Lauryn Isford on Product Growth, Onboarding and Hiring

Lauryn Isford advises growth teams to look for early product actions correlated with long-term retention, ideally examining causation as well. But she cautions against spending too long on a perfect activation metric: a young company’s users and features will change, and its metric should change with them.

Asked by Harry Stebbings to define growth, Isford calls it the practice of kick-starting, fueling and scaling business outcomes. After product-market fit, that can mean bringing in customers, helping them find value and building mechanisms that increase users or revenue.

Isford entered the field through a graduate rotation program at Dropbox, where her first team worked on the self-serve growth engine. She says the discipline had not yet acquired the formal name it has in many consumer and SaaS businesses today.

At Dropbox she learned that growth can be a game of inches. The trial-conversion team studied where customers in different markets, using different payment methods or with different profiles, failed to convert. Small improvements compounded, teaching her the value of precision and rigorous data analysis.

At Blue Bottle Coffee, Isford worked on a subscription delivering whole beans to homes. The team initially assumed online buyers would resemble café customers. Visits revealed a different routine: many home brewers were hobbyists who enjoyed the ritual and did not visit the café that day. That insight changed the team’s marketing and product priorities.

At Facebook, she worked on international growth initiatives including internet.org and an app adapted for the JioPhone launch in India. Her lesson was to question Western assumptions: understanding local needs can produce better products for a global customer base.

At Airtable, she learned to treat product-led and sales-assisted growth as parts of one engine. A user might start with a free plan or small team and later need a more complex deployment. Sales can help an existing user expand value rather than always introducing the product from scratch.

Stebbings asks whether early companies should pick one motion instead of attempting both. Isford agrees that a company still finding product-market fit and beginning to monetize should generally choose a primary motion, after learning through signups, content, community, search or sales experiments. A later sales team will look different if it serves users who already know the product.

Onboarding is about helping people orient themselves and find value. Isford describes activation as a combination of continued engagement or retention and meaningful early behavior: for example, creating a useful Notion document or building something in Airtable. Both can indicate the beginning of a repeatable habit.

Unlike making a button larger, improving activation is hard because it requires teaching. Isford dislikes a barrage of eight or twenty tooltips: users will not absorb that much at once. The task is to identify what someone needs to know now and make the path to relevant value digestible.

For horizontal products with many use cases, Isford recommends asking users what they are there to do instead of inferring intent from an email address. A person using Gmail might still be evaluating software for work. Student, personal and team use can call for different examples and templates.

The exchange should be meaningful: if a product asks for information, it should use the answer to customize the experience. A student does not necessarily need tasks and sprints first, and a professional may not need student examples.

Time to value is critical, she says, because most people who sign up for a consumer or SaaS product through a product-led motion will not return. The first experience cannot teach every feature; it should show a relevant, useful and inspiring slice of what makes the product different.

Stebbings asks whether enterprise users sometimes prefer complexity or a loading screen that demonstrates effort. Isford distinguishes visible power from burden: a product can perform sophisticated setup under the hood while keeping the user experience simple. Enterprise deployments may also involve human help, leaving onboarding to make that time more effective.

Isford says an activation metric should correlate with long-term retention and reflect both return behavior and meaningful use. Her own controversial rule of thumb is that roughly 20% of customers should be able to reach it. She dislikes narrow definitions requiring five or seven actions that only a tiny fraction complete.

A narrow definition may reflect what the company thinks is valuable rather than what users actually value. She recommends studying correlations and, where possible, causation, but revising the metric as the customer base and product evolve instead of seeking an immutable perfect number.

At Airtable, Isford says she examined the data and found that a person who created a substantive base and invited a collaborator was much more likely to use the product six months later. That combination signals both meaningful work and collaboration in a product designed for teams; the example is a correlation, not proof of causation.

Pressed for retention benchmarks, she offers loose rules of thumb rather than universal targets: most signups should return within seven days; roughly 15–20% of teams might still collaborate after a month; and around 10–20% might find value over the longer term. She says the appropriate metric depends on the product.

Common onboarding mistakes include walls of tooltips and passive checklists. Isford estimates that an opt-in educational checklist may engage only around 10% of users, leaving most untouched. She argues for more assertive education, with carefully selected material relevant to most users.

Her preferred patterns are visual, progressive and experiential: ask users to do something, show the resulting product experience, and disclose capabilities gradually. She cites Airtable and Linear as examples she likes, while noting her bias toward Airtable; Stebbings compares gradual learning to the early levels of a game.

On data versus intuition, Isford notes that she has often worked at scaled businesses with abundant data, whereas startups she advises or backs may be pre-launch. Those teams can still gather evidence by asking peers what worked, comparing similar products and learning benchmarks before investing in an idea.

The timing of a growth hire depends on the go-to-market model. For a product-led company, an engineer or technical product manager may work on growth early, setting up logging, experiments and onboarding. More generally, she recommends adding growth capacity after product-market fit and increasing investment when the initial work earns a return.

Infrastructure depends on the job. Google Analytics may be enough to begin with search-oriented acquisition. For pricing, onboarding or other changes inside the product, she wants data engineering and logging, an analytical tool such as Amplitude, Mode or Hex, and ideally an experimentation tool to learn from changes.

