Sarla Birla University: Compensation Benchmarking Basics for an HR Internship Project

Compensation benchmarking means comparing what a role pays in the market with what your organisation pays or plans to pay. For an intern project, the practical version is to collect salaries from public job postings, record the source of every number, and calculate a minimum, median, quartiles and maximum. The result is a range with a stated level of confidence, not an exact market rate, and the sample size always limits how far you can trust it.

What an HR intern is actually benchmarking, and what the answer can be used for

Benchmarking answers one question: what do comparable employers appear to offer for a comparable role? “Comparable” has three parts, namely the job content, the location and the employer type. A fresher analyst in Ranchi at a small company is not comparable with an experienced analyst in a metro at a large firm, even if both are called “analyst”.

The output of an intern project is usually one of these:

• A reference range that a manager can use as a starting point for an offer discussion.
• A check on whether an existing pay band looks far from posted offers.
• A methods note that explains how the range was built, so someone else can repeat it next year.

It is not a pay recommendation. Pay decisions depend on budget, internal fairness, performance, skills and legal requirements, and those belong to the employer. An intern’s job is to make the market evidence clear and honest.

Where public job postings fit, and where they fall short

Job postings are easy to find and free to read, which makes them a good teaching source. They also have limits you must note in the worksheet.

• Many postings give no pay figure. The ones that do may not be typical of the ones that do not.
• Posted pay is an advertised offer, not an agreed one. Final offers can be higher or lower.
• Figures may be monthly, annual, fixed, or total including variable pay. Convert them to one basis and write down the conversion.
• Job titles are loose. Check duties and required experience before including a posting.
• Postings go out of date. Record the date you saw each one.
• Aggregator sites may repeat or estimate figures. Prefer the employer’s own posting and note when a number looks estimated.

For stronger evidence, an employer may subscribe to a paid salary survey. Public postings are a supplement, not a substitute.

The compensation benchmarking worksheet: columns to fill for every posting

Create one row per posting. The columns below make the data auditable and comparable.

Keep a separate tab for excluded postings and the reasons, so the choices you made are visible.

Cleaning the data before you calculate

1. Standardise units. Convert all pay to the same period. A monthly figure becomes annual by multiplying by twelve, and you note whether it is fixed pay or total pay.
2. Handle ranges. For a posted range, record both ends and use the midpoint in the main calculation. Also run the maths on the lows and highs separately to see how sensitive the result is.
3. Remove duplicates. The same vacancy often appears on several sites. Count it once.
4. Check role fit. Exclude postings with very different duties or experience levels from your benchmark role.
5. Flag outliers. Do not delete a very high or very low value just because it is unusual. Record it, check whether it is a typo or a different kind of role, and decide with a written reason.

Calculating a pay range: minimum, median, quartiles and maximum

With the cleaned midpoints in a column, calculate these statistics. Most spreadsheets provide a function for each.

The median and the quartile range are usually more informative than the mean for small and uneven samples. When you report, give the range from Q1 to Q3 as the “typical” band and mention the minimum and maximum separately.

Illustrative worked example: nine postings for an HR executive role

All figures below are invented to show the maths and are not real market data. Suppose you collect nine annual fixed-pay midpoints, in lakh rupees per year, for a fresher HR executive role in one city.

The values are already sorted from low to high. The calculations are:

• Count: n = 9.
• Minimum and maximum: 3.6 and 7.2.
• Median: the fifth value, 4.8.
• Mean: the total is 44.7, divided by 9, which gives about 4.97.
• Q1: the median of the lower four values (3.6, 4.0, 4.2, 4.5), which is 4.1.
• Q3: the median of the upper four values (5.0, 5.4, 6.0, 7.2), which is 5.7.
• Spread: 5.7 minus 4.1 equals 1.6.

A sensible write-up would say: “In nine postings, the median annual fixed midpoint was 4.8 lakh, with the middle half between 4.1 and 5.7 lakh. The full range was 3.6 to 7.2 lakh. The sample is small, so treat this as indicative.” Different spreadsheet programs compute quartiles slightly differently, so state the method you used.

The mean (4.97) is higher than the median (4.8) because of the 7.2 value. If you checked P09 and found it was a role with more responsibilities, you would exclude it with a written reason and recalculate, rather than quietly dropping it.

Sample size cautions that belong in every benchmarking note

• Small n means wide uncertainty. With fewer than about ten postings, one unusual value can move the median. Say how many postings you used, and avoid precise-sounding figures.
• Selection bias. Employers that publish pay may differ from those that do not.
• Time sensitivity. Pay moves, so show the collection period.
• Geography. Mixing cities can hide real differences. Where you can, benchmark one location at a time.
• Do not extrapolate. A benchmark for a fresher role says nothing about a senior one.

