Company Tiers and the Reach, Target, Safe Mix

"Should I apply to Google or to TCS?" is the wrong question, and it is asked every placement season. It treats companies as a ladder of prestige, when what matters for your search is something more practical: what does each kind of company test, and does your evidence match it?

A student with average problem-solving speed but an excellent deployed project may struggle in a large company's timed assessment and do very well with a startup founder who opens the project during the call. A student who is fast at aptitude tests but has built nothing may clear a services company's drive easily and get nowhere with startups. Neither is "better". They are aimed at different doors.

• Five broad groups of employers that hire freshers in India, and what each usually tests.
• How to match your profile scores to a target group.
• The reach, target and safe split, and how to adjust it from your results.
• How data, QA, support and DevOps roles differ from software developer roles.
• How to classify an unfamiliar company quickly, with AI and a check.

These groups describe typical patterns. Individual companies, and even teams within a company, differ, and processes change over time. Use the groups to decide where to start, then read each company's own job page and recent candidate reports (article 26) before you prepare.

The India offices of large global technology, internet and financial-technology companies. They hire relatively few freshers, and many candidates want those roles.

• What they usually test: several problem-solving rounds at medium to hard difficulty, follow-up questions on complexity, sometimes basic design, and a behavioral round. Some publish their values or principles, which their behavioral questions are built on (article 36).
• What helps off campus: a strong problem-solving record, a referral, and proof of work that shows real engineering judgement.
• Watch out for: spending your whole search here. Even strong candidates face long odds because there are few seats.

Well-known Indian companies whose main business is their own software product, in areas such as payments, commerce, delivery, fintech, developer tools and SaaS.

• What they usually test: problem-solving at medium difficulty, often a machine coding or low-level design round (article 31), sometimes a light system design discussion, and a conversation about why you want to work in their domain.
• What helps: a project in or near their domain, clean code, a referral from an engineer.
• Watch out for: neglecting machine coding. Many freshers prepare only for problem-solving rounds and meet a 90-minute "build this" round unprepared.

Smaller companies, often between seed and growth funding, with tens to a few hundred people.

• What they usually test: practical ability. A take-home assignment or live build is common, along with a conversation with a founder or engineering lead about your projects. Problem-solving rounds tend to be easier or absent.
• What helps: proof of work above everything else, curiosity about their product, and the ability to learn quickly on your own.
• Watch out for: company health. Startups can run short of money. Article 41 shows a quick health check before you accept.

Global capability centres (GCCs) are in-house technology and operations centres that global banks, retailers, manufacturers and insurers run in Indian cities. They are not consulting firms: they build and run software for their own parent company, and many hire freshers.

• What they usually test: a mix: an online assessment, a technical interview on fundamentals and your projects, and an HR round. Rounds are often more structured and less intense than at product companies.
• What helps: solid fundamentals, a clean resume, clear communication, and sometimes knowledge of the parent company's industry.
• Watch out for: overlooking them. They are often less crowded than well-known product companies.

Large IT services and consulting firms that build and run software for clients, such as TCS, Infosys, Wipro, HCLTech, Cognizant, Capgemini, Accenture and LTIMindtree. Many have more than one entry track for freshers, with a higher-paid "digital" or "specialist" track that has a harder test.

• What they usually test: an online test with aptitude (quantitative, logical, verbal), basic programming and sometimes a coding section; then technical and HR interviews on fundamentals.
• What helps: steady aptitude preparation, clean basics in one language, meeting the published eligibility, and good communication.
• Watch out for: bonds and training agreements (article 38) and gaps between an offer and the joining date (article 41).

Use your scores from article 2 (0 to 3 in each changeable area).

These are starting points, not verdicts. After three or four weeks your tracker (article 5) will tell you whether your target is right: if you get almost no responses from your target group, step down one group or fix the earliest gap; if you clear most assessments easily, step up.

If you apply only to reach companies, you collect silence and lose heart. If you apply only to safe companies, you may accept an offer well below what you could have reached. A simple split protects you from both:

• Reach, about 20 per cent: one group above your target. Low odds, but occasionally they work, and every assessment teaches you something.
• Target, about 60 per cent: where your evidence matches the bar. Most of your effort goes here.
• Safe, about 20 per cent: one group below. They keep interviews coming, build interview practice and give you a real fallback.

The percentages are a rule of thumb, not a law. Adjust them from your own numbers every few weeks. If safe applications are producing interviews and target ones are not, the evidence says your "target" is really a reach for now.

Rahul is a 2025 graduate. His scores: coding 2, proof of work 2 (a deployed expense tracker with tests), fundamentals 2, resume 2, communication 1. No constraints rule out drives except his batch year.

• Target: Groups 3 and 4: startups and capability centres hiring "0 to 2 years" for backend or full-stack roles.
• Reach: Group 2 roles in fintech, where his expense tracker is relevant, ideally with referrals.
• Safe: Group 5 lateral or off-campus roles open to 2025 graduates.
• His first fix: communication, because it is his lowest score and every group interviews.

In a 50-application month that is roughly 10 reach, 30 target and 10 safe. Article 4 shows how to turn this split into a list of companies.

Many freshers apply only for "SDE" or "software engineer" titles. Other entry roles hire steadily, and some are a sensible way into the industry, especially for non-CS graduates. Their rounds differ:

Two cautions. First, read the job description, not just the title: "associate engineer" can mean development, testing or support. Second, if you take a non-development role as a first step, keep building and keep a record of the technical work you do in it, so a later move stays possible.

You will find many companies you have never heard of. Classify each in two minutes:

1. What do they sell? Their own software product, services to clients, or software for their parent company?
2. How big and how funded? The company's own "about" page, its LinkedIn page (employee count) and reputable news.
3. What does the role ask for? Years of experience, skills, eligibility.

AI can speed this up, but it may be out of date or wrong about smaller companies. Ask it to work only from text you paste:

Here is text I copied from a company's About page, careers page and one news article.Using ONLY this text:1. Does this company mainly sell its own software product, provide services   to clients, or build software for a parent company? Quote the evidence.2. What size and funding stage does the text suggest? Say "not stated" if it   is not in the text.3. Which of these groups fits best: global product, Indian product, funded startup,   global capability centre, IT services. Give one sentence of reasoning.4. List anything I should verify before applying.[paste text]

How to check it: confirm the employee count on the company's own LinkedIn page, and the funding claim in a news article with a date. If the assistant answered from memory instead of from your text (for example, mentioning a product your text never mentioned), ask it again to use only the pasted text.

• Choosing by brand name alone. A famous company that tests something you cannot show yet is a reach, however much you want it.
• Applying only one step below where you could reach. Keep the 20 per cent reach: it costs little and sometimes works.
• Ignoring capability centres and startups. They often have fewer applicants per role.
• Taking an assistant's claims about a small company at face value. It may be guessing. Ground it in pasted text.
• Never revisiting the split. Your tracker should change your targets every few weeks.

• Next: Funnel Maths and Your Master Target List works out how many companies you need and builds the list.
• For the wider fresher job search, see How to Get Your First Job as a Fresher in India: A Step-by-Step Guide.
• If you are moving into tech from another field, see How to Switch to an IT or Data Career at 25 to 35 Without Starting Over.