增长是叠加游戏：Elena Verna 谈产品驱动增长、产品驱动销售和更聪明的免费模式

Elena Verna started in data analytics after graduating with a statistics degree from UC Berkeley, briefly considering actuarial work before deciding she did not want exams as validation. She joined Safeway but found its velocity too slow, then pursued SurveyMonkey for more than eight months to become its first data hire. At SurveyMonkey she grew from senior data analyst to manager and director, building a data organization that could run independently. That foundation led her into product marketing and then into growth, where she became the company’s first growth product manager and later ran growth marketing, growth product and analytics.

After a short stint as head of consumer product at Malwarebytes, Elena moved into advising as a try-before-you-buy way to test teams and industries. Miro invited her to serve as interim CMO, a role that began just before COVID and lasted 11 months, through a repositioning and go-to-market reset. She later took interim leadership roles at Netlify and Amplitude while keeping advising as her main path. Her portfolio has included companies such as MongoDB and Netlify, and she describes advising as pattern matching across many growth experiments.

Elena argues that traditional executive recruiting is broken because both sides sell rainbows and unicorns while hiding the problems they need solved. A leadership hire is a large bet for the individual as well as the company, so she prefers advising or interim work before a full-time conversion. This try-before-you-buy model lets each side test fit and improves retention. She advocates it even though it is not the traditional path into leadership.

Her most fulfilling experience was SurveyMonkey, where she stayed almost seven and a half years and grew with the company through its teenage and graduation phases. Her most challenging was Miro, where hypergrowth meant she had to grow as fast as the company while leading a full repositioning through COVID. She sees a correlation between the hardest challenges and the most gratifying outcomes. At the leadership level, fulfillment comes from taking a company through a challenge to the other side, not just from the title.

Elena describes herself as a growth product manager and a growth marketer, because product-led and marketing-led motions have to work together. She notes that leadership ladders are usually organized around CPO or CMO roles, leaving growth leaders in an undefined space. That mismatch is one reason growth leadership has career challenges. Growth work sits between product, marketing and analytics rather than inside a single traditional function.

She explains that B2C products had to be shiny because they marketed directly to users, while B2B built for an enterprise buyer who wanted utility and efficiency. That created cold interfaces and feature checklists for people who were not even the users. In the last decade, decision power has shifted back toward prosumers and employees. The new wave of B2B products is consumer-like, builds habits and aims at effectiveness, not just cost-efficient job completion.

Consumerization starts with redefining the customer as the user, then redesigning onboarding, time to value, activation and habit loops around that user. Elena measures affinity and advocacy through product NPS and word-of-mouth loops. Slack is a famous example: enterprise buyers did not choose it top-down; employees brought it in and buyers paid after seeing usage. Miro and Amplitude follow similar bottoms-up patterns, with users adopting first and enterprise contracts following.

Elena defines product-led growth by the growth model’s three questions: how to acquire, how to retain and how to monetize. A company can be product-led, marketing-led or sales-led in any of those motions. Bottoms-up is specifically about using usage to generate leads for sales, especially in acquisition and monetization. Product-led is a version of bottoms-up that leans on hand-raisers and product qualification to move toward enterprise contracts.

Her first rule is to nail product-led retention before attempting product-led acquisition. Retention depends on activation and habitual engagement; if users are not forming habits, there are fewer opportunities for referral, content or usage-based acquisition. She also asks whether the product has a one-to-many or collaborative relationship among users. If it does not, product-led acquisition is much harder, and marketing-led or sales-led acquisition may be the better path.

She frames growth motions as a layering game rather than a switch. A sales-led company should not abandon sales to become product-led; it should hunker down on sales and overlay product-led growth to amplify existing growth. The same logic applies in reverse: product-led companies can add sales-led motions without discarding their user engine. The biggest mistake is treating the choice as either/or instead of deciding where to start and what to layer next.

Elena describes a common PLG tragedy: a company gets user love, word-of-mouth, community and self-serve traction, then closes its first enterprise contract with a much larger ACV. The team gets addicted to larger deals and hires more salespeople, forecasting productivity per sales head. Early results look like unicorns and roses because the accounts already have usage and advocates. But the company is often selling into buyers who are already inside its user base.

Slowness arrives when the company runs out of enterprise buyers inside the existing user base. Elena says only 30% or less of the time is the buyer already a user, so sales must go find buyers who have no relationship with the users. That pushes the company upmarket into enterprise marketing, demand generation and larger deals. The original usage engine keeps running, but the company starts disconnecting from the dots that created it.

