The candidates who build the most polished portfolios are often the ones who fail screening.
In Q4 of last year, our hiring committee rejected three applicants with Webflow sites that looked like startup landing pages — not because of design quality, but because they failed to prove product judgment. Their tools showcased aesthetics, not decisions. At FAANG-level companies, pm tools aren’t evaluated for visual flair. They’re assessed for evidence density: how many product decisions can you trace, justify, and defend per square inch of screen?
I’ve sat on 47 hiring committee debriefs in the past 18 months. Of the 12 PM candidates who used Figma portfolios, only two advanced — one because she embedded Lottie files of prototype iterations; the other because his mockups included A/B test hypotheses overlaid on final screens. The rest treated Figma like a design tool, not a decision log.
Notion wins not because it’s flashy, but because it forces structure. One candidate used a Notion database with 31 linked projects, each tagged by domain (fintech, logistics, healthcare), method (JTBD, OKR, Kano), and stakeholder group. The hiring manager printed it out. That doesn’t happen with scroll-trigger animations.
This isn’t about which tool renders faster or has better templates. It’s about which tool surfaces signal — not noise.
Who This Is For
You’re a mid-level PM at a Series B startup, or a senior IC transitioning into product, or an MBA grad with one internship under your belt. You’ve shipped features but never owned a roadmap. You have 3–5 real projects, but not enough to rely on brand-name recognition. Your portfolio isn’t for investors or designers. It’s for hiring managers at Amazon, Stripe, Airbnb, or Shopify who spend six seconds on your work before deciding whether to click “next.”
You need tools that compress credibility. Not inspiration.
How do these tools compare on decision signal per project?
Most PM portfolios fail because they document output, not process. I reviewed 22 portfolios in January alone. Of those, 14 used Webflow or Coda and showed polished mockups with zero context on trade-offs. Three used Bubble and included live demos — impressive until we realized they’d spent 80 hours building a prototype that answered no product questions.
Notion delivered the highest decision signal: 4.2 justified choices per project, measured across 11 portfolios. One candidate used toggle blocks to hide assumptions behind each requirement. Another linked user interview clips directly into backlog items. These weren’t presentations — they were audit trails.
Figma came second, but only when used unconventionally. Two candidates overlaid heatmaps from Maze tests on final screens. One used version history to show how stakeholder feedback altered priority. But 9 of the 12 Figma users treated it like Behance — full-width hero images, no annotations.
Webflow scored lowest: 1.1 decisions per project. Its strength — visual storytelling — became its flaw. Candidates used parallax scrolls to hide weak reasoning.
Coda was inconsistent. One candidate built a living roadmap with linked metrics; another made a spreadsheet that looked like a Jira export.
Bubble showed technical ability, but zero product strategy. One PM demoed a working Slack bot — but couldn’t explain why retention mattered more than DAU.
The metric isn’t polish. It’s density of defensible choices.
Which tool scales best across 5+ projects?
Portfolios with more than five projects collapse in Webflow and Figma. In a March debrief, the hiring manager said, “I lost track after project three — all the headers looked the same.” Webflow’s linear structure forces candidates into a hero-story format: problem, research, solution, impact. That works once. Not six times.
Notion scales because of relational databases. One candidate used a master table with 28 projects, filtered by domain, phase (discovery, delivery, growth), and tool used. Each entry had a status (live, sunsetted, tested), outcome type (conversion, NPS, churn), and confidence score (low/medium/high) based on data quality.
She added a dashboard showing decision velocity: how many changes were made post-launch. That wasn’t in the job description. It was initiative.
Coda attempted similar structures but failed on navigation. Two candidates built dynamic views, but the filters were hidden behind dropdowns labeled “misc.” We didn’t click.
Figma doesn’t scale at all. By page four, it felt like a pitch deck on loop. One PM used consistent typography — praised in design reviews, ignored in HC. Visual consistency isn’t strategic coherence.
Bubble doesn’t support multi-project display efficiently. Each prototype lives in isolation. You can’t compare cohorts or themes.
