What is Product-Led Growth and how is it different from regular SaaS marketing?

PLG is a growth model where the product itself is the main channel for attracting and retaining customers. The user finds a product, registers, receives value - and only then sees the sales team. In the traditional model, sales occur before the user has tried the product. PLG key indicator: if you remove the marketing department for a month, new users will still come through organics and virality.

What conversion from free to paid is considered normal with freemium?

The median for B2B SaaS is 2-5% of all free users. The best PLG products (Notion, Figma, Slack) hold 3-8%. If the conversion is below 1%, the problem is either in the product (no value on the free plan) or in onboarding (users don’t get to the aha moment). If it’s above 10%, most likely the free plan is too limited and scares off potential users at the start.

When does it make sense to launch a freemium and when a trial?

Freemium works for products with network effects, low cost of serving a free user, and viral mechanics (the user invites others). Trial for 14-30 days is better suited for complex products that require onboarding from a CSM, and for B2B with clear decision-makers. Usage-based (pay for use) - for API, infrastructure and analytical products, where the value is directly proportional to the volume. In practice, 70% of successful PLG companies combine models.

What is PQL and how is it different from MQL?

MQL (Marketing Qualified Lead) - a person who has shown marketing interest: downloaded the guide, visited the pricing page, filled out the form. PQL (Product Qualified Lead) is a user who is already working in the product and has achieved a key event that correlates with a purchase. PQL converts into paying customers 3-5 times better than MQL, because the person has already seen the value with their own hands, and not from a landing page.

How to measure a product's viral coefficient?

Viral coefficient (K-factor) = number of invitations from one user × conversion of invitees to registration. If K is greater than 1, the product grows on its own without external marketing. Most B2B SaaS have K = 0.3-0.7: a good addition to other channels, but not an independent engine. Measure through referral codes or unique invite links - Amplitude counts this out of the box.

What NRR does a PLG company need to be considered healthy?

An NRR above 100% means that expansion and upselling are blocking the churn - the product is growing even without new customers. For the early stage (up to 1 million ₽ MRR), the NRR norm is 90-110%. For a mature PLG product the target is 120%+. Hubspot holds 105%, Snowflake has historically shown 150%+. If the NRR is below 80%, the product is losing money from its existing base faster than attracting new ones.

All articles

StrategyMarch 3, 202611 min

Product-Led Growth 2026: freemium, product virality and PLG metrics

How to build growth through the product itself: freemium, trial, viral mechanics. When PLG works and when it doesn't. Trial→paid conversion benchmarks, NRR and retention for B2B SaaS.

PLG is not just freemium. This is when the product itself is the main sales channel: the user registers, receives value, and only then meets the seller. Marketing doesn’t sell the product—the product sells itself.

I have advised three SaaS companies that have decided to migrate to PLG in 2024-2025. One project is an HR platform with a check of 180,000 ₽/year, the second is an analytical tool for e-com, the third is a B2B messenger for production teams. The free→paid conversion before the introduction of PLG for all three was below 1%. After properly built onboarding and PQL logic - in the range of 3-7%. I tell you what worked and what I had to throw away.

If the user did not understand the value of the product in 10 minutes, no amount of retargeting will help.

1. What PLG is (and what it is not)

The most common misconception about PLG is that it is just a free plan. No. Freemium is one of the PLG tools, not PLG as a whole. Product-Led Growth is a strategic decision in which the product becomes the main driver of user acquisition, activation and retention. Sales and marketing exist, but they work on top of the product funnel, not instead of it.

Practical test: if you turn off all paid advertising channels tomorrow, will there still be new registrations in a month? In Slack, Figma, Notion - yes, because users invite colleagues from within the product. In a typical corporate SaaS with “demo on request” - no, because each lead requires manual work from the sales team.

