[ ]productinterview
Run a free mock

product sense · standard

Design a product for commuters

Design a product for commuters.

Updated Jun 2026 Calibrated to the strong-hire bar

Eighty percent of candidates answer this by describing Transit App or Google Maps with one extra feature. The answer fails not because the features are wrong but because the candidate never names which commuter they are designing for, never identifies a job that existing products leave unfilled, and never builds something that would survive the first week of daily use. Those three gaps are exactly what the interviewer is scoring.

Scope it before you solve it

The question is deliberately open. A strong candidate asks one clarifying question before choosing a direction: “Are we building from scratch or extending an existing platform, and what is the company’s moat: real-time location data, employer relationships, or transit agency partnerships?” That question signals you understand that build decisions are constrained by assets, not just user needs.

Then name the segment decision, because there are three genuinely different commuter archetypes requiring completely different products:

  • The anxious scheduler. Needs certainty above everything. The commute is a chain of commitments and any single point of failure cascades. Typically a $90K+ knowledge worker, in the office three days per week, who books rides rather than waiting. Highest willingness to pay.
  • The productive drifter. Treats transit time as the only quiet window in their day: reading, calls, focused work. Their pain is interruption and unreliable seat availability. Fixed transit pass holder.
  • The cost-maximizer. Trading time for money at a calculable rate, optimizing spend versus travel time on every trip. Likely to churn if price changes.

Pick one and defend it. The candidate who tries to serve all three builds nothing memorable.

For a Lyft-adjacent company, the anxious scheduler is the right segment: highest revenue potential, employer benefits eligibility, and the strongest product fit for a ride-booking platform.

The job to be done

The anxious scheduler’s job is not “know where my train is.” It is: leave home at exactly the right moment and arrive without a single point of uncertainty. The pain is not information deficit. It is the cognitive weight of managing a plan that might collapse at 7:43am on a train they cannot miss.

This is why proposals that surface more data (delay alerts, crowding maps, crowding by car) miss the point. The user already has data. They have four apps. What they do not have is someone who has already handled it.

US commuters average 27.6 minutes one way nationally. In major metros (NYC, SF, Chicago, DC) the median one-way trip runs 35 to 47 minutes. That is the time budget the product must fit within, and it includes the pre-departure window where anxiety peaks.

What to build: Commute Guarantee

Not a feature. A commitment.

The user sets their must-arrive time and home location once. Every morning, before they check their phone, the system has pre-computed their optimal trip, pre-booked the rideshare leg if needed, and sent one glanceable notification: “Leave at 7:48. Your ride is reserved.” No app to open. No decision to make.

This is the habit loop: the notification is the trigger, the pre-planned trip removes the variable action, on-time arrival is the reward. Within two weeks it replaces the cognitive load the user was carrying as active planning. That relief is the retention moat, not the feature set.

Two design constraints that almost every candidate skips. First, 60 to 70% of transit commuters have hands or eyes occupied during the trip. This product must be audio-confirmable (“Your 7:48 ride is 4 minutes out”) and glanceable on a watch face, not a screen you navigate. Second, a pre-booked guarantee requires supply reliability. If it fails once, trust collapses permanently. This is the hard problem: not building the notification, but making the commitment credible enough that the user stops double-checking it.

The seam problem

Candidates who design for car commuters miss the richest design space. The highest-anxiety moments in a multimodal commute are the seams: the gap between subway exit and office entrance, or parking lot to train platform. That last-mile uncertainty is where maximum anxiety concentrates and where product can deliver maximum relief. No existing product owns that moment cleanly, which makes it the most defensible surface to build on.

The 2026 reframe

In 2026, real-time GTFS feeds, Mapbox routing, and LLM-powered natural language trip planning are commodity. Proposing “an app that shows delays and lets you pick the least crowded car” describes what Google Maps already does. That answer signals a candidate who has used apps but has not thought about why people stop using them.

The differentiating question in 2026 is not feasibility. It is viability and lovability. Feasibility is free.

Viable: the IRS pre-tax commuter benefit is $315 per month per employee. Selling a Commute Guarantee subscription through HR rather than the app store means lower acquisition cost, higher LTV, and a natural renewal cycle tied to employment. Lyft Business and corporate commuter benefit programs already exist as the GTM infrastructure. A strong answer builds on top of that rather than reinventing B2C acquisition.

