· PM Editorial · Product Sense · 5 min read
Design a Product for Senior Citizens: Metrics and North Star
How to choose a defensible north star metric, supporting input metrics, and counter metrics for a senior-focused product, plus A/B testing pitfalls specific to this population.
Why Metrics Are Where Most Answers Fall Apart
Ask ten candidates to design a product for senior citizens and most will land on a reasonable solution. Ask them to define how they’d measure success, and half go blank or default to a generic “DAU” answer that ignores everything specific about the population they just spent five minutes describing. Updated for July 2026, this article gives you a repeatable metrics framework for this prompt — and for demographic-based product sense questions generally.
This is part three of a four-article series. If you haven’t yet, start with the complete answer structure and the segmentation framework. For structured PM interview frameworks, see The 100x Product Manager Interview Playbook (Amazon: https://www.amazon.com/dp/B0DBC1FQWH?tag=sirjohnnymai-20).
Start From the Problem, Not a Generic Metric Template
The biggest metrics mistake in this prompt is reaching for engagement metrics that could apply to any consumer app — DAU, session length, retention — without connecting them back to the specific problem you chose to solve. If your solution targets isolation among active retirees, “time in app” is actually a weak, even perverse, proxy: a lonely senior spending three hours a day scrolling isn’t necessarily a win. Time in app must be justified against the actual outcome you’re chasing.
Choosing a North Star Metric
A north star metric should capture the core value delivered, be a single number the whole team can rally around, and be resistant to gaming. For a senior-isolation-focused product, walk the interviewer through 2-3 candidates before landing on one:
| Candidate north star | Why it might work | Why it’s flawed |
|---|---|---|
| Daily active users | Simple, comparable across products | Doesn’t capture whether isolation actually decreased |
| Weekly meaningful social interactions per active user (calls completed, messages exchanged with a real contact, video chats held) | Directly tied to the core outcome (reduced isolation) | Requires defining “meaningful,” slightly harder to instrument |
| Self-reported loneliness score (periodic in-app survey, e.g. UCLA Loneliness Scale short form) | Most directly tied to true outcome | Survey fatigue, self-report bias, slower feedback loop than behavioral metrics |
The strongest interview answer picks “weekly meaningful social interactions per active user” as the north star, and mentions the self-reported loneliness score as a slower-moving validation metric run quarterly to confirm the north star is actually tracking the real outcome, not just a proxy behavior.
Input Metrics: What Drives the North Star
Once you have a north star, break it into 2-4 input metrics your team can actually act on day to day:
- Contact list connection rate: % of users who’ve connected at least one family member or friend contact in-app
- Prompted-call acceptance rate: % of contextual “call your daughter” nudges that result in a completed call
- Time-to-first-interaction: how quickly a new user has their first meaningful contact after onboarding
- Caregiver co-setup rate: % of accounts set up with help from an adult child or caregiver, since this correlates with long-term retention in this population
Naming input metrics shows the interviewer you understand the causal chain between what the team builds and what the north star measures — not just a metric in isolation.
Counter Metrics: What Could Go Wrong
This is the section that separates a good answer from a great one. Every metrics framework needs counter metrics — signals that the product is winning on the north star while quietly causing harm elsewhere.
| Risk | Counter metric to watch |
|---|---|
| Product becomes a crutch that reduces real-world in-person contact | % of interactions that are in-app vs. facilitated in-person meetups |
| Family members feel obligated/guilted into responding to nudges, causing resentment | Opt-out rate on receiving-end (family member) accounts |
| Scam/fraud exposure increases as seniors engage more with unfamiliar contacts | Reported fraud/suspicious contact incidents per 1,000 users |
| Overly aggressive nudging increases anxiety rather than connection | Uninstall rate correlated with nudge frequency |
Mentioning even two of these unprompted is a strong signal. It shows you’re not just optimizing a number, you’re thinking about the actual human on the other end of the metric — the core skill product sense interviews are testing for.
A/B Testing Considerations Specific to This Population
Testing methodology questions often follow the metrics discussion, and seniors as a test population have real quirks worth naming:
- Smaller effective sample sizes: adoption curves are slower, so tests may need to run longer to reach significance
- Higher variance in tech comfort within the segment: consider stratifying by digital literacy, not just standard demographic splits, or a single test can wash out real signal
- Ethical sensitivity around vulnerable populations: be cautious A/B testing anything touching safety features (fall detection, medication reminders) — some teams treat these as non-negotiable rollouts rather than experiments
- Caregiver-mediated adoption: because a caregiver often sets up the account, your “user” in the experiment might behaviorally be two people, which can muddy interpretation of engagement metrics
Putting It Together: A Sample Metrics Close
A strong closing statement in the interview sounds like this: “My north star is weekly meaningful social interactions per active user. I’d drive that through onboarding metrics like contact-connection rate and nudge-acceptance rate, and I’d watch counter metrics like in-app-only interaction share and family opt-out rate to make sure we’re building genuine connection, not just engagement. I’d validate the north star quarterly against a self-reported loneliness score to confirm we’re tracking the real outcome.”
That’s roughly 60 seconds of talk time and it hits north star, input metrics, counter metrics, and validation — the full stack interviewers are listening for.
Common Mistakes
- Picking DAU/MAU as north star without connecting it to the actual problem
- No counter metrics at all — this is the single most common gap
- Ignoring the caregiver as a second user whose behavior also needs metrics
- Suggesting standard fast-iteration A/B testing on safety-critical features without caveats
Closing Thought
Metrics questions in product sense interviews aren’t really about analytics — they’re a test of whether you understand what “success” means for the specific human you designed for two minutes earlier. Always loop your metrics discussion back to the segment and problem you chose. It’s the throughline that makes an answer feel like one coherent argument instead of four disconnected interview-prep talking points bolted together.
For a deeper, more systematic bank of metrics frameworks used across FAANG and high-growth company interviews, see The 100x Product Manager Interview Playbook (Amazon: https://www.amazon.com/dp/B0DBC1FQWH?tag=sirjohnnymai-20).