Rai University Students: Using RICE Scoring to Prioritise Features in a Short Internship

RICE scoring ranks features by multiplying Reach by Impact by Confidence and dividing by Effort. In a 14-day internship it works only if you fix your scales on day one, write down the evidence behind every number, and add a sanity check that compares the ranking with your gut. This guide gives you a calibrated sheet, a ten-feature example and a defence script for your manager.

Why RICE suits a two-week product internship

A short internship gives you little time and little data. You probably cannot run experiments or wait for a release to see results. What you can do is make your reasoning visible. RICE forces you to turn opinions such as “this feels big” into four numbers, each tied to a stated source. A manager may disagree with a number, but they can see how you got it, and that is what makes you credible as an intern.

The formula is:

RICE score = (Reach × Impact × Confidence) / Effort

• Reach: how many users or events the feature touches in a fixed period.
• Impact: how much it moves the goal for each person reached.
• Confidence: how strong your evidence is, as a percentage.
• Effort: total work in person-days (or person-weeks) across design, engineering and testing.

The calibrated RICE scales to fix before you score anything

Most RICE sheets fail because every scorer uses private scales. Agree the scales below with your manager before you score, and keep them unchanged for all features. The values are a common convention, not a rule; adjust them if your team already has its own.

Rules that keep the scales honest

• Choose one goal for the whole sheet (for example, “increase completed sign-ups”). Impact is judged only against that goal.
• Use one time window for Reach. Mixing monthly and quarterly figures breaks the ranking.
• Never give 100% Confidence without a measured number behind it.
• Effort should come from the builder. If you must guess, say so and lower Confidence.
• Write the evidence in a column. A score with no note beside it is not defensible.

A 14-day plan for running RICE inside your internship

Treat prioritisation as a mini project with its own timeline. The free two-week online internship from CareerSwami follows the same 14-day length, so this plan fits it, and it works for any short product placement.

1. Days 1-2: Agree the goal, the Reach time window and the scales. Collect the feature list.
2. Days 3-5: Gather evidence: available analytics, five or more user conversations or reviews, and any support or feedback data.
3. Days 6-7: Ask engineers for effort estimates. Draft first-pass scores.
4. Days 8-9: Run the sanity check (below), then revise scores where the evidence supports it.
5. Days 10-11: Review with your manager or mentor. Record every change and the reason.
6. Days 12-14: Write a one-page recommendation and present it.

Worked RICE example with ten features and a sanity-check column

This example is illustrative. The app, the numbers and the evidence are invented to show the method, not drawn from any real product. Imagine a small food-ordering app for campus canteens, and the goal is “increase completed orders per month”. Reach is users per month, Effort is person-days.

Scores are calculated as Reach × Impact × Confidence ÷ Effort. For feature 2: 3,500 × 2 × 0.8 = 5,600, divided by 8 gives 700.

What the ranking tells you

1. Save UPI as default payment (700)
2. Fix checkout crash (600)
3. Reorder in one tap (480)
4. Search by dish name (286)
5. Order-ready notification (280)
6. Faster menu loading (253)
7. Redesign home screen (127)
8. Loyalty points (75)
9. Group ordering (32)
10. Dark mode (25)

Notice that the crash fix has the most reliable evidence even though it ranks second. A good recommendation says this out loud, and may suggest doing the crash fix first because its Confidence is 100%.

How to run the sanity-check column on your RICE ranking

The sanity-check column catches the cases where arithmetic produces a confident but wrong answer. Ask these questions for each row, then write a short note.

• Gut test: Does the rank match what the team expects? If not, which input explains the gap?
• Goal link: Does Impact refer to the chosen goal, or to a different benefit such as satisfaction?
• Evidence test: Is Confidence of 80% or more backed by at least one real source?
• Effort test: Did the builder give the estimate?
• Sensitivity test: If Impact drops one step or Effort doubles, does the feature fall out of the top five?
• Dependency test: Does a high-ranked feature depend on a low-ranked one?

Sensitivity is the strongest check. In the example, feature 10 (home screen redesign) holds a mid-table rank only because Impact is 2. At Impact 1 its score halves to about 63 and it drops to the bottom three. That fragility is worth flagging to your manager.

Rai University, Ahmedabad: internship process and the Corporate Resource Cell for product management interns

Rai University in Ahmedabad treats internship as part of the degree. Its placement cell page states that satisfactory completion of an internship is mandatory for the degree, so a short product management internship is something to plan, not an optional extra. The Corporate Resource Cell is the university’s route for this, and its overview page says it partners with organisations to arrange internships and placements and runs workshops, resume sessions and mock interviews.

The page lays out a seven-step process. The student requests a No Objection Certificate from the college, sends a profile to industries, waits for the industry to confirm a slot, submits a joining report, submits a training report, receives a certificate and finally submits a feedback form. For a student who plans to apply RICE scoring to a backlog, that sequence matters: the NOC and the confirmed slot come before the first day, and the reports come after the last.

