How to write a cover letter for an internship (2026 template + 6 industry examples)

There are a lot of internship cover letter guides on the internet, and most of them are template farms — generic four-paragraph skeletons with mad-lib variables for "company name" and "role." This isn't that. The data on what actually works (and what doesn't) is good now, and the structure that follows from the data is specific and short.

Three things to know before we start:

• 83% of hiring managers read most cover letters they receive, and 94% say cover letters influence interview decisions (ResumeGenius 2026 survey).
• The optimal length is 250-300 words — the consensus recommendation across ResumeGenius, Indeed, Kickresume, and Resume.io's published guides. There's no peer-reviewed study on this specific question — see the dedicated length post for an honest breakdown of the sources.
• 66% of recruiters spend under 30 seconds reviewing a cover letter before deciding whether to continue (ResumeGenius 2026 hiring manager survey).

Everything below follows from those three facts.

The structure: four short paragraphs, ~250 words

The four-paragraph structure isn't a template-farm convention — it's what fits in 250 words while doing the four jobs a cover letter has to do. Each paragraph has one job. If a paragraph is doing two jobs, it's too long. If it's doing zero jobs, cut it.

Total: 170-270 words. Aim for the middle of the range.

Paragraph 1 — the hook

The opening sentence does most of the work. It needs to:

1. Name the role exactly as posted.
2. Name the company.
3. Give one specific reason you want to work there that you couldn't paste into 50 other letters.

The third part is what separates the letters that get read from the letters that get filed. "I'm passionate about your mission" is filed. "I'm applying because of the engineering blog post on how you handled the Black Friday capacity surge — that's the kind of incident-response work I want to learn" is read.

Bad opening: I am writing to express my interest in the Software Engineering Internship at your esteemed company.

Good opening: I'm applying for the Software Engineering Internship at Stripe because of the engineering blog post on the Black Friday capacity surge — the way the team scaled the cursor pagination on charges was exactly the systems-thinking I want to learn.

The good version is 38 words and immediately demonstrates that you've actually read something the company published. That signal is rare in 2026 because most cover letters are AI-drafted from a sparse prompt that didn't include any specific company research.

Paragraph 2 — the proof

This is the longest paragraph and the most important one. The job: pick one experience or project from your resume and explain it with the detail your resume couldn't fit. The cover letter version of a project bullet has the why, the how, and the result with context.

Resume bullet: Built a Postgres-backed feature flag service in TypeScript that 4 product teams adopted; rollout time dropped from days to under an hour.

Cover letter version of the same project: Last summer at Acme, I built the company's first feature flag service. The team had been deploying flag changes by editing a YAML file in the repo and waiting for the next release — sometimes 3 days for a flag flip. I shipped a Postgres-backed service in TypeScript with a small admin UI, and four product teams adopted it within the first month. Rollout time dropped from 2-3 days to under an hour. The interesting failure mode I didn't anticipate was webhook delivery races during deploys — that's the part of the work I'd want to dig deeper on at Stripe.

Three things this version does that the resume bullet can't:

• Context — explains why the work mattered (3-day rollouts were a real problem).
• Specificity — names the technical decision (Postgres-backed, admin UI in TypeScript).
• Forward connection — links the experience to the company you're applying to.

If you don't have full-time work experience, the proof paragraph uses a project, hackathon, club initiative, or course assignment instead. The structure is identical.

Paragraph 3 — the fit

One line connecting your trajectory to the role's trajectory. This is where many cover letters bloat into a four-paragraph mini-autobiography. Resist that. One sentence:

I've been moving from full-stack feature work toward systems and infrastructure over the past two years, and the Stripe role's mix of payments-domain work and capacity engineering is the next step I want.

That's 35 words. It tells the reader where you've been, where you're going, and why this specific role is the bridge. Don't write three sentences when one works.

