Google vs. Infosys: Two H-1B Business Models

When people talk about "the H-1B program," they usually have one of two very different companies in mind without realizing it. Some picture a technology firm hiring a specialized engineer it couldn't find quickly enough at home. Others picture a global outsourcing firm rotating thousands of interchangeable workers through client projects. Both pictures are real. They just describe different business models that happen to use the same visa.

Google and Infosys are the cleanest examples of each. Comparing them side by side explains far more about the H-1B debate than any single statistic, because almost every disagreement about the program is really a disagreement about which of these two models you have in mind.

This article walks through how each company uses H-1B sponsorship, why their filings look so different, and what a job seeker or researcher should actually take away from the contrast.

Two companies, two reasons to sponsor

Google LLC is a product company. It builds and sells its own software, hardware, and cloud services, and it earns money from those products. When Google sponsors an H-1B worker, that person is typically a direct employee working on Google's own products at a Google site. The job is the destination, not a placement.

Infosys Limited is an IT-services and consulting company. Its product is, in effect, labor: it signs contracts to staff and run technology projects for other companies — banks, retailers, insurers, manufacturers — and supplies the engineers who do that work. When Infosys sponsors an H-1B worker, that person is usually billed out to a client and may sit at the client's office, not Infosys's own campus.

That single difference — who consumes the worker's labor — drives almost everything else you'll notice in the data.

How the filings differ

Because the two companies sponsor for different reasons, their Labor Condition Applications (the wage-and-job filings employers must submit before petitioning for an H-1B) tend to look different in a few predictable ways.

Volume and concentration. Services firms file at enormous scale because their headcount is their revenue engine. The more billable consultants on visas, the more projects they can staff. Product companies file in large numbers too, but sponsorship is a recruiting tool, not the core of the business model. You can see how the largest sponsors stack up below — note how many of the top names are outsourcing and consulting firms rather than household-name product companies.

Job titles. Product-company filings skew toward narrowly defined, higher-band roles: software developers, research scientists, security and infrastructure specialists. Services-firm filings cluster in roles built for client deployment and rotation — computer systems analysts, software developers, and the catch-all computer occupations, all other category that consultants are frequently filed under.

Wage level. This is the number job seekers care about most, and it's where the models diverge most sharply. The DOL requires employers to pay the higher of the "prevailing wage" for the role and location or the actual wage paid to similar workers. Product companies competing for scarce senior engineers routinely file well above the prevailing wage because they're bidding against other product companies. Services firms, whose margins depend on the spread between what they pay a consultant and what they bill a client, have a structural incentive to file closer to the prevailing-wage floor.

Location. Product-company filings concentrate around their own campuses — for Google, heavily in California and Washington. Services-firm filings spread across the country, following wherever their clients happen to be: Texas, New Jersey, Illinois, and many others.

Why the wage gap isn't a simple "good vs. bad" story

It's tempting to read higher filed wages as proof that one company is "better" for workers and the other is exploitative. The reality is more structural.

A senior research engineer at a product company and an early-career consultant at a services firm are not doing the same job, so comparing their filed wages directly is misleading. Much of the gap reflects a difference in the mix of roles and seniority each company sponsors, not just generosity or stinginess at the same role.

That said, the incentives are genuinely different, and they matter. A product company's reason to sponsor is "we want this specific person." A services company's reason is "we need billable capacity." The first model pulls wages up because the employer is competing for an individual. The second model pushes wages toward the legal minimum because the employer is competing on price for a contract. Neither is illegal — the prevailing-wage system is designed to set a floor precisely because lawmakers anticipated the second incentive. But it explains why the same visa produces such different outcomes.

For a worker, the practical question isn't "which company is moral?" It's "which model am I entering?" An offer from a product company is usually an offer to do that company's work at that company's pay scale. An offer from a services firm is often an offer to be deployed to clients, sometimes at locations and on projects decided after you start.

What the broader rankings show

The Google-versus-Infosys contrast scales up to the whole program. The highest-paying sponsors and occupations are dominated by specialized, product- and research-heavy roles, while the highest-volume sponsors include many services firms. Looking at the best-paid occupations makes the role-mix point concrete: the top of the pay distribution is full of specialized scientific, managerial, and senior engineering work — not the entry-level consulting roles that drive raw application counts.

You can explore any individual employer's filing history on the site to see where a given company sits on this spectrum. Some firms blend both models — a consulting arm plus an internal product team — and their filings show a split personality as a result.

How to use this when you read H-1B numbers

A few takeaways that hold up regardless of the specific figures:

• Ask what the company sells. If it sells products, sponsorship is recruiting. If it sells labor, sponsorship is inventory. That tells you most of what to expect from the filings before you read a single number.
• Compare like with like. A wage only means something next to the same role, level, and metro area. Use the occupation and state pages to find the right benchmark instead of comparing two companies' overall averages.
• Volume and wage level are different questions. A firm can be the largest sponsor in the country and still file near the prevailing-wage floor; another can sponsor far fewer people at much higher pay. Don't let one number stand in for the other.
• The "model" is the unit of analysis. Most arguments about whether H-1B "helps" or "hurts" American workers are really arguments about the services model versus the product model. Naming which one you mean makes the conversation far more precise.

Takeaway

Google and Infosys aren't opposites because one is good and the other is bad. They're opposites because they use the same legal tool for two unrelated purposes: one to hire specific people for its own products, the other to supply billable labor to its clients. Once you can tell those two models apart, the rest of the H-1B data stops looking like a single confusing pile of numbers and starts looking like two distinct stories that happen to share a form. When you read about a sponsor — on this site or anywhere else — start by asking which story you're in.