The page also says the training report must be signed by both the Internship Supervisor and the Faculty Mentor. It is evaluated on originality, an adequate and purposeful write-up, organisation and presentation, learning experience, and the link between practical work and theory. A prioritisation exercise with clear reach, impact, confidence and effort estimates, written up with the reasoning, fits those criteria well.

The same page suggests finding internships through company websites, Internshala, nearby firms, direct emails, alumni, faculty and the placement office. It does not state credits or duration, so students should ask their school for those numbers.

What to confirm with Rai University before a product internship

• Request the NOC from the college before accepting any offer in writing.
• Ask the Corporate Resource Cell for the current internship guidelines and report format.
• Confirm the duration and credit value that apply to your school and semester.
• Agree with the supervisor who will sign the training report and when.
• Keep the RICE worksheet and decision notes as appendices for the training report.

How to defend your RICE scores to a manager

Expect pushback. Managers often challenge the number that surprises them most. Prepare a short script for each of your top three and bottom three features.

• State the goal first: “I scored everything against completed orders per month.”
• Name the evidence: “Reach comes from last month’s checkout funnel.”
• Admit weak spots: “Confidence is 50% because I have only two interviews.”
• Show sensitivity: “If Impact were 1 instead of 2, this drops to rank seven.”
• Invite correction: “If you have data that changes Reach, I’ll update the sheet.”

Changing a score when someone supplies better evidence shows judgement. Changing it simply because someone senior frowned does not, so record the reason for every revision.

RICE scoring limits an intern should state in the recommendation

• RICE ranks relative value, not strategic fit. A low-scoring compliance or security fix may still be mandatory.
• It rewards features that reach many users, so it can undervalue improvements for a small but important segment.
• Scores look precise but rest on estimates. Present them as a ranking with ranges, not as exact truths.
• Effort estimates are often too low. Ask what is excluded.
• Close scores (within about 15%) should be treated as ties and decided on dependencies or risk.

Template to copy: RICE sheet columns and recommendation outline

Use these columns in a spreadsheet:

• Feature name and one-line description
• Goal it supports
• Reach, with time window and source
• Impact (0.25 to 3), with reason
• Confidence (%), with evidence list
• Effort (person-days), with name of estimator
• RICE score (formula column)
• Rank
• Sanity-check note
• Change log (date, old value, new value, reason)

Your one-page recommendation can follow this outline: goal and scales; top three features and why; one feature you would deliberately delay; the riskiest assumption; and the next piece of evidence you would gather.

Frequently Asked Questions

Can I use RICE if I have no analytics data in a short internship?

Yes. Estimate Reach from proxies such as app store reviews, support tickets, survey responses or user counts for a related flow. Label them as assumptions and lower Confidence to 50% or below so the score reflects the weak evidence.

What is a reasonable Confidence score for a feature backed by five user interviews?

Usually 50% to 80%, depending on how consistent the interviews are and how well they match the goal. Five interviews that agree can justify 80%; mixed answers justify 50%. Only measured data should earn 100%.

Should Effort include design and testing or only engineering time?

Include everything needed to ship: design, build, testing, review and rollout. Using engineering time alone makes features look cheaper than they are and inflates their scores.

How do I handle two features with almost identical RICE scores?

Treat scores within roughly 15% as a tie. Break it using dependencies, risk, the strength of the evidence, or which feature teaches the team more quickly.

Is RICE better than simple value-versus-effort sorting?

RICE adds Reach and Confidence, so it separates a feature that helps many users from one that helps few, and a well-evidenced estimate from a guess. For a very small list, a value-versus-effort grid may be enough and quicker to explain.

What if my manager disagrees with the ranking?

Ask which input they disagree with, then compare evidence. Update the sheet if they provide better data and log the change. If the disagreement is about strategy rather than numbers, note it in your recommendation as a factor RICE does not capture.

How can I show a RICE exercise as part of my CareerSwami internship work?

Keep the scales page, the scored sheet, the change log and the one-page recommendation together as a single portfolio document. Describe the goal, your evidence and what you would test next, so a reader can follow your reasoning without extra explanation.

Does RICE work for non-software products or services?

Yes, if you can define the unit for Reach (customers, visits, orders) and estimate effort in person-days. The same scales apply, but be extra careful with Impact, which needs a clearly stated goal.

Related Product Management guides

• Mahatma Gandhi Chitrakoot Gramoday Vishwavidyalaya: Product Thinking for Rural-Facing Apps
• Parul University Interns: Running a Competitor Teardown and Writing It Up in Three Pages
• Central University of Andhra Pradesh Students: From Internship Tasks to Strong Resume Bullet Points
• All Product Management internship guides