The crucial mistake is letting go of product usage growth while chasing enterprise buyers. Resources are finite, so companies choose enterprise reps over growth PMs or ABM specialists over growth marketers, and those personas often do not overlap. Elena says around 80% of companies let go of product-led initiatives in favor of enterprise, then see enterprise pipeline slow because they forgot the hand that feeds them. Some correct quickly; others pivot into a top-down sales machine and enter a dangerous reinvention.

She also warns that PLG ACVs are not comparable to top-down sales ACVs because PLG enters accounts earlier, so it is an expansion game rather than a land-the-whole-enterprise game. Overcorrection shows up when the product roadmap focuses exclusively on enterprise features. That is a telltale sign that the company has stopped innovating for the users who act as selling agents inside their organizations. PLG companies must protect the user engine even as they add enterprise sales.

Elena predicts that product-led sales will box out top-down sales organizations within another 10 years. The reason is that it focuses on user value that drives expansion, and its cost of acquisition is the cost of acquiring a user rather than a buyer. Pipeline comes from usage instead of marketing-qualified leads, which creates more hooks for acquisition, retention, product discovery and adjacent use cases. She sees this as a one-step-forward model where users are empowered to make decisions and act as sales agents.

She points to self-serve moving upmarket: Atlassian closes $200K deals self-serve, and she paid over $24K a year for Drift on a credit card without talking to anyone. Product-led sales pushes pressure onto the product and users, not just onto a traditional sales team. Enterprise buyers still matter for budgets, spend monitoring and unit economics, but Elena argues that decision-making has moved to users.

In her definition, product-led sales starts with acquisition and self-serve usage, whether freemium or paid. The company then looks for network effects inside an account and builds a product qualification model scoring usage volume, velocity, feature breadth, behavior and demographics. Hand-raisers are the first pipeline; sales also outbounds to users who behave like hand-raisers but have not raised their hand. Sales uses the product itself as enablement material rather than starting with demos and dinners.

Traditional sales-led growth starts with marketing-qualified leads or outbound prospecting, then uses demos and relationship-building to convince a buyer before usage occurs. That often creates utilization problems because the company sells as much as possible to the organization and then has to catch up to the contract value. Product-led sales instead captures intent from users and lets usage reveal where enterprise expansion is likely. Elena expects enterprise buyers to be boxed out of decision-making even though they still hold budgets and monitor spend.

Elena defines freemium as ongoing free usage with feature or usage walls, while a trial is the same free usage with a time bound. The time bound is superficial because the time it takes one person to try a product varies enormously by context. A startup user may finish a project in 24 to 48 hours, while an enterprise user with dependencies may need a month. A seven-day trial therefore serves small startups and alienates large enterprises that need more time to reach value.

MongoDB is her favorite example of freemium that is really a trial in disguise, but without a time limit. It offers a free shared community cluster where nobody would put a production application, yet anyone can prototype for as long as needed. Production use naturally moves into paid monetization. She applies the same logic to high-friction products such as Amplitude, Twilio, Contentful and Snowflake: the friction depends on the segment, and self-serve innovation keeps pushing provisioning costs down.

Freemium needs a strategic reason beyond conversion-rate optimization. At Miro, three editable boards are enough for some users who will never pay, but they share boards and acquire other users. At SurveyMonkey, half the user base was free, retained and healthy, and it drove 100% of acquisition. LinkedIn uses a strong freemium base and network effects to build sales navigator and recruiting products on top.

Her framework asks whether free drives indirect monetization or virality, whether it covers a commoditized feature every user needs, whether it delivers the aha moment, and whether it creates habit loops. Anything that promotes the growth model should probably be free; anything that creates friction for growth should probably be gated. The exact answer depends on the monetization strategy, but the test is how free helps the growth model without sacrificing monetization potential. This is why she says the biggest mistake is treating freemium as only conversion-rate optimization.

Elena believes growth is an evolution, not a revolution, so the first growth hire is often best found internally. A data analyst, existing product manager or technically inclined engineer who understands the business can test the first hypotheses. External hires are not forbidden, but they risk slapping a previous company’s growth model onto a product with different mechanics. She says copied models can have a 99% rejection or failure rate with customers and employees.

External growth leaders need three to six months, sometimes longer, to learn what the growth model should be. If they are pressured for immediate step-function change, they will copy-paste the only proven pattern they know. Elena notes that even at SurveyMonkey, a Qualtrics growth model would not have worked because the products and markets were intertwined differently. She says she still has to pull back at Amplitude after four months when a familiar tactic does not produce the expected step change.

She says a startup should bring in a dedicated growth person only after strong product-market fit, with retention evidence from at least six months to almost a year and some green shoots in acquisition or monetization. Hiring someone to invent the first growth model is like hiring a product person and saying “I don’t know what to build, go build it.” The founding team must create the first growth model itself. Her final advice is to find patterns in problems, minimize context switching and build frameworks for repeatable, sustainable growth.