 figure counts jobs located in France; it does not measure total employment at Teleperformance, Concentrix, or the other global providers. The report shows France’s employment base shrinking faster than French-market revenue. It does not tell us how much of that gap came from AI.

TCS has now said how it plans to compete with that kind of pressure.

It wants forward-deployed engineers to make up 1% to 1.5% of its 593,798-person workforce. That works out to roughly 5,900 to 8,900 people working inside client environments. TCS has not said how many will be new hires.

CEO K. Krithivasan told Reuters:

"What you need is a deep knowledge of the customer environment to make it work. That is where we differentiate ourselves. This has nothing to do with cost arbitrage."

I would not take the last sentence literally. Employee costs were 58.3% of TCS revenue last quarter, and the company added 9,279 people. Labor scale still matters.

The quote tells buyers what TCS wants them to pay for next. On the same earnings call, management said clients are already receiving 10% to 15% productivity gains and that more contracts are moving toward output- and outcome-based pricing.

The proposed FDE group would sit between the buyer and a delivery workforce approaching 600,000 people. That group helps decide which work becomes software, which work stays human, how the result is measured, and what TCS gets paid for.

Philippine operators should count how many people they can put inside a client’s process with that mandate. AI training on the delivery floor helps with execution. Pricing power comes from being allowed to change the process and put a price on the result.

If that authority stays with the client, the platform vendor, or the foreign parent, the Philippines still gets plenty of work. It gets less say over how that work is defined, priced, and reduced.

[Bernd Rennebeck @brennebeck]

Sources:

Moneycontrol on Cognizant outcome-based BPO contracts:
https://t.co/F0vL1jpHr5

Cognizant Q1 2026 earnings-call transcript, including the labor-based to outcome-based language and AI-infused rate cards:
https://t.co/YQ0zxdSJjU

Related piece:
From Hours to Outcomes: The Business Model That Survives AI
https://t.co/hH5s0SFt3T

[Bernd Rennebeck @brennebeck]

Moneycontrol reports that 45% of Cognizant's new BPO contracts are now outcome-based.

Cognizant's own investor language points the same way. On its Q1 call, the company said it is "shifting our economics from labor-based to outcome-based models" and described AI-infused rate cards that blend human effort and digital effort.

This is a pricing signal.

For twenty-plus years, the offshore services bargain was built around the seat: coverage, hours, staffing ratios, labor cost, SLA compliance.

AI makes more of the operating work measurable. Resolved cases. Cleared invoices. Claims processed. Turnaround time. Exception rate. QA accuracy. Cost per transaction.

Once buyers can measure those units more tightly, they push contracts toward results. They still need people, but the people are increasingly tied to exceptions, judgment, QA, escalation, and client-specific context.

That changes the Philippines' problem.

If a provider is only selling cheaper labor inside someone else's workflow, outcome-based AI pricing squeezes it.

If a provider owns the workflow, data loop, evaluation set, governance, and client outcome, there is still margin to capture.

This is what 'from rented labor to owned capability' means in contract terms.

[Bernd Rennebeck @brennebeck]

Yesterday, LTM (formerly LTIMindtree) launched BlueVerse Currency, an outcome-based pricing model for agentic AI work.

For BPO and services operators, the important part is the pricing logic. LTM says the model is tied to measurable business results and shared productivity gains, while bundling people, agents, platforms, accelerators, and compute into one commercial construct.

That pushes the AI shift into the contract.

For a long time, services growth followed a familiar equation: more demand created more seats, more billable effort, and more revenue.

This model points somewhere else. More output will not automatically mean more headcount. More value will accrue to whoever owns the workflow, measures the result, and captures the productivity gain.

Philippine BPO and IT-BPM operators should pay attention to that.

The opportunity is still large. But scale by itself gets weaker when the margin starts moving toward workflow ownership, governance, data, and domain context.

The next competitive gap may be less about who can supply the most labor and more about who can remain essential after AI compresses the labor required.

The open question is whether this becomes real contract behavior or stays mostly positioning.

My read: once a major services firm starts pricing AI this way in public, the seat-based model is already under pressure.

#BPO #AI

[Bernd Rennebeck @brennebeck]

TCS just said out loud what the outsourcing industry has been trying to soften for months.

This week, chairman N. Chandrasekaran said TCS could eventually have as many AI agents as employees, and that hiring will slow as more work shifts to agents. TCS says annualized AI revenue has already crossed $2.3 billion.

This is a real operator signal from one of the world's biggest services firms.

My read: growth is moving up the stack, not into the payroll.

For years, the outsourcing equation was simple. More work meant more seats. More seats meant more revenue.

Now more work can mean more agents, better workflow design, slower hiring, and higher output per employee.

That is why I keep coming back to workflow ownership.

The industry has stopped pretending AI is just a copilot story. The model is shifting from labor scale to workflow ownership.

The Philippines will keep doing BPO work. The installed base is too large and the workforce too experienced. But labor scale alone is no longer enough.

The next phase of value will go to firms that own the workflow, the QA layer, the governance, and the customer relationship above the seat count.

If Philippine firms build there, they keep more of the upside.

If they do not, growth happens in the parent company while the Philippines keeps the execution burden.

#BPO #AI

[Bernd Rennebeck @brennebeck]

Davao Today ran a worker-side warning that agentic AI could wipe out the Philippine call-center industry. Around the same time, Genpact's Global Agentic AI Officer said the link between revenue growth and headcount growth is breaking, and called it "by design, not by accident."

Worker-side fear and vendor-side strategy are now describing the same shift: more process volume won't automatically mean more seats. The seat-based model is being repriced, and the Philippines still has a window to respond.