The first hiring question is whether the goal is acquisition or more value and revenue after signup. The former points to a data-driven growth marketer; the latter calls for someone able to work with engineering or write code. A person equally strong in both is hard to find, particularly when making only one hire.

For a product-growth candidate, Isford suggests a take-home exercise: give them the website and product access, then ask for ideas to improve growth. An experienced practitioner should offer several new ideas, draw useful comparisons and identify potential gaps in onboarding or pricing even without internal data.

A red flag for her is excessive reliance on experiments to validate every small change. Early teams need judgment about where to invest and may need substantial changes to the product, not just more tooltips, friction tests or pricing tweaks. A strong growth leader can also propose new features with a business thesis.

She favors placing growth within product, especially at early and middle stages. A shared engineering organization should apply the same quality bar to growth work as to core product work, and co-location helps people learn from one another instead of leaving growth on an island.

One hiring mistake is assuming consumer growth experience transfers directly to SaaS, or vice versa. Isford experienced a steep learning curve moving from Meta to Airtable and recommends hiring someone familiar with the relevant context when possible.

In consumer products, changing the prominence of a button or removing a signup step may meaningfully affect optional behavior. In workplace SaaS, a user asked by a manager to deliver a document may use the tool despite small friction. Consumer-quality design still matters, she says, but SaaS growth often requires larger changes aligned with user intent.

Isford says Notion had just relaunched its template gallery with more than 4,500 new templates, compared with about 500 previously, and featured outside creators. The gallery included more than 100 habit trackers. She presents this as a bet on the many ways students, professionals and hobbyists use a horizontal product.

Stebbings asks how users can discover the right template among so many choices. Isford’s team hypothesizes that people enjoy browsing and selecting something personal, drawing inspiration from Canva. She expects diverse preferences rather than a 95/5 concentration, while acknowledging that the outcome remains to be seen.

Notion calls post-mortems retrospectives, and they can happen at the level of a pair, a team or the company. For a major launch, Isford allows time for results, then asks participants what went well, what could improve, what went badly and whether the original hypothesis delivered. Surprises in customer behavior matter too.

She distinguishes missing a goal because a hypothesis was wrong from missing it through poor execution. The first can produce valuable learning if the team understands why and changes course; the second calls for a close look at how to execute better next time.

As a rough rule, she says around 20% of growth experiments produce a metric win. That differs from the 70% goal attainment rate her teams aim for: if ten teams set suitably ambitious but practical goals, about seven would hit theirs. Teams therefore need enough ideas for the minority of successful experiments to add up.

Stebbings asks whether growth should come from small increments or rare, transformative bets. Isford says it depends on the company. A business with too few signups to meet its ambitions, or an already well-optimized funnel, may need bold new bets; a business with obvious conversion, onboarding or payment gaps can gain substantially from compounding improvements.

Even a small reduction in churn can compound over years. The choice starts with the customer and the product’s present performance, not an abstract preference for big bets or small optimizations.

How long should an experiment run? Isford says sample size and the need for statistical significance matter; a narrow audience might take months or longer to reach significance. She focuses on desired behavior and guardrails, and says an early signal may sometimes justify shipping without significance. Such a signal remains less certain than a conclusive test.

Isford says angel investing and advising help her study growth problems beyond her own product. Seeing different tactics and loops gives her perspective she believes makes her a better operator. Her investment and advisory roles also give her a commercial interest in the companies she works with.

One lesson is to match acquisition to the ideal customer profile. A YouTube or community strategy might attract enthusiastic individual users without reaching a large enterprise buyer. The audience reached at the first touch must be one the product and go-to-market motion can serve through the rest of the funnel.

She sees an emerging product-led opportunity in developer tools and believes it can be a substantial market. Isford invests as an angel, usually meeting companies through her specialized product-led growth work and referral network; she also says that some pessimism about investing in 2023 may be unwarranted, a prediction she leaves open to time.

She encourages operators to invest if they understand why they are doing it and can articulate a niche where they add value. That makes it easier for a company to judge whether the operator belongs on its cap table.

In the quick-fire round, Isford says acquisition, onboarding and conversion or monetization remain core growth investments. She considers the familiar small credit for referring a friend less current, though referral programs still exist.

For someone starting a growth leadership role, she recommends spending three months learning the business, available data and customers, attending sales conversations and asking what colleagues have tried. Each customer base differs, so lessons from a previous job can create bias.

Sales-assisted and self-serve motions can compete: a dollar from self-serve may trade off against two or five dollars from a sales-assisted contract. Isford says leaders should make those trade-offs explicit and optimize for the whole business. Asked about commissions for product growth, she says predictable results can be harder to deliver through product changes and is curious about variable compensation models.

She admires Pinterest’s growth organization, although she says she does not know the specifics of its internal work. On productivity tools, she sees collaboration becoming the default and expects products increasingly to support both private space and working with others.

Isford praises Figma’s FigJam launch as a smart way to reach a different part of the market with a product complementary to its core offering. Stebbings agrees, and the conversation ends.