--- title: "Product-led growth (PLG) in SaaS: how it works and when it fails" description: "How product-led growth turns free users into paying customers, the metrics that show it working, a funnel worked through, and the cases where it fails." canonical: "https://saascron.com/plg-saas/" language: "en-US" author: "Ricardo Machado" author_url: "https://saascron.com/author/ricardo-machado/" published: "2026-10-22T17:00:00-04:00" updated: "2026-10-22T17:00:00-04:00" category: "Growth & Marketing" tags: ["Plg", "Freemium", "Growth", "Unit economics"] --- # Product-led growth (PLG) in SaaS: how it works and when it fails > How product-led growth turns free users into paying customers, the metrics that show it working, a funnel worked through, and the cases where it fails. You have a product people can try on their own, a small budget and no sales team, and someone has told you to "go product-led". You need to know what that means in practice, which numbers prove it works, and when it will not work for your product. Product-led growth (PLG) is a way of selling software in which the product itself does most of the selling: people sign up, get value on their own and pay when they need more. PLG is a go-to-market choice, not a feature. It shapes your pricing, your onboarding and who you hire. This guide explains the mechanics, works through a funnel with numbers and lists the signs that a product needs salespeople instead. For the basics of the delivery model behind it, see [What is SaaS?](https://saascron.com/what-is-saas/). ## Product-led growth in plain terms Think of it like the tasting stand at a supermarket. Nobody pitches you. You try a piece of cheese, you like it, and you put a block in your basket. The stand works only if the sample is good, quick to taste and close to the shelf. PLG applies the same idea to software. Four pieces make it up: - **Free entry.** Either a free trial (time-limited access to the paid product) or freemium (a permanent free plan with limits). The two are not interchangeable: a trial ends on a date, a free plan ends when you hit a limit. - **Self-serve onboarding.** A new user reaches a useful result without a call, a demo or a setup fee. - **Self-serve purchase.** A card form and a public price list, so a user can pay in minutes. - **In-product expansion.** More seats, more usage or a higher tier, bought from inside the product as the team grows. The opposite model is sales-led growth: a salesperson qualifies the buyer, runs a demo and negotiates a contract. Many companies run both, which this guide covers below. ## How a PLG funnel works A PLG funnel has four stages. Each one has a metric, and each metric needs a definition you write down before you measure it. 1. **Sign-up.** Someone creates an account. Count accounts, not page visits. 2. **Activation.** The user completes the action that shows the product working for them. For a scheduling tool, that might be "booked a first meeting through the link". Activation rate is activated users divided by sign-ups in the same cohort. 3. **Conversion.** The account starts paying. Free-to-paid conversion is paying accounts divided by sign-ups (or by activated accounts; say which). 4. **Expansion.** The paying account adds seats, usage or a higher tier. That is expansion revenue, the added revenue from existing customers. Some teams also track product-qualified leads (PQLs). A PQL is a free or trial account whose usage matches the pattern of accounts that later paid, such as five active users in one workspace. In a hybrid model, PQLs are the accounts a salesperson calls first. ## What two public filings show Two software companies described this model in their registration statements before going public. Their numbers show the scale a free entry can reach, and how a self-serve company still adds sales later. **Slack, Form S-1 (2019).** As of January 31, 2019, Slack reported "more than 600,000 organizations with three or more users". Of those, more than 88,000 were Paid Customers and more than 500,000 used the free plan ([Slack Form S-1](https://www.sec.gov/Archives/edgar/data/1764925/000162828019004786/slacks-1.htm)). The filing described a "self-service approach, for both free and paid subscriptions". It also said Slack would expand its direct sales force to reach large organizations, with efforts "targeted at C-suite executives and business unit leaders". The paid share is a rough calculation, because both counts are lower bounds: 88,000 / (88,000 + 500,000) = 88,000 / 588,000 = about 15%. About one organization in seven paid. The other six were the free entry. **Atlassian, Form F-1 (2015).