TheShortlist^™

Most pre-seed decks land on a go-to-market slide that lists the same things: content, paid ads, Reddit posts, LinkedIn, conference booths, podcast tour. It reads like a list of channels a founder would do if they had nothing else.

We tested whether that motion correlates with venture-scale outcomes across 65 hand-researched companies: 50 portfolio winners (Andreessen Horowitz, Sequoia, Redpoint, plus mid-tier and late additions) and 15 named failures (WeWork, Theranos, Quibi, Fast, Juicero, Magic Leap, Jawbone, Webvan, Convoy, Olive AI, Bird, Better.com, Pets.com). Each company's primary GTM motion at first institutional check was coded as product-led, sales-led, network-led, channel-led, or marketing-led.

The failure cohort, sorted by motion

Eleven of fifteen named failures were running marketing-led at the moment of first institutional check. The pattern was visible in the deck before the company knew it had a problem.

• WeWork. Brand and community marketing led every funding round. Adam Neumann was the brand, the experience was the brand, the marketing budget was a primary line item. The product (subleased commercial real estate with snacks) never carried itself.
• Quibi. Hollywood-studio marketing, billboard saturation, $1.75B raised against a content-marketing bet that famous talent plus short episodes would create a habit. No product mechanic was forcing return visits.
• Fast. Celebrity-fronted TV and Twitter ads, one-click checkout as a category. The marketing spend scaled. The product retention did not, because Shop Pay already owned the install base.
• Pets.com. The iconic sock puppet ad and Super Bowl spot remain a marketing case study. Customer acquisition cost ran well above lifetime value the entire time. The marketing worked; the unit economics did not.
• Jawbone. Industrial design and brand marketing against Fitbit and Apple. Once design parity collapsed, there was no software or network behind the product to hold customers.
• Bird. Consumer brand, urban-rider marketing, paid scooter visibility. Lime, Spin, and city regulators arrived. Brand alone could not protect against operationally identical competitors.

Five other failures (Theranos, Webvan, Magic Leap, Convoy, Olive AI) ran sales-led or channel-led motions and failed for category reasons unrelated to GTM. Marketing-led was not the only failure mode. It was the most common one.

The winning cohort, sorted by motion

Product-led growth dominated the winners. Eighteen of fifty winners ran a primary product-led motion at first institutional check, against one of fifteen failures.

• Stripe. Developer-led adoption. One engineer ships an integration. The team standardizes. The company commits. Marketing spend was minimal until the company was profitable at scale.
• Snowflake. Per-query consumption pricing meant the product sold itself in usage data. Sales overlay came later for enterprise expansion. The first thousand customers signed before a marketing budget existed.
• Figma. Designers brought Figma into companies, then designers became managers and standardized teams on it. No outbound until the company was already replacing Sketch and Adobe.
• Slack. One team adopted, then the next. Stewart Butterfield famously refused to spend on paid acquisition because the product retention told them the loop was self-sustaining.
• Notion. Community-led, template-driven, viral inside specific use cases (engineering wikis, founder note systems) before a single ad ran.
• Datadog. Free tier for individual engineers, paid tier for teams. The same person who installed it became the budget owner inside their company.

The remaining thirty-two winners ran network-led (Airbnb, Stripe in its merchant phase, Uber), sales-led with a clear ICP (Salesforce, ServiceNow, Workday), or channel-led with named distribution partners (Robinhood through Plaid, Plaid through banks). Marketing was not the lead motion for any of them at the first institutional check.

Why does marketing-led correlate with failure?

The mechanism is straightforward. Marketing-led growth scales ad spend, not retention. If the product does not pull users back on its own, the company is paying for every cohort indefinitely. When the funding environment tightens, the spend cannot stop, and the company either runs out of cash or pivots to a real motion too late.

Product-led growth, network-led growth, and sales-led growth into a specific buyer all share one property: the cost of acquisition decreases per cohort as the loop matures. A product-led company gets better referrals per user year over year. A network-led company gets liquidity that makes each new user more valuable to the existing base. A sales-led company gets reference selling, lower friction, and shorter cycles. Marketing-led does not have this property structurally.

This is not new. Andrew Chen at Andreessen Horowitz documented the law of shitty clickthroughs (paid acquisition channels decay over time) in 2012. Sequoia's PMF framework (v2, 2024) treats sustained product-led retention as the test of whether the company has earned the right to raise. Eugene Wei has written for a decade about the difference between distribution that works once and distribution that compounds. The pattern is well-documented. Pre-seed founders default to marketing-led anyway because the channels feel concrete and the dashboard moves.

What this means if you are applying to a fund

The partner reading your deck is going to look at the GTM slide for one thing: does the loop compound, or does it depend on a budget. The slide that fails by default is the one that lists five paid channels and three brand activations. The slide that passes names one motion that the founder can articulate as a compounding loop, then names the early evidence the loop is real.

1. If you are product-led, name the wedge user and the friction point that activates them. "Engineers at fintech startups, configured in under fifteen minutes, integrated into CI within a week." That is the motion.
2. If you are network-led, name the side of the network that comes first and what makes the other side want to join. Two-sided networks die at the supply-demand seam. Specify which side you are seeding and why it wins.
3. If you are sales-led, name the ideal customer profile by company size, role, and pain. "Heads of revenue at series-B fintechs with under twenty AEs." That is a motion. "Enterprises" is not.
4. If you find yourself listing five marketing channels, you are not yet positioned to raise venture capital. Marketing is a layer on a motion that already works, not a motion of its own. If marketing is your primary distribution, the partner reading your deck will assume the product retention is too weak to carry itself, and the data in this issue is why.

The exception worth naming: consumer brands at scale (Hermès, Apple, Hims) run marketing-led motions and win. They run them on top of either a powerful brand moat or a category-defining product. At pre-seed, neither moat is yet built, so marketing-led at the first institutional check is correlated with failure for a structural reason: the founder has no compounding loop to layer marketing onto.

Further reading. Andrew Chen, "The Law of Shitty Clickthroughs" (a16z, 2012). Sequoia Capital, "PMF Framework v2" (sequoiacap.com, 2024). Eugene Wei, "Status as a Service" (eugenewei.com, 2019). Hamilton Helmer, 7 Powers: The Foundations of Business Strategy (Deep Strategy, 2016). Reid Hoffman, Blitzscaling (Currency, 2018) on distribution-led growth.