Strategy Before Goals: Ravi Mehta's Product Strategy Stack, Tinder Lessons and Selective Micromanagement

Ravi Mehta has been a product leader at Tripadvisor and Facebook and chief product officer at Tinder. He is now co-founder and CEO of Outpace, a coaching company. Asked what surprised him most about moving from large companies to founding one, he starts by challenging the idea that startups are faster. His own company felt slower at first, with fewer engineers and no base of existing users to research.

What startups really have, Mehta says, is low latency rather than high velocity. Big companies can always do more work, but a small team can have an idea one day and test it the next. He likens it to driving: a car at high speed has a wide turning radius, while a startup turns tightly. His test for any team is how long it takes to go from deciding a button text change is worth making to seeing results.

Decision-making changes too. Companies with large audiences can lean on experiments and statistical significance, but at a startup that approach is too slow or too noisy. Mehta shifted from an experimental mindset to a conviction-oriented one: ask whether you have enough data for informed conviction, commit, and then see whether to double down or change direction. Otherwise a small team can drown in market research and pricing variants.

The third surprise was social. People tend to build their careers around a particular company stage, so a network built at Facebook or Tinder is not necessarily useful for founding a company. Early-stage networks of founders, freelancers, angels and investors are different, and so are the growth tactics they use. Mehta recommends plugging in early, for example through the indie hacker community and the Everything Marketplaces community.

Mehta created the product strategy stack because PMs kept asking him how to choose between two features of similar size. When he dug in, the problem was usually that the strategy was not understood, or had never been fully defined. The stack splits commonly blurred terms into five distinct layers so teams can see which one is missing.

At the top is the company mission, a qualitative and aspirational statement of the change the company wants to make. Below it is company strategy, which is rigorously logical: the specific plan for achieving the mission. Product strategy is the connective tissue between what the company wants and what the product team does day to day. It informs the roadmap, and the roadmap in turn informs the goals.

The stack works in two directions. To define strategy, a PM works top-down. To diagnose problems, they work bottom-up: if goals are being missed, the roadmap may be wrong, and if the roadmap is wrong, the product strategy may be unclear. The cause may ultimately be that the team does not understand the company strategy. Mehta folds vision into mission, and Lenny adds that mission is what you are trying to do and vision is what the world looks like once you have done it, so the key is not to overthink either.

The framework began at Tripadvisor, where the team had to plan a major trip-planning feature, a problem several startups had failed to crack. Instead of working bottom-up through experiments, the team defined what they wanted to achieve, how, and in what steps. They also declared that the strategy doc would not be complete without wireframes, which was a first for them in a strategy context.

Mehta explains the reasoning with an architect analogy. No one would hire an architect who described a house only in words, because everyone would imagine something different. A blueprint does not need high fidelity, but it shows how the pieces fit. Since most products are pixels on a screen, you need to know how those pixels will be organised.

His litmus test is the mobile navigation bar, which typically holds only four or five items. Ask different people what those items would be given a strategy written in words and you will get different answers. Drawing it out forces trade-offs early, because there are always limited pixels. For PMs without a designer, Mehta suggests sketching on paper or using the wireframing tool Balsamiq, which can produce conceptual wireframes in an afternoon.

To show the stack in action, Mehta contrasts Tinder with Hinge. Both are freemium dating apps, yet their missions differ. Hinge is designed to be deleted once you find a long-term partner, while Tinder's mission is to make single life more fun and to stay on people's phones throughout their 20s and early 30s.

Their company strategies share a freemium model with some monetised features, but they acquire customers differently. Hinge leans on television ads, while Tinder relies more on influencer and event-based marketing. Their product strategies diverge more sharply: Tinder is a fast swipe-based experience with simple profiles, while Hinge deliberately moved beyond swiping to richer profiles with prompts that spark conversation.

Roadmaps overlap in places. Both invested heavily in video chat after the pandemic, expecting people to spend more time online before meeting in person. Goals are very similar at a high level, since both measure meaningful conversations, but the product mechanics beneath those goals differ. Mehta says the stack shows when a decision should copy competitors and when it should not.

A counter-example is search filters. Almost every dating app lets you filter by occupation, income, religion or height, and Tinder has largely resisted. The product philosophy was that Tinder should not become a search engine for people. Users often say they met people they would never have found had they been able to filter.

Mehta describes a discovery story from Tinder. The team noticed that a single-digit percentage of users generated a large share of a la carte revenue, spending hundreds of dollars a month on features like Boost and Super Like when average spend was around $30. The team's hypothesis was that these were wealthy users who did not care about the money.

Tinder's monetisation has two parts. Subscriptions such as Tinder Plus and Tinder Gold let users break the rules, for example by seeing who has already swiped right on them. A la carte products include Super Like, which makes a match about three times more likely, and Boost, which shows a profile more often. Researchers identified the heavy spenders and ran a usability study to find out why they paid so much.

The assumption was wrong. These users were not richer than average. They were people with an intense use case, such as military personnel who moved often, salespeople in different cities, or newcomers to a city. They compared the price not to other subscriptions but to the cost of dating, where a few dates a month could easily cost a couple of hundred dollars.

Two products followed. Tinder Platinum became a third subscription tier bundled with consumables, and a premium Super Like with a note let a user send a first message before matching. Because the team understood how users valued the utility, it priced the new offering well above its earlier expectations. Mehta's broader point is that PMs should spend more time simply calling users to understand their psychology.

