TL;DR: Product led growth marketing breaks when ad platforms optimize for cheap signups. Amplitude found over 98% of new users go inactive within two weeks for half of all products. Define one activation event, feed it back to Google Ads, bid to PQLs, and score every channel on CAC payback.
What product led growth marketing asks of paid media
In product led growth, the product does the selling. Someone signs up, uses the tool, and buys because it worked. Paid media has one job: put the right people in front of the product.
That job sounds simple. It breaks in a predictable way. Ad platforms optimize toward whatever event you feed them. Feed a signup event to Google Ads Smart Bidding and it will find you the cheapest signup on the internet. Cheap signups are cheap for a reason.
The result is a dashboard that looks healthy above a product that looks empty. Cost per signup falls 30%. Activation falls faster. Two quarters later the sales team says paid trials never close, and nobody can point to the month it went wrong.
This is a build order for avoiding that. Define activation, send it back to the ad platform, bid to it, then score every channel on payback instead of volume. For the wider program view, start with our SaaS PPC playbook, then come back here for the activation piece.
Why signup volume and activation pull against each other
Signup volume is easy to buy. Widen the match types, loosen the audience, add display, cut the form to one field. Volume climbs inside a week.
Every one of those moves also widens the gap between what the ad promised and what the product actually does. That gap does not show up as a refund request. It shows up as a dead account.
The size of the leak is public. Amplitude’s 2025 Product Benchmark Report, published September 30, 2025 from data across more than 2,600 companies, found that for half of all products analyzed, over 98% of new users are no longer active at the two-week mark.
The bar for good is lower than most founders expect. In the same report, getting 7% of a signup cohort to return on day seven puts a product in the top 25%. For enterprise products, median day-7 retention is 2.1% and the 90th percentile is 12.4%.
Read that carefully, because it cuts both ways. Most of the leak is onboarding and product, not media. Paid cannot fix it. Paid can absolutely make it worse, and usually does, by buying the users least likely to survive the first session.
The activation math most PLG campaigns skip
Cost per signup hides the only number that matters: what you paid for a user who actually used the product. Here is that arithmetic, built from two published inputs.
The media inputs come from iClick’s 2026 benchmark tables. Those are first-party numbers from accounts we manage, not industry medians. For SaaS and B2B software, our portfolio averages a $4.88 CPC and a 3.80% conversion rate. The retention inputs come from Amplitude.
| Input | Value | Where it comes from |
|---|---|---|
| Monthly paid budget | $50,000 | Worked example |
| Average CPC, SaaS and B2B software | $4.88 | iClick 2026 benchmarks, first-party |
| Clicks | 10,245 | Arithmetic |
| Signup conversion rate | 3.80% | iClick 2026 benchmarks, first-party |
| Signups | 389 | Arithmetic |
| Cost per signup | $128 | Arithmetic |
| Still active on day seven at the enterprise median, 2.1% | 8 users, $6,120 each | Amplitude 2025 Product Benchmark Report |
| Still active on day seven at the 90th percentile, 12.4% | 48 users, $1,036 each | Amplitude 2025 Product Benchmark Report |
Same budget, same ads, same landing page. The cost of a real user moves by roughly 6x, and none of that movement appears anywhere in the ad platform. That is the whole problem in one row.
Bid adjustments will not decide which end of that spread you land on. What you optimize toward will.
Five steps to run product led growth marketing without tanking activation
1. Define activation as one event with a deadline
Activation is not a login. It is the first moment the product delivers the thing the ad promised. Pick one event, attach one time limit, and write it down before launch.
Concrete versions: connected a data source within 24 hours (Segment, Fivetran), invited a second seat within seven days (Figma, Notion), sent a first campaign within 72 hours (Klaviyo, Mailchimp), completed a first successful API call within 48 hours (Stripe, Twilio).
One event, one deadline, agreed before the numbers arrive. Renegotiate the definition after a bad month and you no longer have a metric, you have a mood.
Most teams never get here. ProductLed’s benchmark study of more than 600 SaaS businesses, published February 5, 2025, found activation is tracked only 34% of the time.