That window only matters if the response is active — and active means more than another upskilling program. The soft landing comes from using today's BPO base as a bridge to something more locally owned and AI-native:

- outcome-based services instead of billing by the hour
- ownership of the workflow itself, end to end
- stronger data and QA loops
- operators who turn domain knowledge into products and managed outcomes

AI is already touching Philippine BPO. The open question is who captures the productivity gain: Filipino operators, or whoever owns the next layer of value offshore.

What's the first workflow Philippine operators should own instead of just staff?

#FutureOfWork #Philippines

[Bernd Rennebeck @brennebeck]

Full article: https://t.co/7s0NqDqoNA

[Bernd Rennebeck @brennebeck]

The Philippines' biggest conglomerates built the offices the BPO sector filled, the banks that held its wages, and the malls its workers spent in.

They didn't build the software.

Now AI is restructuring the sector from underneath — and it's the institutions whose balance sheets are wrapped around BPO payroll that have the most to lose.

New long-form piece on who has to build next, and why the math points toward the conglomerates doing it.

[Bernd Rennebeck @brennebeck]

https://t.co/v7I0mPBTXZ

[Bernd Rennebeck @brennebeck]

https://t.co/v7I0mPBTXZ

[Bernd Rennebeck @brennebeck]

Check out the podcast at https://t.co/8jpJQ39cfj and read more at https://t.co/Tkx7eeVzwO

[Bernd Rennebeck @brennebeck]

"For India, it's existential.
For the Philippines, we're a byproduct."

TCS, Infosys, Wipro — Indian founders & shareholders.
When they automate, the upside stays in India.

The Philippines built a $40B BPO industry answering to boards in Charlotte & Paris.
As AI nationalism rises, those companies will optimize for their shareholders. We’re not in that room.

Sat down with John Soriano on the Oblique Podcast to break it down.

What do you think — are we at risk, or can the PH turn it around?

Links in comments 👇

#FutureOfWork #AIPhilippines

[Bernd Rennebeck @brennebeck]

Full article on the blog: https://t.co/BTtwLzxlut

[Bernd Rennebeck @brennebeck]

India built TCS. The Philippines built a very good tenancy.

Both countries started with similar advantages 25 years ago:

- English-speaking talent pool
- Rising college graduation rates
- Hunger to build something bigger

India chose capability ownership.
Philippines optimized for labor rental.

TCS today: $30B+ market cap, owns IP, owns client relationships, owns outcomes
PH BPO today: $40B industry, zero Fortune 500 multinationals HQ'd in Manila, profits exit the country

Same starting point. Radically different trajectories.

Here's the thing: in a world of infinite labor arbitrage, the tenancy strategy makes sense. You can rent cheaper than they can build.

But AI collapses the infinite labor assumption.

When labor gets commoditized by software, the advantage goes to whoever owns:

- The customer relationship
- The domain IP
- The outcome definition
- The pricing model

Not whoever can execute the cheapest.

The Philippines has 18-24 months to stop being TCS's supply chain and start being the next TCS.

That requires capital, policy, and urgency in proportions the country hasn't mobilized before.

What would it take to make that actually happen?

#Philippines #Futureofwork

[Bernd Rennebeck @brennebeck]

https://t.co/Idv4rgPDwb

[Bernd Rennebeck @brennebeck]

https://t.co/Idv4rgPDwb

[Bernd Rennebeck @brennebeck]

Full article: https://t.co/SX1KuL1XqJ

[Bernd Rennebeck @brennebeck]

The Ghost GDP piece traced where the money goes when AI replaces a Filipino worker. This one traces where the damage goes.

IT-BPM accounted for 44 percent of all Philippine office transactions in 2024. Metro Manila vacancy was already 19.8 percent at year-end. BSP data put banking-system real estate exposure at ₱3.3 trillion as of March 2025. When a sector that large contracts, the pressure doesn't stay inside it.

The sequence is mechanical: BPO headcount shrinks --> office vacancy rises --> developer revenues and REIT cash flows deteriorate --> household credit tightens --> urban retail and transport slow --> local government tax bases erode --> bank balance sheets take the hit. Each link is downstream of the one before it.

The secondary cities are where this lands hardest. Cebu, Davao, Iloilo — places where BPO payroll is the dominant source of formal middle-class income and there is no diversified economic base to absorb a demand shock.

Most policy conversations about AI and BPO treat this as a labor market problem. The blast radius says it's a real estate problem, a banking problem, a fiscal problem, and a current account problem — all at the same time.

New article on the blog. Link in comments.

#Philippines #Futureofwork

[Bernd Rennebeck @brennebeck]

https://t.co/IwP7C8yUCj

[Bernd Rennebeck @brennebeck]

https://t.co/IwP7C8yUCj

[Aaron Levie @levie]

https://t.co/5zTvYqHztE

[Bernd Rennebeck @brennebeck]

Full analysis — the mechanics, the scenario modeling, the investment gap, and why the 1.9M people in BPO are better understood as a founding class than a casualty list.

https://t.co/K567abrilV

[Bernd Rennebeck @brennebeck]

The Philippines' BPO sector just crossed $40 billion in revenue.

GDP is up. Employment is stable. The headline numbers look fine.

Underneath those numbers: AI is shifting income from Filipino paychecks to foreign shareholders. Output stays. The money leaves.

I spent a year modeling what happens next. The short version: GDP keeps rising while the middle class hollows out. The sector's own success hides the adjustment until it's too late to manage.

[Bernd Rennebeck @brennebeck]

https://t.co/xUtBlaq1DV