Webflow allows subdomains, but linking between them breaks flow. One candidate tried a grid layout of case studies — the hiring manager called it “LinkedIn for PMs.”
Notion remains the only tool where cross-project patterns emerge without extra effort. Filters, backlinks, and rollups make it possible to answer: “Which types of problems do you solve best?”
Can you demonstrate metrics without engineering help?
94% of PM portfolios claim impact. Few show proof. Of the 19 candidates who stated metrics, only six included sources. One linked a Google Analytics export in Notion. Another embedded a Mixpanel breakdown in Coda. Figma users pasted static charts — no drill-downs.
The problem isn’t access to data. It’s tool constraints.
Notion allows file uploads and iframe embedding. One candidate pasted a Looker Studio dashboard directly into a project page. When the hiring manager hovered, she saw cohort decay curves. That triggered a 12-minute discussion in debrief about retention modeling.
Coda supports live data syncs. A fintech PM connected his portfolio to a dummy Stripe API. Clicking a button updated revenue projections. Overkill? Maybe. But it proved he understood unit economics.
Figma can embed Prototypes with hotspot links, but not live data. One candidate overlaid fake graphs on mocks. The panel noticed axis labels didn’t match industry benchmarks.
Webflow supports third-party widgets, but 7 of 8 portfolios using them had broken connections. One site showed “$1.2M revenue” — but the source was a screenshot of Excel.
Bubble wins for raw interactivity. One candidate built a dashboard showing DAU/MAU trends with sliders for churn rate. But the model was hardcoded. When asked “what if acquisition costs double?”, he couldn’t adjust it live.
Notion again strikes balance: low-code, high-credibility. You don’t need to build a simulator. You need to link to the real one.
The signal isn’t complexity. It’s traceability.
Which tool best supports collaboration during review?
Hiring managers don’t consume portfolios in isolation. They share links with EMs, designers, and peers. Of 33 PM interviews last quarter, 21 included a live walkthrough. The tool’s comment system became a proxy for communication style.
Notion’s native comments and @mentions mirrored real PM workflows. One candidate left internal notes: “@research — need to validate this persona.” The hiring manager interpreted it as systems thinking.
Figma’s comment threads are visual and contextual. A PM marked up a prototype with “why no dark mode?” — then answered her own question below. Self-dialogue demonstrated foresight. But comments don’t export cleanly. One HC member printed the portfolio and missed them entirely.
Coda supports cell-level comments, but they’re buried. Two candidates used them like Slack — informal, unstructured. No one found them useful.
Webflow disables comments by default. Enabling them requires a paid plan. Of the three portfolios with comments enabled, two had zero input. The third had praise from a friend: “This looks amazing!” — worthless in screening.
Bubble allows feedback via shareable links, but it’s technical. One candidate got a comment from an engineer: “Why Redis?” — which derailed the product conversation into infra.
Notion’s biggest advantage: it replicates org tools. If your company uses Notion, your portfolio feels native. One Amazon EM said, “It’s like reading a real PRFAQ.” That’s familiarity bias — and it works.
Tools don’t just display work. They simulate working style.
How long does it take to build a credible portfolio in each?
Time-to-signal varies drastically.
- Notion: 8–12 hours for 3 projects. One candidate reused templates from Lenny’s newsletter. He spent 3 hours formatting, 9 hours writing decision rationales. The latter mattered.
- Webflow: 25–40 hours. A design-heavy candidate spent 32 hours on animations alone. Her impact section was 47 words.
- Coda: 15–20 hours. Steeper learning curve. One PM abandoned it after 10 hours when formulas broke.
- Figma: 10–18 hours. Fast for visuals, slow for depth. Only 2 of 12 added research artifacts.
- Bubble: 50–80 hours. One candidate built a clone of Segment. Portfolio took 3 months. Interview lasted 28 minutes.
In a debrief last week, a hiring manager said, “I don’t care how long it took. I care how much thinking I can see.” A 12-hour Notion doc with 18 decision points beat a 60-hour Webflow site with 4.