Paragraph 4 — the close

A specific next step in 20-40 words. Avoid:

• Passive close: I look forward to hearing from you.
• Faux confidence: I would be a perfect fit for your team.
• Corporate boilerplate: Please don't hesitate to contact me at your earliest convenience.

Use instead:

I'd welcome the chance to walk you through the feature flag service at a 20-minute call next week. Thanks for reading.

Specific (you've named the duration), low-pressure (a 20-minute call costs them less than a 60-minute one), and signs off like a person who'd be easy to work with.

Six worked examples by industry

Each example below is a real-shaped cover letter at 250-280 words for an entry-level/internship application. Names and companies are illustrative; the structures are designed to be copied and adapted.

1. Software engineering intern

Dear Stripe Engineering team,

I'm applying for the Backend Software Engineering Internship at Stripe because of the engineering blog post on how the team handled the Black Friday capacity surge — the cursor-pagination work on the charges API was exactly the systems-thinking I want to learn.

Last summer at Acme, I built the company's first feature flag service. The team had been editing a YAML file in the repo and waiting for the next deploy — sometimes 3 days for a flag flip. I shipped a Postgres-backed service in TypeScript with a small admin UI, and four product teams adopted it in the first month. Rollout time dropped from 2-3 days to under an hour. The interesting failure mode I didn't anticipate was webhook delivery races during deploys — that's the part of the work I'd want to dig deeper on at Stripe.

I've been shifting from full-stack feature work toward systems and infrastructure over the past two years, and Stripe's mix of payments-domain depth and capacity engineering is the next step I want.

I'd welcome the chance to walk you through the feature flag service at a 20-minute call next week. Thanks for reading.

— Aanya Sharma

2. Data analyst intern

Dear Notion Data team,

I'm applying for the Data Analyst Internship at Notion because of the public retro you posted on the activation-funnel rebuild — the SQL framing of "what defines an activated team" is the kind of analyst-judgment problem I want to spend a summer on.

In a class project last semester, I worked with a 12M-row event log from a public dataset to identify the top three onboarding drop-off points for a fictional B2B product. I wrote SQL against a Postgres replica, built a dbt model that refreshed nightly, and presented findings to a panel of 4 industry mentors. Two of the mentors said the third drop-off point — at the team-invite step — wasn't one their teams had identified despite having larger datasets. The framing I used (event-funnel + cohort retention overlay) is something I'd bring into the Notion role.

I'm targeting a career in product analytics specifically — work that connects measurement to product decisions, not dashboards in isolation.

Could we set up a 20-minute call next week? I'd like to walk through the dbt project. Thanks for reading.

— Marcus Chen

3. Marketing intern

Dear HubSpot Marketing team,

I'm applying for the Content Marketing Internship at HubSpot because of the recent series on B2B SEO experiments — specifically the post on canonical answers vs comparison content, which I've been trying to apply in a side project.

For the last three months I've run a small SEO experiment for my college's career office: 8 blog posts on internship-prep topics, all structured with TL;DR blocks, FAQ schema, and citation-ready paragraphs. The blog went from 0 to 1,400 monthly organic visitors in 12 weeks. The interesting finding: posts with comparison tables outperformed listicles by 2.3x in time-on-page despite shorter average length. I'd want to keep running these kinds of experiments in the HubSpot Internship.

I'm specifically interested in B2B SaaS content marketing because the audience cares about substance over style — that's the writing I want to do.

I'd welcome the chance to walk through the SEO experiment data at a 20-minute call next week. Thanks for reading.

— Priya Iyer

4. Finance intern

Dear JPMorgan Investment Banking team,

I'm applying for the Investment Banking Summer Analyst position at JPMorgan because of the deal team's work on the recent healthcare-IT M&A transaction — the cross-border financing structure was the kind of complexity I want exposure to early in my career.

