strategy · hard
How would you price and GTM a new product?
How would you price a new product and build its go-to-market strategy?
Most candidates treat this as two questions stapled together: a pricing slide, then a GTM slide. Interviewers at senior levels are checking whether you understand that price is the first strategic statement about a product. The model, metric, and number you pick force the distribution motion, the sales approach, and the success metrics downstream. Get the sequencing right and the rest of the answer follows.
The sequencing that matters
Price cannot be set before the ICP is locked. If you do not know who buys, you do not know what outcome they value, and without a valued outcome you are guessing at a number. The correct order:
ICP first. Not “small businesses” but “ops-heavy SMBs with 20-200 employees currently running three SaaS tools that do not talk to each other.”
Quantify the value delivered. If the product saves 8 hours of manual reconciliation per week at a $60/hr loaded labor cost, that is roughly $25K/year in recovered time per seat.
Derive the price from the value, not the cost. At a 10x ROI ratio, a customer would pay up to $2,500/year. Pricing at $1,200 makes the buy decision straightforward and preserves room for expansion revenue. Your customers do not care about your server costs.
Choose a pricing model that matches where value is actually created. Per-seat works when value scales with headcount. Usage-based works when value scales with output volume. Charging per seat when value comes from data processed creates friction and misaligns incentives. Stripe charges a percentage of transaction value because that is exactly where value is created.
State the pricing objective explicitly. Share capture and margin maximization are both valid but they are not compatible at launch. This choice determines whether you price to grow or price to monetize, and it has to be made before you pick a channel.
Choose the GTM channel from the pricing model, not before it. Enterprise price-points ($10K+ ACV) require a sales-assisted motion. PLG and freemium require a self-serve activation funnel. You cannot bolt a self-serve channel onto an enterprise pricing model and expect it to work.
Define success metrics tied to the pricing model. Trial-to-paid conversion at 30 days. Expansion MRR (are customers adding seats or usage, which validates the value metric). Time from signup to first value event (tells you whether onboarding activates value fast enough).
Structure a strong answer
strong
"Before I get to the model, I want to lock two things: product type and business objective. Is this B2B or consumer? Are we in share-capture mode in year one, or do we need to monetize from day one?
Assuming B2B, ops-heavy SMB, share-capture mode: I'd quantify the value first. If we save 8 hours per week of manual reconciliation at $60 per hour loaded cost, that is roughly $25K per year per seat recovered. I'd price at $1,200 per year: about 5x ROI for the customer, enough margin that the CFO approves without a committee. That is well under the value delivered, intentionally. We are buying adoption data and case studies, not maximizing year-one revenue.
The value metric is per seat because value scales with headcount here. If this were an AI summarization tool where value scales with documents processed, I'd go usage-based instead.
For launch, I'd offer a founder cohort at 30% off, locked in permanently. That generates NPS data, usage telemetry, and case studies for outbound. At $1,200 ACV, the deal is too small for a full enterprise sales cycle but too big for purely self-serve. So we do light sales-assist: outbound to ops directors on LinkedIn, a demo-gated trial, and a waitlist we work through personally for the first 50 accounts.
Success metrics: trial-to-paid at 30 days (primary), expansion MRR by seat (validates the value metric), time from signup to first value event (tells us if activation is working). If expansion MRR is flat at 90 days, the per-seat model may be wrong and I'd revisit with cohort analysis, not aggregate revenue."
weak
"I'd research what competitors charge, price in that range, maybe do freemium, then launch with social media and a press release, and track DAU and revenue." This fails on every dimension: pricing becomes a lookup table rather than a strategic statement; freemium is chosen with no connection to the business objective or ICP; channels are listed rather than chosen based on the pricing model; success metrics are vanity defaults with no tie to what was actually priced. The candidate never states what problem the product solves or for whom, so every subsequent choice is unanchored. Interviewers describe this pattern as "memorized a checklist, exercised no judgment."
The 2026 reframe: pricing is now the viability argument
In 2026, feasibility is nearly free. You can build almost anything with AI. That collapses the old PM triangle: viable and lovable are now the bar, and feasibility rarely filters anything out. The pricing question in an interview is now primarily a viability test. Can you name a market that will pay, a price that covers inference costs at scale, and a distribution motion that gets you to the usage volume where unit economics close?
For AI products, this is concrete. Inference cost is a real floor, and usage-based pricing (per query, per token, per output) is now standard because flat-rate risks margin compression at scale. A strong answer names this explicitly: “At $0.002 per query, our gross margin at $20 per month requires that most users stay under 500 queries per month. Here is how we design the activation flow and usage caps to make that likely.” Candidates who answer as if it is 2019, treating pricing as a marketing decision and GTM as a launch checklist, signal they have not absorbed the shift.
One asymmetry worth stating aloud: pricing too cheap is a one-way door. It permanently anchors perceived value downward and is harder to fix than pricing too high. That asymmetry should appear in your answer, especially at staff or GPM level.
What signals seniority by level
APMs should show the sequencing (ICP before price before channel) and be able to name one WTP research method (Van Westendorp, Gabor-Granger, or conjoint) without just listing all three.
Senior PMs should demonstrate the pricing-model-to-channel dependency and state the pricing objective before picking a model.
Staff and GPM candidates should address the follow-up: “You chose freemium, what if free users never convert?” The answer is cohort analysis by activation event, not aggregate conversion rate, plus a specific trigger that moves a user to a paid conversation. They should also name regulatory and compliance pre-launch checks in fintech, healthtech, or any AI product subject to output liability. That detail is a tell for senior thinking.
See also proving viability for the viability math that underpins any pricing answer, and feasibility is free for why the 2026 PM bar has shifted to the viable-and-lovable axis.