AI Raised the Server Bill. Zoho Chose a Pay Freeze Over Layoffs.

This story was written with AI assistance and edited by me.

On October 13, Zoho co-founder and chief scientist Sridhar Vembu said on X that the company will not give its usual annual pay increase this year. He said server prices have increased fourfold, AI usage is raising expenses quickly, and the outlook for software is uncertain. Zoho has also cut its marketing budget, and Vembu said the company has not grown its employee count in 18 months. He wrote, "I have apologised for not being able to avoid this."

On October 15, CEO Mani Vembu said the company chose to skip raises rather than lay people off or raise prices, because the business is still growing and it has over a million customers to support.

Those are the facts as reported. Everything below is my opinion.

My position: whatever you think of the decision, Zoho is saying out loud something many companies are not. AI has a cost, and someone pays it.

The hidden bill

Most of the AI conversation is about savings. Do more with fewer people, ship faster, cut hours. Zoho is describing the other side: the compute, the servers and the usage that make those savings possible are expensive and rising.

If a company that runs its own data centers feels that pressure, smaller businesses buying AI tools by usage will feel it too.

I would not claim every company faces a fourfold server cost. That is Zoho's situation, in its own words. But the general point holds: usage-based pricing can grow faster than anyone planned, and a pilot that looks cheap can get expensive at scale.

Raises or layoffs

Mani Vembu said layoffs are common but did not feel right while the business is still growing. Spreading the cost across everyone, instead of removing some people, is a defensible choice. It is also a hard one. Employees lose real purchasing power, and some will leave.

The public reaction was sharp, and I understand why. A pay freeze is painful even when it is honestly explained.

What I give them credit for is saying it plainly. Many companies reach the same decision and hide it in vague language. Employees can handle bad news better than they can handle being kept in the dark.

What leaders can take from it

Put AI spend in the budget as its own line, with an owner, and review it every month.

Measure the return. If a tool saves ten hours a week and costs more than those hours are worth, that is not a win.

Decide in advance who carries the cost if it rises: customers, shareholders or employees. Do not let it get decided by accident.

And talk to your people early. The worst version of a hard decision is one they learn about from the news.

A fair counterpoint

To be fair, the critics have a point too. Asking employees to absorb costs while the company says it is still growing is a tough message, and a pay freeze compounds with inflation. Mani Vembu acknowledged there is no single right answer and said every software business will be having this discussion. I agree with that. My view is only that being open about the tradeoff is better than hiding it.

I also want to avoid overreading one company. Zoho is unusual: it runs much of its own infrastructure and has a long record of unconventional choices. What other firms can take from it is the discipline of naming the cost, not copying the specific decision.

The takeaway

Zoho's choice will be judged over the next few years, and the company itself said this period will be decisive. The lesson for everyone else is simpler: count the cost of AI as carefully as the benefit. When I grew a firm from $6M to $60M in under three years, the budgets that held up were the ones that named new technology costs early and assigned an owner. Strong people on a clear budget raise the whole team.

If you lead a team that buys AI tools, pick one line item this month and put a person's name next to it. Review the spend against the hours it actually saved. That one habit prevents a surprise later.