** Atlassian stated that it had no "direct salesforce" and that its sales model did "not include traditional, quota-carrying sales personnel". It reported more than 51,000 customers and total revenues of $319.5 million for fiscal 2015, which ended June 30, 2015 ([Atlassian Form F-1](https://www.sec.gov/Archives/edgar/data/1650372/000104746915008450/a2226437zf-1.htm)). Dividing one by the other gives a rough ceiling on revenue per customer: $319,500,000 / 51,000 = about $6,265 a year. That is a low-price, high-volume business, which is the shape PLG fits. ## Worked example: a PLG funnel **Example:** a labeled hypothetical. A team collaboration tool gets 10,000 sign-ups a month. Of those, 25% activate within seven days, and 6% of activated accounts start paying within 90 days. A paying account has four seats at $10 per seat per month. | Stage | Calculation | Result | |---|---|---| | Sign-ups | given | 10,000 accounts | | Activated | 10,000 x 25% | 2,500 accounts | | Paying | 2,500 x 6% | 150 accounts | | Revenue per paying account | 4 seats x $10 | $40 a month | | New monthly recurring revenue (MRR) | 150 x $40 | $6,000 | | Overall free-to-paid | 150 / 10,000 | 1.5% | Now change one input at a time. - **Activation from 25% to 35%:** 10,000 x 35% = 3,500 activated; 3,500 x 6% = 210 paying; 210 x $40 = $8,400. New MRR rises by $2,400, or 40%. - **Price from $10 to $12 per seat, same conversion:** 150 x 4 x $12 = $7,200. New MRR rises by $1,200, or 20%. In this example, the onboarding fix is worth twice the price rise. The arithmetic points to a pattern worth testing: growth can sit in the first 10 minutes of the product, not only on the pricing page. Run the same calculation with your own numbers before you decide where to spend. ## The hybrid model: PLG plus sales Pure PLG often stalls above a certain contract size. A buyer who signs a large annual contract wants a security review, a custom agreement and someone to call. Slack's filing shows one answer: keep the self-serve entry and add a sales team for larger organizations. In a hybrid model, the product brings in accounts and the PQL signal decides which ones a salesperson contacts. The free plan stays open to everyone. Sales spends time only where usage already shows intent. For how this changes acquisition cost and payback, see [LTV and the LTV:CAC ratio](https://saascron.com/saas-ltv/). ## When PLG fails: a fit test PLG is not a fit for every product. Use this table to test yours. Each row is a condition from the mechanics above; the more rows land in the right-hand column, the more you need sales. | Question | PLG fits when | Sales-led fits when | |---|---|---| | Can one person get value alone? | Yes, in one session | Value needs a team, data import or setup project | | How long until the first useful result? | Minutes to a day | Weeks | | Who decides to buy? | The user, on a card | A committee, through procurement | | What does a contract cost? | Low enough to expense | High enough to need approval | | Does use spread inside a company? | Yes, people invite colleagues | No, each seat is a separate decision | | What does a free user cost you? | Little (light hosting, no support) | A lot (compute, data, support time) | The last row decides whether you can afford a free plan at all. If each free account costs real money in compute or support, a time-limited trial limits the bill. The trade-offs between free plans, trials and paid tiers are set out in [SaaS pricing models](https://saascron.com/saas-pricing-models/). ## Common mistakes 1. **Calling a free plan "PLG" without self-serve onboarding.** If users must book a demo to see value, the free plan only collects email addresses. Why it happens: a free tier is easy to add, onboarding is hard. Instead: measure activation first and fix the path to the first result. 2. **Measuring conversion against the wrong base.** Free-to-paid against all sign-ups and against activated accounts can differ fourfold, as in the example (1.5% against 6%). Why it happens: dashboards pick one without saying which. Instead: write down the base and the time window next to every rate. 3. **Giving away the feature people would pay for.** If the free plan solves the whole problem for most users, few upgrade. Why it happens: generous limits look good at launch. Instead: put the limit where a team's need grows, such as seats, history or admin controls. 4. **Hiring sales too late, or too early.** Without sales, large accounts stall in procurement; with sales too early, reps chase tiny accounts. Why it happens: PLG is treated as all or nothing. Instead: add sales when PQLs and contract sizes justify it, as Slack's filing describes. For how free entry, conversion and churn add up month by month, see [the SaaS business model](https://saascron.com/saas-business-model/). ## Sources - Slack Technologies, [Form S-1](https://www.sec.gov/Archives/edgar/data/1764925/000162828019004786/slacks-1.htm), 2019, figures as of January 31, 2019. - Atlassian Corporation Plc, [Form F-1](https://www.sec.gov/Archives/edgar/data/1650372/000104746915008450/a2226437zf-1.htm), filed November 9, 2015. ## In three sentences Product-led growth lets people sign up, reach a useful result and pay on their own, through a free trial or freemium plan, self-serve onboarding and in-product expansion. In the worked funnel, raising activation from 25% to 35% adds 40% more new MRR, twice the effect of a 20% price rise. PLG fits products that one person can value in minutes and buy on a card; when contracts need committees and setup projects, add sales on top of the free entry, as Slack did. --- Canonical: