The stipend question gets asked once per hire and answered differently each time, until the day two interns doing the same work compare numbers. A stipend policy is how you avoid that conversation — and it also makes hiring faster, because a manager can price a role without an approval thread.
This page covers how to build the bands, what deviations to allow, and the policy text itself.
How to build your bands
| Step | What to do | Common mistake |
|---|---|---|
| 1. Group roles into families | Four to six families — engineering, data, design, marketing, business, operations. Not one band per job title. | A band per title becomes unmaintainable within two cycles. |
| 2. Set a tier-1 metro base per family | Use live market data for the role, not last year's internal number. | Anchoring on what you paid before, in a market that moved. |
| 3. Apply city multipliers | One multiplier per city tier, published. Roughly 0.75–0.9 for tier-2 against a tier-1 base. | Paying tier-2 rates for a remote role that competes nationally. |
| 4. Add a duration adjustment | Slightly higher for six-month and longer internships; slightly lower for one-month. | Paying a one-month intern a full band rate when they barely clear ramp-up. |
| 5. Publish the bands internally | Managers should be able to price a role without asking anyone. | Keeping bands in HR only, which reintroduces the approval thread you were removing. |
| 6. Review annually | Early-career rates move. A band left for three years is a recruiting problem you cannot see. | Reviewing only when a candidate declines. |
The sample policy — copy this
[COMPANY NAME] — INTERNSHIP STIPEND POLICY Effective: [date] | Owner: [People Team] | Review: annually 1. PRINCIPLE All internships at [Company] are paid. Stipends are set by role family, city and duration — not by negotiation with the individual candidate. 2. BANDS Stipends follow the published band table (Annexure A), which sets a monthly range for each role family in each city tier. The published band is stated in the job posting as a number. 3. SETTING THE FIGURE WITHIN A BAND 3.1 Default to the midpoint of the band. 3.2 The upper half of the band may be used where the role requires a specific scarce skill, or where the internship is six months or longer. 3.3 The lower half is used only for internships shorter than [8] weeks. 4. DEVIATIONS 4.1 Any figure outside the published band requires written approval from [role]. 4.2 The reason is recorded. Deviations are reviewed at each annual band review. 5. EQUAL TREATMENT Two interns in the same role family, city tier and duration receive the same stipend. Individual negotiation does not change the figure. 6. EXTENSIONS AND CONVERSIONS 6.1 An extension continues at the same stipend unless the band has been revised. 6.2 On conversion to full-time, compensation follows the full-time band for the role; the internship stipend does not anchor it. 7. PAYMENT Stipends are paid by the [Nth] of the following month by bank transfer, with a payment advice to the intern. 8. ACADEMIC-CREDIT INTERNSHIPS Internships taken for academic credit are paid on the same bands. Academic credit does not reduce the stipend. ANNEXURE A — BAND TABLE Role family | Tier-1 metro | Tier-2 city | Remote -------------------|----------------|----------------|---------------- Engineering / Data | ₹[x] – ₹[y] | ₹[x] – ₹[y] | ₹[x] – ₹[y] Design | ₹[x] – ₹[y] | ₹[x] – ₹[y] | ₹[x] – ₹[y] Marketing | ₹[x] – ₹[y] | ₹[x] – ₹[y] | ₹[x] – ₹[y] Business / Sales | ₹[x] – ₹[y] | ₹[x] – ₹[y] | ₹[x] – ₹[y] Finance / Ops | ₹[x] – ₹[y] | ₹[x] – ₹[y] | ₹[x] – ₹[y]
Clause 5 is the whole point
Equal treatment within a band is what a stipend policy is actually for. Without it, the intern who negotiates hardest is paid most, which correlates with confidence and background rather than with contribution — and the disparity surfaces within weeks, because interns in the same cohort talk. The cost of the disparity is not the money; it is the cohort finding out that the company's stated fairness is negotiable.
Frequently asked
Should interns doing the same work be paid the same?+
Yes, and a policy is how you make that true rather than aspirational. Individually negotiated stipends reward negotiation confidence rather than contribution, and the disparity always surfaces — interns in a cohort compare notes within the first fortnight.
Should we pay less for an academic-credit internship?+
No. Academic credit is between the intern and their university; it does not reduce the work you receive. Reducing the stipend because the intern also gets credit is a policy that is difficult to defend when stated out loud.
How often should stipend bands be reviewed?+
Annually at minimum. Early-career rates move quickly in India, and a band left unreviewed for three years shows up as a recruiting problem — slower fills, more declines — long before anyone connects it to the band.
Does the internship stipend set the conversion salary?+
It should not, and the policy should say so. Anchoring a full-time offer on an internship stipend produces offers below market for your own best candidates, who then decline and go elsewhere — the most expensive possible way to lose a hire you already assessed for six months.
What if a candidate asks for more than the band?+
Say the band is published, explain what determines position within it, and hold. Almost all early-career candidates accept a clear, consistent answer. What they do not accept, in retrospect, is discovering that someone else asked and got it.