2. Send activation back to the ad platform
Smart Bidding cannot optimize toward a number it never sees. Activation happens inside your product, days after the click, so somebody has to ship it back.
Two mechanics do this on Google Ads. Offline conversion import ties a product event to the Google Click ID (GCLID) you captured at signup and stored on the user record. Enhanced conversions for leads does the same job using hashed first-party data such as an email address.
Google Ads Help recommends starting with enhanced conversions for leads rather than standard offline conversion import, and reports a median 10% increase in measured conversions when advertisers supply first-party data alongside GCLIDs.
One date belongs on your engineering roadmap. Google states that from June 15, 2026, offline conversion import and enhanced conversions for leads uploads migrate to the Data Manager API and are blocked in the Google Ads API. If your upload runs through a script somebody wrote in 2022, that script has an expiry date.
3. Bid to product qualified leads, not signups
Once activation data flows back, make it the conversion action bidding chases. Signups drop to a secondary, view-only column. This feels wrong for about three weeks. Volume falls, cost per signup rises, and someone senior asks what happened.
What happened is that you stopped paying for users who were never going to activate.
The payoff shows in the ProductLed data. Free trials that use PQLs convert to paid at 25% on average, against a 9% median free-to-paid rate across all models. In the $5K to $10K ACV band, PQL-driven trials convert 39% of the time.
Adoption is still thin: only 24% of product-led companies report using PQLs at all. If you need the definition and a scoring model before you build one, read our breakdown of what a product qualified lead is.
4. Split campaigns by what the searcher already knows
Do not group campaigns by channel. Group them by how much the person understood before they clicked. Three buckets, three budgets, three activation rates.
- Brand. They typed your company name. Highest activation, lowest volume, cheapest clicks. Never let this traffic sit inside a catch-all campaign that takes credit for it.
- Category. They typed the job to be done, such as customer data platform or invoice automation software. Real intent, real competition, and the bulk of a healthy PLG budget.
- Problem. They typed a symptom, or they never typed anything at all (paid social, YouTube, display). Cheapest signups, weakest activation. Fund it as a test line, not a growth line.
Blend those three and the problem bucket eats the account, because it always wins on cost per signup and always loses on activation.
Performance Max needs a specific warning. Our benchmarks page notes that PMax cannibalizes Shopping in more than 80% of the accounts we review. Watch for the same behavior against branded search in a SaaS account: PMax absorbs your cheapest, highest-activation traffic and reports it back as incremental.
5. Judge every channel on CAC payback
Cost per activated user is a good middle metric. It is not the finish line. The finish line is how long the business waits to get its money back.
The 2026 SaaS and AI Performance Benchmarks report from Aleph and Benchmarkit, built on full-year 2025 actuals from 342 SaaS and AI-native software companies, puts median B2B SaaS CAC payback at 16 months. Top-quartile companies recover in six months or fewer. The bottom quartile takes 24 months or more.
Run that per channel, not per account. A channel at $60 per signup with a 22-month payback is worse than one at $190 with a 9-month payback, every single time, and the first one will look better in every weekly report you send. Our guide to the CAC payback period walks through the formula.
Where the signup came from changes whether it activates
Channel choice is an activation decision wearing a media costume.
Search puts you in front of someone who already named their problem. It costs more per click, and it should. iClick’s first-party 2026 benchmarks put SaaS and B2B software CPC at $4.88, against a $2.32 average across all the Google Search accounts we manage.
Paid social interrupts instead. That is the trade, not a flaw. Our portfolio CPM for SaaS and B2B software is $14.50, and LinkedIn sits higher still. HockeyStack Labs, analyzing $28 million in LinkedIn ad spend across more than 70 B2B SaaS companies, published CPCs between $10.48 and $15.72 on December 15, 2025.
Those signups also arrive with less context about what they just signed up for. They need a heavier onboarding sequence to hit the same activation rate, and that cost belongs in your channel math. We put the full comparison in LinkedIn Ads vs Google Ads for SaaS.
What to check in the first 90 days
You need 90 days to know whether a PLG paid program works. Check five things, in this order.
- Activation rate by campaign, never blended. If one campaign sits at 4% and another at 21%, you have a budget decision, not a product problem.
