Marcus Webb

Product Manager · Brooklyn, NY

Marcus Webb

The highest-leverage thing a PM can do is figure out the right problem.

Five years building B2B products across data infrastructure, cloud analytics, and workflow automation. I combine structured discovery with quantitative rigor to turn ambiguous problems into measurable outcomes. The work below shows how.

How I Think About Product

Atlas Data · 2024

The growth plateau wasn't a feature gap.
It was an onboarding problem.

Context

Atlas Data's self-serve query builder had plateaued at 600 MAU after launch. Leadership wanted to invest in advanced charting. I had a different hypothesis. The data showed a plateau, but it couldn't tell us why -- that required watching real users try to use the product.

The discovery

I recruited 32 enterprise users across 4 segments and ran a 6-week structured discovery sprint -- contextual inquiry (watching people use the product) combined with problem-space interviews. The finding was stark: 78% of churned users never got past the query configuration screen. They didn't know which data source to select. The product had the features. Users couldn't find the starting line.

The decision

I presented the findings to the VP of Product with a clear recommendation: pause advanced charting, invest in guided onboarding and query templates. The supporting data was decisive -- users who completed their first query in under 5 minutes had 3.2x higher 30-day retention. The tradeoff was clear: short-term feature work versus fixing the foundation. I backed it with retention cohorts and the VP approved the pivot.

Outcome

We shipped a guided setup wizard and 12 pre-built query templates. MAU grew from 600 to 1,400 in 8 months. The advanced charting work was rescheduled to H2 -- with much better context on what charts users actually needed, informed by the usage data from the new cohorts.

32 users interviewed
600 → 1,400 MAU in 8 months
3.2x retention differential
Atlas Data · 2024

Restructuring pricing to align price with value --
without losing a single customer to churn.

Context

Atlas Data's platform tier used flat per-seat pricing. Heavy users paid the same as light users, creating margin pressure on one end and sticker shock on the other. The pricing model had worked at 50 customers. At 200+, it was starting to break.

The analysis

I structured the problem by asking: what usage metric most strongly correlates with expansion and retention? After analyzing 6 months of data, the answer was query volume. I benchmarked 8 competitors' pricing pages and designed a hybrid: base platform fee plus usage-based query tiers. Then I built a migration calculator. The critical number: 80% of existing customers would see a lower or equal bill. That's the number that made the business case survivable.

The rollout

Partnered with sales to create a grandfathering plan for the 20% who'd see increases. Ran a 4-week beta with 15 accounts before general rollout -- enough to catch edge cases without losing momentum. The beta surfaced two billing-display bugs and one customer communication gap we fixed before going wide.

Outcome

Net revenue per account increased 23% within two quarters. Transition churn was 1.2%, well below the 3% threshold we'd set as the guardrail. The usage-based component also unlocked a new growth loop: accounts that grew usage naturally upgraded without sales intervention. That wasn't the goal, but it became one of the most valuable second-order effects.

+23% revenue per account
1.2% transition churn
15 beta accounts
Nimbus Analytics · 2022

8 experiments to cut time-to-value from 22 minutes to 9.

Context

Nimbus had a 45% drop-off rate between signup and first dashboard. The onboarding was a 7-step wizard that frontloaded configuration before showing any value. New users were leaving before they experienced the product. Classic time-to-value problem.

The approach

I mapped the existing flow and identified 3 steps that could be deferred: team setup, notification preferences, billing. Then designed a "value-first" variant that connected a sample data source automatically and dropped users into a pre-built dashboard. But I didn't stop at one test. I ran 8 sequential A/B experiments over 3 months -- each building on the previous winner. Deferral order, sample data scope, copy variations, progressive disclosure patterns. Sequential testing because the design space was too large for a single multivariate test.

Outcome

Winning variant reduced time-to-first-dashboard from 22 minutes to 9 minutes. 7-day activation rate went from 34% to 51%. The sample data sandbox became a permanent feature -- sales started using it in demos, which was an outcome nobody planned for but everyone benefited from.

22min → 9min time-to-first-dashboard
8 experiments run
34% → 51% activation rate

Career

Atlas Data

2023 — Present

Senior Product Manager · Series C · 200 people

Own the self-serve analytics product line. Grew the query builder from 0 to 1,400 MAU. Led pricing restructure (+23% revenue/account). Drove cross-functional performance initiative (4.2s to 1.1s p95 latency, +31% query volume). Built the org's experimentation framework, now used by 4 product teams. Mentored 2 associate PMs through a structured onboarding program I designed -- both shipped independently by month 5.

Nimbus Analytics

2021 — 2023

Product Manager · Series B · 120 people

Owned the dashboard and visualization product. Shipped 14 features across 6 cycles. Redesigned sharing workflow (+52% adoption). Ran 8 onboarding A/B tests (22min to 9min time-to-first-dashboard). Authored SOC 2 compliance requirements that enabled the first 3 enterprise contracts ($540K ARR). Overhauled API documentation, cutting support tickets 38%.

Streamline

2019 — 2021

APM → Product Manager · Seed Stage · 35 people

First product hire. Built the development process from zero -- specs, sprint cadence, release checklists, monthly product reviews. Ran 60+ discovery interviews in 6 months. Shipped the workflow template marketplace (became #2 acquisition channel, 22% of signups). Led the v2 editor redesign (-40% support tickets, +28% task completion). Promoted from APM to PM in 14 months.

University of Virginia

2015 — 2019

B.S. Systems Engineering · Minor in Economics

Dean's List (6 semesters). VP of the Product & Design Club. TA for Decision Analysis. Capstone: hospital patient flow simulation that reduced average wait time by 18%.

What Informs My Thinking

Behavioral Economics

Kahneman's work on cognitive biases has directly shaped how I design experiments and user flows. When I killed Atlas Data's charting roadmap in favor of onboarding, that was a bet -- informed by data, not guaranteed by it. Thinking in Bets by Annie Duke formalized an intuition I already had about treating product decisions as probabilistic, not deterministic.

Chess

USCF rated, ~1650 Elo. Chess teaches you that every decision has second and third-order effects. The pricing restructure at Atlas wasn't about revenue this quarter -- it was about setting up the expansion motion for the next two years.

Basketball

Weekend pickup league in Prospect Park. I play point guard -- it's more about reading the floor than scoring. Finding the open player, creating space, making the right pass. That's a decent metaphor for cross-functional product work.

Mentorship

I volunteer with Code2040, mentoring underrepresented students in tech. I grew up in Richmond, VA -- mixed race, navigating different contexts. That gives you a perspective on empathy and code-switching that translates directly to working across teams and stakeholders.

On the shelf right now

Thinking in Bets changed how I run experiments. Nudge is why I spend too much time on default states. Inspired is the PM bible but I think it undersells the quantitative side.

Thinking in Bets — Annie Duke Nudge — Thaler & Sunstein Inspired — Marty Cagan The Design of Everyday Things — Don Norman Predictably Irrational — Dan Ariely

Let's Talk

Looking for product roles in B2B SaaS, data infrastructure, developer tools, and AI/ML. Based in Brooklyn. Open to remote, hybrid in NYC, or relocating to SF or Seattle.

marcus@marcus-webb.com