Mehta admits that placing goals after the roadmap in his stack is contrarian, and some people have objected strongly. In many companies goals start the process, with a target such as raising revenue by X or retention by Y. He argues that this puts all the team's energy on moving numbers without any structure describing what success looks like or why.

He uses a road-trip analogy. You do not start a trip by saying you will drive 250 miles; you pick a destination, such as Las Vegas from Los Angeles, and the 250-mile figure only matters in that context. If hitting a goal pulls the product away from the destination, the team needs an explicit conversation about either leaving that gain on the table or changing the destination. Without that, decisions about direction slip by unnoticed.

Lenny asks how to prioritise a roadmap without goals. Mehta says strategy usually contains quantifiable elements. At Tripadvisor the strategy was to get people to plan trips directly on the site rather than hopping back and forth to Google. Yet teams optimising for immediate bookings kept pushing users straight down the transaction path, which undermined the long-term relationship. With the strategy clear, they could size and prioritise opportunities and then measure progress.

Mehta has seen many companies struggle with OKRs. One cause, he says, is the rule of always focusing on outcomes over outputs. A PM should ultimately be measured on outcomes, but that does not mean committing this quarter to an outcome nobody knows how to move. He calls the line between what a team knows and does not know the frontier of understanding.

He sorts goals by the risk they address. Understanding risk means you do not know what the levers are, so the goal should be to learn. Dependency risk means you know the levers but lack the tools. Execution risk means you have the hypothesis and resources but may not deliver, so a goal might be to run 20 experiments. Strategic risk means the hypothesis itself may be wrong.

Lenny points out that many managers will not accept a quarter spent on understanding. Mehta suggests keeping the outcome goal but breaking the quarter into stages of understanding, hypothesis forming, execution and validation. If the goal is missed, you can explain where things went wrong and what to commit to next. Leadership is outcome-driven, he says, but wants confidence that outcomes will be hit, and a clear learning path provides that.

Mehta adds a two-by-two: did you hit your goals, and do you know why. The ideal is hitting goals and knowing why. Hitting goals without knowing why will eventually catch up with you, and when you miss, you should at least be building understanding. He has written about this as setting goals with NCTs instead of OKRs.

Mehta developed his PM competency framework at Tripadvisor, which was newly public and needed to grow its product team faster than hiring experienced Boston PMs would allow. The company created a rotational program that hired people out of undergrad and business school into rotations across zero-to-one, growth and infrastructure teams. That required a clear definition of what product management is and a plan for building the skills.

The result is 12 competencies in four areas. Product execution covers functional specification, product delivery and product quality. Customer insight covers fluency with data, voice of the customer and user experience design. Product strategy covers owning business outcomes, product vision and roadmapping, and strategic impact. Leadership covers stakeholder inclusion, team leadership and managing up.

The same structure applies from associate PM to CPO, though the specifics change. An APM thinks about execution as their own daily work, while a CPO thinks about the systems that help teams execute. Lenny calls it the simplest and clearest definition of the PM role he has seen, and Mehta offers a downloadable kit on his website for self-assessment and archetypes such as growth PM versus discovery PM.

The framework also enables what Mehta calls exponential feedback, feedback with compounding returns. Fixing a surface symptom helps once, but understanding the underlying behaviour helps a person diagnose themselves going forward. He suggests rating yourself as needs focus, on track or outperforming on each competency, asking your manager to do the same, and using disagreements as the start of a real conversation. He also advises giving managers permission to give raw, real-time feedback.

Mehta sees two failure modes for new managers. The first is real micromanagement, which signals distrust and limits the manager to only a couple of reports. The second is a fully hands-off approach, where the new person gets autonomy but not the context, guardrails or frameworks to channel their effort.

He notes that some of tech's most innovative leaders, including Steve Jobs, Elon Musk and Mark Zuckerberg, are known micromanagers, because details like the text on a button can matter. He prefers to think of seniority as increasing dynamic range: a CPO spends time on strategy but can also zoom into specifics. His preferred state is scalable leadership, where the leader is confident in the direction and the team has autonomy.

When confidence is missing, the answer is selective micromanagement: get detailed in a tactical and temporary way, while building a framework the team understands so you can pull back. What people usually call micromanagement, Mehta calls micro-mismanagement, where neither side has control or clarity. Success is when the team uses the framework to reach answers better than the leader's own.

A related idea is that employees constantly manage two variables with their boss: alignment and confidence. High on both brings full support. If you are misaligned but have your manager's confidence, you can still win permission to go your own way. If you have neither, something has to change.

Mehta says Outpace expected AI to matter eventually, but advances in models from OpenAI, Stable Diffusion and Midjourney in 2022 pulled that timeline forward. Outpace pairs content with a human coach: each week a participant does a 20 to 30 minute session with a short audio lesson and interactive exercises. Participants' written answers become the raw material for AI suggestions.

Coaches can ask OpenAI's models to suggest feedback on a participant's response and then tailor it to the person. By changing the prompt, the team can simulate different leadership styles: action-oriented lists of next steps, sympathetic responses, inquisitive follow-up questions or informative frameworks. Mehta sees the value not in replacing people but in giving them a strong starting point, in text and image generation alike.

In the lightning round, Mehta recommends the books Hooked and Working Backwards, and his favourite interview question is to ask candidates about a product they love. He lists Airtable, Webflow, Superhuman, Descript and Balsamiq as his team's tools. To evaluate a coach, he suggests asking about the client they are proudest of helping, which reveals what they value.