PLG doesn't mean no sales either. The best PLG companies use a “product-led sales” model: the sales team gets involved exactly when the user has already received value and is ready to talk about money. CAC in this model drops by 40-60% compared to classic outbound, because you are calling warm people, not cold ones.

2. Three models: freemium, trial, usage-based

The choice of monetization model within PLG determines everything: onboarding, activation metrics, upsell moment. Each model has its own conversion mathematics and its own risks.

Model	Conversion free→paid	Benefits	Risks	Better for
Freemium	2–5% (median), 8–12% (top)	Low barrier to entry, virality, organic growth	High cost of free users, diluting value proposition	Products with network effects and low COGS
Trial (14–30 days)	15–25% (with active onboarding)	A clear deadline motivates action, it’s easier to explain the value	Pressure on the user, strong onboarding is needed, high churn after the trial	Complex products, B2B with decision makers, high average bill
Usage-based (pay-as-you-go)	40–70% (those who started paying pay)	Revenue growth is proportional to client success, low entry barrier	Unpredictable MRR, complex financial model, risk of “unexpected bill”	API, infrastructure, analytics, AI tools

In practice, companies mix models. Figma - freemium from trial to pro plan. Stripe is usage-based with a fixed subscription fee for added features. Slack is freemium with a limit on message history, which creates natural pressure to upgrade. Look at what your business creates and how users get value - and choose from there.

One of the projects I worked with sat on trial for 7 days and complained about the conversion of 8%. We have increased the trial to 21 days and added a personalized checklist in onboarding. Conversion increased to 19% in two months. The point was not in the length of the trial, but in the fact that in 7 days users simply did not have time to reach the aha moment.

3. Viral mechanics inside the product

“Viral loop” in PLG is not a referral program with cashback. This is a situation built into the product where the user himself wants to invite others because it makes the product better for him. Three working patterns:

The first is collaboration. The user creates something valuable (document, design, report) and wants to show it to a colleague. Figma lives on this: sharing a file with the editor is the most common action in the product. Every new collaborator is a potential user.

The second is public results. The product generates something that the user wants to publish: a public link to a dashboard, an embeddable widget, a training certificate. Each publication is an advertisement for a product, zero costs for it.

The third is the network effect. A product gets better when it has more participants. Slack is useless without a team. Notion with a team is 5 times more valuable than Notion for one. This is the most powerful “viral loop” - but also the most difficult to build: you need the value to actually grow with each new participant.

The viral coefficient (K-factor) is calculated simply: the average number of invitations from one user, multiplied by the percentage of those who accepted the invitation. K above 1 - the product grows without additional investments in marketing. Most B2B SaaS have K = 0.3-0.6: a good addition to other channels, but not the only driver. And that's okay - PLG doesn't require K greater than 1 to work.

4. PLG metrics: PQL, time to value, NRR

A standard marketing report (CAC, CPL, ROAS) for a PLG company hides half the picture. Three additional metrics are needed. Read more about unit economy SaaS - a separate article, here only PLG-specific.

PQL (Product Qualified Lead) — a user who reached a key event in the product and showed behavioral signs of willingness to pay. Not “downloaded the white paper”, but “created 3 projects and invited a colleague.” PQL converts to paid accounts 3-5 times better than MQL because the person has already experienced the value. Determining PQL is the most important analytical task when starting a PLG: take the last 50 converting users, find the general behavior in the first 7 days.

Time to Value (TTV) — median time from registration to the first “aha” moment. If TTV is more than 30 minutes, onboarding kills PLG. Goal: 5-10 minutes for simple products, no more than 30 minutes for complex ones. Measure via Amplitude or Mixpanel: time between event “registered” and event “first key action”.

NRR (Net Revenue Retention) — percentage of revenue retained from a cohort of clients, taking into account upsell, downgrade and churn. An NRR above 100% means that the product is monetizing existing customers better than losing them. This is the main indicator of “product-market fit” for PLG. Hubspot holds ~105%, Snowflake at its peak showed 158%. For Russian B2B SaaS in 2026, 95-110% is considered the norm.

