Interview Prep Hub

I: You're the PM for our stock trading product. Gross profit has dropped significantly over the last month. What would you do?

A: Let me play that back: gross profit — revenue minus direct costs — fell significantly within one month. Before I structure: is that absolute gross profit rather than margin? And is the goal to diagnose and restore it, or is there a target attached?

I: Absolute gross profit. Diagnose it and tell me what you'd do.

A: Two quick context questions. First, how does the product monetize — per-trade commission, spread, subscription, or a mix?

I: Assume a commission on each trade, a percentage of trade value.

A: And the main direct costs — I'd expect broker/execution fees, market data, payment processing. Is that the right picture?

I: Broadly yes. The biggest line is what we pay our executing broker.

A: Great. May I take a minute to structure? (structures) Since GP = revenue − direct costs, I'll first establish which side moved. Revenue = number of trades × average trade value × commission rate. Costs = variable per trade — mainly broker fees — plus direct fixed. A one-month sharp move with no mention of product changes makes me suspect a cost- or mix-side shock more than demand; I'll still check revenue first, in one question: what did revenue do over the month?

I: Revenue grew about 5%, same as the month before.

A: So revenue is healthy and growing — for GP to fall, costs must have grown materially faster. I'm deprioritizing the revenue branch. On costs: how did fixed and variable behave?

I: Fixed costs are flat. Variable costs grew 0%, then 2%, then 10% in the last three months.

A: Variable +10% against volume +5% — that means cost per trade is rising, so it's not pure scale; it's a rate or a mix problem inside variable costs. You said broker fees are the biggest line — how is the broker fee structured? Flat per trade, percentage, per share?

I: There's a schedule: trades under 100 shares cost us 0.05% of the dollar volume; trades of 100 shares or more cost 5 cents per share.

A: Interesting — that's a step function. Let me test what it does to unit cost. Take a $1,500 trade: under 100 shares it costs 0.05% ≈ $0.75. The same $1,500 as 150 shares at $10 costs 150 × 5¢ = $7.50. Ten times the unit cost for the same trade value. So if the trade mix shifted toward 100+-share trades, variable cost would explode while revenue — a % of value — wouldn't move. Do we have volume by trade size?

I: Small trades grew about 10% over the period. Large trades grew about 900%.

A: That's the root cause then, and it's mechanical: the mix shifted massively into the per-share tier, where unit economics are ~10× worse, and our commission — priced as % of value — doesn't pass that cost through. Gross profit compresses even as revenue grows. Quick sanity check on the "why now": did anything drive large-trade adoption — a feature, a campaign, market conditions?

I: Assume a viral trend — a few high-priced stocks got very popular with active traders. Why don't you move to what you'd do about it.

A: To recap the diagnosis in one line: a cost-side mix shock — the 100-share fee step multiplied by a 9× surge in large trades. I'll propose three fixes across pricing, the broker contract, and product, each with sizing, trade-offs and a KPI. First, short term — reprice the affected segment: introduce a per-share or tiered commission above the step, or include only N free large trades in Standard and shift heavy traders to a paid plan. If large trades are now, say, a third of volume at 10× unit cost, passing even half of that through restores most of the GP gap; I'd A/B it on new users first to measure churn elasticity. Trade-off: UX friction for our most active traders — mitigate with plan bundling rather than raw fees. Second, mid term — attack the cost itself: renegotiate the broker schedule with our new volume as leverage, or dual-route large orders to a venue priced on value; also smart order handling — where compliant, splitting a 150-share order into two sub-100 executions moves it back to the 0.05% tier. That needs a compliance/best-execution check — I'd flag it to legal before shipping. Third, monitoring: GP per trade by size tier as a weekly guardrail metric with an alert, so a mix shift never again reaches the P&L unseen. Owner: my squad; repricing ships within a sprint behind a flag; broker renegotiation is a quarter-long track.

I: Good. Can you wrap up?

A: Recommendation: fix the large-trade unit economics — reprice above the fee step now, renegotiate or re-route execution next, and add tiered GP monitoring. Reasons: revenue is growing and healthy; the entire GP decline is a variable-cost explosion from the 100-share fee step interacting with a 900% surge in large trades — unit cost is ~10× above the step while pricing doesn't reflect it. Risks: churn among high-value active traders if repricing is blunt — mitigate via plans and A/B elasticity testing — and broker-relationship terms limiting order splitting. Next steps: quantify the exact mix and GP bridge, run the pricing experiment, open the broker conversation, ship the monitoring dashboard. That closes the case for me — happy to go deeper on any branch.