In my Corporate Finance course, I led a 4-person team's full pitch on a hypothetical acquisition of a mid-cap SaaS company. We built a 5-year DCF model, a comp set of 12 public-market peers, and a sensitivity analysis on the synergy assumptions. Our valuation came in 18% below the market quote of the target. The professor — a former Goldman MD — used our deck as the model answer in the next year's section. I'd want to bring the same modeling depth to live deal work this summer.

I'm targeting investment banking specifically because the volume and pace of pattern recognition early in your career has no equivalent in another finance track.

Could we schedule a 20-minute call next week? I'd like to walk through the pitch deck. Thanks for reading.

— Daniel Park

5. Design intern

Dear Linear Design team,

I'm applying for the Product Design Internship at Linear because of the recent post on the keyboard-first interaction model — that's the bar I'd want to be designing toward, not the consumer-app default.

Last semester I redesigned the issue-triage flow for a small open-source project I contribute to. The original flow took 14 steps to triage a new issue (label, assign, prioritize, link, comment); my redesign collapsed it to 5 with a keyboard-driven palette. I tested the redesign with 8 maintainers; the average triage time dropped from 4.2 minutes to 1.6 minutes per issue. The maintainers shipped a stripped-down version of the design two weeks later. I'd want to bring the same workflow-first thinking to Linear.

I've been moving from generalist UI design toward systems-and-tools design specifically over the past year. Linear is the company doing this work at the highest standard.

I'd welcome the chance to walk you through the triage redesign at a 20-minute call next week. Thanks for reading.

— Sara Kim

6. Research assistant (lab)

Dear Dr. Patel and the Computational Biology Lab,

I'm applying for the Research Assistant position in the Patel lab because of the recent paper on transformer-based protein structure prediction — the failure-mode analysis on disordered regions was exactly the kind of careful boundary-mapping I want to learn.

In Professor Lin's molecular biology course last semester, I led a small replication of the published method on a held-out test set of 240 proteins. I implemented the model in PyTorch, ran the evaluation pipeline, and produced a written analysis identifying three structural motifs where the model's predictions consistently underperformed. Professor Lin's feedback was that the disordered-region failure cases I flagged were the same ones the original authors mentioned in their follow-up workshop talk — so the boundary I mapped was real. I'd want to bring this kind of careful negative-result work to the Patel lab.

I'm planning to apply to PhD programs in computational biology in two years; this position would be the experience I cite in those applications.

Could we schedule a 20-minute meeting next week? Thanks for reading.

— Rohan Mehta

Common mistakes (and how to fix each)

1. Generic opening. "I am writing to express my interest..." — fix by naming one specific thing the company has published, shipped, or done that connects to the work.

2. Restating the resume. "As you can see from my resume, I have experience in..." — fix by picking one project and going deep on a single dimension the resume couldn't fit.

3. Buzzwords with no content. "I am passionate about driving results in a fast-paced environment" — fix by replacing every buzzword sentence with one specific noun (a project, a tool, a decision, a number).

4. Inventing claims. AI-drafted letters that say "I led a team of 8" when you led a team of 2. Fix by writing the cover letter from real material only; the moment you're tempted to inflate, ask whether you'd defend the claim in an interview.

5. Length over 350 words. The data is clear: 250-300 is the sweet spot. Fix by cutting paragraph 3 (fit) to one sentence and paragraph 1 (hook) to one sentence + one specific reason.

6. Generic close. "I look forward to hearing from you at your earliest convenience" — fix by naming a specific time frame and a specific call duration.

Try the AI cover letter — first one is free

If you've drafted the bullet points but the four paragraphs aren't coming together, Laxu Resume's tailoring tool generates a per-application cover letter alongside the resume rewrite. The first tailoring is free with full output unlocked — including the cover letter, ~250 words, referencing the JD and your actual experience without inventing claims.

For the deeper guides on the writing craft, see:

• How long should a cover letter be? (the data-backed answer)
• Cover letter examples by industry: 8 templates that actually worked
• How to tailor your resume to a job description
• What is the STAR method? Definition, template, and 12 sample bullets