The relationship between metrics works like this: good TTV leads to a high PQL-rate → PQL converts to paid ones → paid ones increase usage → NRR grows above 100%. If any link breaks, look for the problem there. Most often it is TTV that breaks.

5. When PLG doesn't work

PLG is not a silver bullet. I've seen three companies spend 6-12 months "going PLG" and lose momentum because the product fundamentally didn't fit the model.

PLG does not work for products with long setup times. If to start working with the product you need integration with corporate ERP, training the team for 3 days and signing an SLA, the user physically cannot receive the value on his own. What is needed here is white-glove onboarding, not self-serve.

PLG does not work when the value is visible only after a large amount of data. Analytical platforms that become smarter after 3-6 months of use do not convert well into trials: the user leaves without waiting for the result. The output is demo data or “quick win” mechanics that provide the first value in the first session.

PLG does not work with purchasing committees of 5+ people. If the decision to purchase is made by a lawyer, CTO, CFO and two heads of departments, one self-serve PLG user will not push through. A sales cycle is still needed here, it’s just that PLG helps you come to the first call with an already trained champion within the company.

If you recognize your product in one of these points, PLG as the main model will not suit you. But PLG elements (normal onboarding, product analytics, PQL alerts) are useful in any SaaS. More about this in the material about building a sales funnel for B2B.

6. How to implement PLG into an existing product

The transition of an existing SaaS to PLG is not a one-time project, but 6-18 months of iterations. I worked with an HR platform that went this way in 9 months: from “everything through a demo call” to “60% of new clients are activated without sales participation.” That's what really helped.

The first three months are only analytics. Set up product analytics (Amplitude, Mixpanel or even Metabase on top of your database), determine the aha moment through cohort analysis, measure the current TTV. Without this data, any changes in onboarding are guesswork. Do not touch the product at this stage.

Months 4-6 - onboarding. Cut the path to the first aha moment in half. Remove any fields from the registration form that are not needed for the first session. Add an in-app checklist with three points (not twenty). Show empty states demo data - the user should see how the product looks “live”, not a blank screen.

Months 7-9 - monetization and PQL. Define PQL criteria and set up alerts in CRM. Add in-app upgrade offers when the limit is reached or after a key action. Launch trigger emails based on usage. These tools are described in more detail in the guide to CRM marketing.

On the HR platform, specific results after 9 months: TTV decreased from 45 minutes to 12 minutes, the trial→paid conversion increased from 9% to 23%, the share of independent activations (without the participation of sales) - from 15% to 61%. CAC decreased by 47% with the same marketing budget.

7. PLG-readiness checklist

Before you announce a “switch to PLG”, check the product against this list. If there are less than 7 points out of 10, first eliminate the gaps, then change the monetization model.

The user receives the first value in 10 minutes without the help of a manager
There is a clearly defined aha moment and it is measured through product analytics
Onboarding does not require filling out a form with more than 4 fields
You can try the product without a credit card or contract
There is at least one viral mechanism (sharing, collaboration, public results)
TTV is measured and is no more than 30 minutes (the goal is up to 10 minutes)
The PQL criterion was determined based on data, not on intuition.
Configured PQL alerts in CRM or sales tool
NRR is calculated monthly and is at least 90%
There is an in-app upgrade mechanism without the participation of a sales manager

PLG is an investment in a product that is returned through a decrease in CAC and an increase in NRR. A properly built PLG funnel allows you to spend less on paid traffic and more on developing the product itself. About how to calculate the unit economy of SaaS and when PLG starts to pay off - in the article about unit economy 2026. About how to build marketing for B2B SaaS on the Russian market - in the material about B2B SaaS marketing in Russia.

If you want to analyze a specific PLG funnel or calculate whether your product should switch to this model, write to Telegram or through form. Starting consultation - 0 ₽.