- Time from signup to activation. If it stretches week over week, the ads are promising something onboarding does not deliver in session one.
- Conversion lag. Product events land days after the click. Give the platform 30 days of clean uploaded data before you touch a bid strategy.
- Budget share in the problem bucket. If it drifts past 25% without a matching activation rate, it drifted because it looked cheap.
- CAC payback per channel. Recalculate monthly, on booked revenue, not pipeline.
What it looks like when this works: across B2B SaaS accounts, iClick’s published results include a 30% CAC reduction, a 60% lift in trial signups, and 2x demo volume. Trademark Bob founder Brad Walz credits us with dropping CAC by 30% in under 90 days. Talent Room CMO Oliver Auerbach says we 2x’d demo requests in Q1 without increasing spend. Across 250+ clients and $61M+ in managed spend since 2017, our portfolio average conversion goal lift is 41%.
Three mistakes that cost the most
- Optimizing to signups because that data is already wired up. Convenience is not a strategy. The two-week engineering task to upload activation events is the highest-return work in the whole program.
- Judging the program at 30 days. With a seven-day activation window plus a 30-day learning period, month one is noise. Month three is signal.
- Treating every SaaS account the same. A seed-stage self-serve tool and a 40-seat enterprise platform need different bidding, different activation events, and different payback targets. That is why we split our SaaS paid media work across B2B, early-stage, and enterprise accounts.
Start with the number that settles the argument
Every debate about PLG paid spend ends at one figure: how many months until the business gets its money back. Median B2B SaaS takes 16 months. Top-quartile companies take six.
Find out where you sit before you change a single campaign. Run your numbers in the free CAC payback calculator. It takes about two minutes, there is no sales call, and the answer tells you whether your activation problem is actually a media problem.
Related on iClick
Sources
- Amplitude, The 7% Retention Rule Explained (2025 Product Benchmark Report, September 30, 2025, 2,600+ companies)
- ProductLed, Product-Led Growth Benchmarks (February 5, 2025, 600+ SaaS businesses)
- Aleph and Benchmarkit, 2026 SaaS & AI Performance Benchmarks, CAC payback period (full-year 2025 actuals, 342 companies)
- Google Ads Help, About offline conversion imports and enhanced conversions for leads
- HockeyStack Labs, 2025 LinkedIn Ads Benchmark Report for B2B Marketers (December 15, 2025, 70+ B2B SaaS companies, $28M spend)
- iClick Advertising 2026 PPC Benchmarks, first-party data from managed accounts
- iClick Advertising case studies
Frequently asked questions
What is product led growth marketing?
Product led growth marketing is acquisition built around the product doing the selling. Ads drive a free trial or freemium signup, the user experiences value directly, and the purchase follows. Paid media’s role is narrow: send people who can reach that value quickly, then measure success by activation and revenue rather than by signup volume.
Should PLG companies bid on signups or demos?
Bid on neither at first. Bid on activation, the first in-product event that proves value. Signups are too cheap to be meaningful and demos ignore self-serve buyers. Once activation events upload cleanly to Google Ads, Smart Bidding chases users who actually use the product. ProductLed reports PQL-driven trials convert to paid at 25%.
How long before a PLG paid campaign shows real results?
Plan for 90 days. A seven-day activation window plus a 30-day Smart Bidding learning period means month one is mostly noise. Month two shows direction, month three shows signal. Judging performance at 30 days almost always pushes teams back toward cheap signups, which is the exact failure the activation metric exists to prevent.
What activation rate should paid signups hit?
Set the target from your own product, not an industry average. Amplitude’s 2025 Product Benchmark Report found 7% of a cohort returning on day seven puts a product in the top 25%, while for half of products over 98% of new users go inactive within two weeks. Compare paid cohorts to your organic ones.
Does Performance Max work for product led growth SaaS?
Use it carefully and never as a catch-all. iClick’s benchmarks show PMax cannibalizes Shopping in more than 80% of accounts reviewed, and the same absorption risk applies to branded search in SaaS. Run brand as its own campaign, exclude brand terms from PMax, and feed PMax activation conversions rather than raw signups.