Lovable, in 2026 terms, means the product has handled everything before the user woke up. That is the shift from usable (tells you what is happening) to lovable (has already taken care of it). Hybrid work compressed peak commute demand into Tuesday through Thursday, creating a new “three-day commuter” segment that older prep resources miss entirely. Their commute is irregular enough that they have no muscle memory for it. They leave at 8:30 on a Tuesday twice a month and have no reliable sense of what that route looks like on that day. The Commute Guarantee is equally valuable here: the system knows their calendar; they do not have to remember to plan.

strong

"Before I go anywhere, one clarifying question: are we building greenfield or extending an existing platform, and what is the company's moat: real-time location data, employer relationships, or transit agency partnerships? That shapes the build. [Waits.] Got it. Let me narrow to a segment. There are three commuter archetypes: the anxious scheduler who needs certainty above everything, the productive drifter who uses transit time, and the cost-maximizer trading time for money. For a Lyft-adjacent company I would focus on the anxious scheduler: the $90K+ knowledge worker commuting three days per week who books rides and qualifies for pre-tax employer commuter benefits. Their core job is not 'know where my train is.' It is: leave home at exactly the right moment and arrive without a single point of uncertainty. The pain is cognitive load, not information deficit. They have four apps. What they don't have is someone who has already handled it.

The product I would build is a Commute Guarantee: not a feature, a commitment. The user sets their must-arrive time and home location once. Every morning, before they check their phone, the system has pre-computed the optimal trip, pre-booked the rideshare leg if needed, and sent one glanceable push: 'Leave at 7:48. Your ride is reserved.' No app to open, no decision. The habit loop: notification is the trigger, the pre-planned trip removes the variable action, on-time arrival is the reward. Within two weeks it replaces the active planning they were doing every morning. That relief is the retention moat.

Two constraints I would name: 60 to 70% of transit commuters have hands or eyes occupied. This has to be audio-confirmable and watchOS-glanceable, not a screen to navigate. And pre-booking requires supply credibility. If the guarantee fails once, trust is gone permanently. I would test in two or three dense corridors, NYC, SF, Chicago, before scaling, and partner with one transit agency for predictive GTFS data so we are anticipating, not reacting.

On viability: the IRS pre-tax commuter benefit is $315 per month. Selling through HR is lower CAC, higher LTV, and renews with employment. That is the real GTM, not app-store acquisition. Lyft Business already exists as the infrastructure. I would extend it rather than build parallel.

Metrics: north star is weekly recurring commute trips per active user, meaning did they use the Guarantee every commute day this week or did they fall back to improvising. Secondary: morning notification to booking conversion rate, which tells me whether they are trusting the pre-plan or second-guessing it by opening other apps. And commute-hour NPS versus off-peak NPS, because the product should be measurably better at the moment that matters. Employer-sponsored activation rate is the viability signal. If that number is not growing, the GTM is not working."

weak

"I'd design a mobile app for daily commuters that shows real-time train delays, lets you pick the least crowded car, and sends push notifications when your train is running late. I'd add a feature to save your favorite route. For metrics I'd track DAU and app store ratings." This fails on four counts. First, it describes Transit App, Moovit, and Google Maps without differentiating. Second, it names no user segment, so the features cannot be grounded in a real unmet job. Third, the features are reactive: they tell the user what is happening but the user still has to decide, act, and worry. There is no habit loop and no retention mechanism beyond the hope that users keep checking. Fourth, DAU and app store ratings are vanity signals that say nothing about whether the product improved the commute. Interviewers hear this answer from the majority of candidates. It signals someone who has used apps but has not thought about why people stop using them.

Metrics that close a strong answer

Do not close with DAU. Close with:

  • Weekly recurring commute trips per active user: did they use the Guarantee every commute day this week, or did they fall back to improvising?
  • Morning notification to booking conversion rate: are they trusting the pre-plan, or opening four other apps to verify it?
  • Commute-hour NPS vs. off-peak NPS: is the product measurably better at the moment that matters, or just generally fine?
  • Employer-sponsored activation rate: the viability signal that tells you whether the B2B GTM is actually working.

The jobs-to-be-done framework is the right lens for reframing “know where my train is” as “leave home without uncertainty.” For the 2026 context on why feasibility is no longer the hard PM question, see feasibility is free. For what lovable means when usability is the floor, see lovable, not just usable.

Asked at

Related