A useful rule is to write a confidence label next to the final range: low if you have fewer than about ten clean postings or mixed sources, medium for a larger sample from varied employers, and high only when you have corroborating data from a proper survey.

Source notes: how to document every number so someone else can check it

Every figure in your final table should be traceable. Use a source note format like this one.

1. Source type: employer page, job board, government portal or other.
2. URL and the date accessed.
3. The exact pay wording copied from the posting.
4. Any conversion you applied, written as a formula.
5. Whether the figure was stated by the employer or appears to be an estimate by the site.
6. Your include or exclude decision and reason.

Store screenshots only if the site’s terms allow it, and never record personal data of recruiters or candidates.

Presenting the benchmark: a one-page summary template

Sarla Birla University students: fitting a benchmarking project into the MBA internship requirement

Sarla Birla University, which appears under the spelling Sarala Birla University on its official site, is a private university based near Ranchi in Jharkhand. A web page that mirrors the university’s MBA information, hosted on a different address from the official domain, says the MBA is a two-year, four-semester programme with HR listed among its specialisations. The same page says a summer internship between the first and second year is compulsory, that students submit a summer internship report, and that they may find hosts through the campus process or on their own.

The page also describes a Training and Placement Cell that collaborates with organisations on internships and training. Because that information appears on a mirror page and was not confirmed on the official domain, treat it as a pointer rather than a rule, and check the current version with the university before planning around it.

For a benchmarking project, the practical fit is straightforward. A compensation benchmarking exercise is a bounded, report-friendly task, and it can serve as one component of a longer HR internship rather than replace it. Because the university’s own rules on what counts as an acceptable internship project were not verified, do not assume that a standalone desk-based exercise will be accepted as the internship itself.

If you want a practice run before the compulsory internship, the free two-week online internship from CareerSwami can host a small version of this worksheet, and offers an offer letter and a completion certificate after successful completion.

Verification steps for Sarla Birla University students planning a benchmarking project

1. Ask the MBA department office for the current internship guidelines and the report format, and note the year printed on them.
2. Ask in writing whether the summer internship may be fully or partly remote, and whether a compensation project in an HR function is acceptable.
3. Ask the Training and Placement Cell how host organisations are approved and what paperwork they must sign.
4. Confirm with your faculty guide whether public-data projects need a host organisation’s letter.
5. Check that the HR specialisation or major and minor choice you plan matches the current programme structure, since the mirror page describes the structure in two different ways.

Handling salary data ethically during an internship

• Use public information for the benchmark. Do not collect pay details that colleagues or candidates shared privately.
• If the employer gives you internal pay data, keep it separate from the public data and never publish it in a university report.
• Do not present benchmark figures as a statement of what any individual should be paid.
• Follow the terms of any website you collect data from, and do not scrape sites that forbid it.

Frequently Asked Questions

How many job postings do I need for a compensation benchmark?

There is no fixed number, but fewer than about ten clean postings gives a low-confidence result. State the count and the confidence label in your note.

Should I use the mean or the median salary?

Use the median as your headline figure, and show the quartiles. The mean is pulled by unusually high or low postings, especially in small samples.

How do I convert a monthly salary in a posting to an annual figure?

Multiply by twelve, and record whether the posting meant fixed pay or total pay including variable components. Note the formula next to the converted value.

What do I do with a posting that gives a salary range?

Record both ends, use the midpoint in the main calculation, and also run the maths on the lows and highs separately to see how sensitive the result is.

Can I use a salary estimate shown by a job website?

Treat it as low confidence and label it as an estimate. Prefer figures stated by the employer in the posting itself.

Does a benchmarking project count as my Sarla Birla University summer internship?

That depends on the university’s current rules, which you should confirm in writing with your department. A benchmarking exercise is more likely to work as one task within a host-based HR internship.

How should I handle outliers in salary data?

Check them first, since they may be typing errors or different roles. Keep them in a separate tab with the reason you excluded or retained them, and show the result both ways if the choice matters.

Related Human Resources guides

• Dr. C.V. Raman University: Final-Year Students' Guide to Converting an HR Internship Into a Job Offer
• Thiruvalluvar University: Communicating in Tamil and English in HR Internship Tasks
• Madhabdev University: Organising a Small Workplace Event as an HR Internship Project
• All Human Resources internship guides