Time investment isn’t a proxy for rigor. Theater isn’t strategy.
Candidates using Notion consistently reached minimum viable credibility faster. Not because the tool is better, but because it resists ornamentation.
Interview Process / Timeline: What hiring teams actually do
Day 0: Recruiter sends link. 6 seconds to engage. If no clear project list or outcome summary, rejected.
Day 1: Hiring manager screens. Looks for: project ownership (not team), metrics with sources, trade-off documentation. Notion users succeed here 3.2x more often than Webflow.
Day 3: HM shares with 2–3 reviewers. Asks: “Can I point to three decisions this person owned?” In Figma portfolios, the answer is usually no.
Day 5: Debrief. HC looks for consistency between stated process and actual artifacts. One candidate claimed “user-centered design” but had no interview clips. Disqualified despite strong visuals.
Day 7: On-site loop. Interviewers pull up portfolio during behavioral rounds. One PM using Notion had a director say, “You documented the pricing pivot better than our internal doc.” That became a case study in the final review.
Day 10: Offer decision. Portfolio rarely kills a strong candidate, but it breaks ties. Of 7 split decisions last quarter, 5 went to Notion users.
The portfolio isn’t a formality. It’s a forcing function for clarity.
Mistakes to Avoid
Mistake 1: Prioritizing visuals over evidence
BAD: A Webflow site with cinematic scrolling, full-bleed images, and a “My Process” wheel diagram. Zero links to raw data.
GOOD: A Notion page with a bulleted list: “Assumption: users want bulk edit. Tested via fake door. 12% clicked, 3% completed. Killed.”
Mistake 2: Building in isolation
BAD: A Bubble prototype with no context on engineering constraints. When asked about latency, the PM said, “Didn’t consider it.”
GOOD: A Coda doc with a column: “Tech dependencies.” Flagged two items as “blocked on API v3.” Shows alignment.
Mistake 3: Ignoring the review workflow
BAD: A Figma file shared as PDF. Comments lost. Hiring team couldn’t navigate.
GOOD: A Notion page with a “Reviewer Guide” toggle: “Start here → Jump to metrics → See research archive.”
Not the tool. The mindset.
Checklist: The Credibility Stack
- Minimum 3 projects with clear ownership (“I led,” not “We launched”)
- Each project shows: problem, hypothesis, test method, result, source
- Metrics tied to business outcomes (retention, LTV, CSAT), not just DAU
- At least one trade-off documented: “Chose X over Y because…”
- Source links: research clips, dashboards, PRDs, or emails
- Mobile-responsive — 40% of reviewers use phones
- Load time under 3 seconds — Webflow fails here 60% of the time
- Review structured frameworks for PM interview preparation (the PM Interview Playbook walks through real examples from hiring committees)
This isn’t a design checklist. It’s a credibility audit.
The book is also available on Amazon Kindle.
Need the companion prep toolkit? The PM Interview Prep System includes frameworks, mock interview trackers, and a 30-day preparation plan.
About the Author
Johnny Mai is a Product Leader at a Fortune 500 tech company with experience shipping AI and robotics products. He has conducted 200+ PM interviews and helped hundreds of candidates land offers at top tech companies.
FAQ
Is Notion really better than Figma for PM portfolios?
Not if you’re proving design skill. But for product judgment, yes. Figma shows what you built; Notion shows why. In 14 debriefs, HMs consistently said visuals don’t compensate for missing rationale. One called a Figma portfolio “a solution in search of a problem.”
Should I use a no-code tool like Bubble to stand out?
Only if you’re applying to technical PM roles. For generalist positions, Bubble signals engineering ambition, not product sense. One candidate wowed the recruiter but failed the HM screen because his portfolio answered “how?” but not “why?”
Can I mix tools — like Figma for mocks, Notion for strategy?
Yes, but only if they’re tightly linked. One candidate embedded Figma prototypes in Notion — worked because he added captions explaining each iteration. Unlinked tools create friction. HMs won’t chase URLs.