I’m seeking insights from growth leads and A/B testers who’ve run real-world experiments in SaaS or ecommerce. Can you share one growth experiment you ran that tied directly to product loops or funnel performance? Please include: (1) the hypothesis you tested, (2) what you changed (onboarding, paywall, pricing, feature, etc.), (3) the metric or funnel stage it impacted (activation, retention, revenue, or referral), and (4) what you learned, even if it didn’t go as planned. We want to showcase this among experiments that growth teams can use right away.

One of the most effective growth experiments we ran was testing demand with a newsletter before investing in the SaaS itself. Hypothesis: if we could breakeven (or profit) on user acquisition through a newsletter funnel, then there was enough market pull to justify building the SaaS for end users. What we changed: instead of launching straight into product, we built a Beehiiv newsletter and layered Boost + SparkLoop (affiliate marketing for other newsletters) on top. That let us offset ad spend with affiliate referral revenue, meaning growth was essentially self-funding. Metric: Once the loop worked, we optimized the newsletter it for SEO through a domain name rebrand into www.PRpackages.io. we eventually ranked #1-2 for "PR package" and "PR packages" across almost every country, bringing daily signups for our newsletter profitably. What we learned: starting with the newsletter gave us profitable acquisition, real proof of demand, and SEO authority. By the time we built and launched the SaaS, we already owned the audience, free traffic + SEO rankings, and brand trust — making the product funnel perform far better than if we had gone product-first.

We ran a 'Review Loop Acceleration' experiment to test the hypothesis that if we incentivized exhibitors to leave reviews immediately after the event, i.e., when their experience of the event was still fresh in their minds, we would increase volume and quality. This would then increase the strength of our core product loop as we would sign up more organizers as a result. We re-designed our post event onboarding flow by introducing a 48hr time-bound call to action with a non-monetary reward like leaderboard position. We also added smart nudges via email and in-app and removed the friction of having to complete entire profiles to submit a review. This had a direct effect on our activation and referral funnel stages. We observed a 37% increase in review volume and a 60% decrease in time-to-first-review that began to act as social proof helping organizers close deals more quickly. The main insight was that timing was everything. The prompt failed during the event when exhibitors were busy. We also learned that incentives of the non-monetary variety (i.e., tied specifically to reputation and visibility) drove more usage and engagement than simply throwing money at users. This demonstrates that optimizing for the user's emotional peak delivers a really strong, virtuous growth loop.

At Magic Hour, we tested whether showing new users AI-suggested project templates during onboarding would increase activation, compared to a blank slate dashboard. It boosted first-week retention by about 18%, and I realized firsthand that reducing cognitive load upfront matters more than dazzling someone with limitless options early on.

Here's a neatly tested war story from the trenches of growth experiments. This strategy can shamelessly drop into a "growth hacks we actually tried" slide in the deck: Experiment: SaaS onboarding and the paywall timing The hypothesis was like this: if we delayed the paywall prompt until after users hit their "free trial". Mostly, like using one key feature at least twice. More users would use the app and eventually convert. What we changed after the testing. So originally, users hit a paywall just after signing up. So we made a shift to something meaningful. So the paywall is triggered after completing two meaningful actions in the app. Funnel Stage: In primary, we got new signups completing both actions, and in secondary, we received paid conversions within 14 days. Outcome: Activation jumped noticeably, and conversions rose modestly. The twist: revenue per user dipped slightly because late paywalls encouraged "try before you buy" behaviour. Net takeaway? Delaying friction can boost activation, but you need to balance it against monetisation velocity.

We once hypothesized that layering social proof into onboarding would make users feel the platform was more trustworthy right away. By adding a simple note like '300 shoppers saved this deal last week,' 7-day return visits rose significantly, but interestingly, initial sign-ups dipped, teaching us there's a balance between trust-building and overwhelming new users.

We kicked off a test with one of our fashion brand clients years ago. I wanted to find out if showing limited stock would affect the purchase decisions of their customers since they haven't tried it yet. The idea was to show that "low stock" warnings would push people to buy faster. My thinking was that unsure customers would make up their minds quicker if they thought the product might disappear earlier than expected. So, we set up a stock warning near the buy button showing "Only X Left in Stock" for items under 10 units. Now,we placed it right below the purchase button so people would catch it when choosing. Our test mainly hit our client's conversion stage of the funnel, changing browsers into customers. We then tracked the upticks in add-to-cart actions and checked which ones resulted in completed orders. To see the results, we compared the pre-test and post-test purchase rates. We also monitored the revenue contribution from products that displayed the low-stock message. After our careful evaluation, the purchase rates went up for items that were truly almost sold out, proving our idea was right. Shoppers really liked real stock warnings to be included on the site because they give them time to decide to purchase it now or later. We saw that urgency messages are best kept to items that are genuinely running low, not used as a blanket rule.

At my health and wellness company, I decided that experimentation would be more effective than guessing, and so we ran a test related to an onboarding process in which we hypothesized that early education on sauna health benefits directly increased our client retention. We found a lot of first-time users didn't realize all the ways they best help when it comes to pain relief, relaxing and breathing exercises. We redesigned the first three touchpoints of our onboarding: Instead of opening with a bland welcome, we dropped a brief alert experience explaining what was happening in your body when you were in the heat and a simple breathing practice that you could try in your first session. The effect was maximal in the activation phase. We followed session completion rates and the rate of users who booked a follow-up in the first two weeks. The marks perceived weren't astronomical, because we hadn't changed the product; it was just the customers immediately recognizing and seeing value in the practice. What I learned was that small educational interventions directly related to the product loop (in our case, how more people feel comfortable spending time in saunas multiple times per week is part of wellness) can move the needle QUICKER THAN ADDING NEW FEATURES. What I would recommend to other teams: consider the emotional or psychological state you need customers to be in from the outset — and build that dynamic into your right-out-of-the-box experience.

Regarding the hypothesis, we improved onboarding using interactive prompts that increased the speed of user activation. Regarding the changes, we implemented a step-by-step guide to the key features of the product and a description of the static features. Also, the metric Activation (the number of users who completed the first successful use case). And as a conclusion, our engagement increased by 30 percent, but we found that too long guides scared some users away. The next step is to optimize the length and make it personalized.

I once ran an experiment where we shifted patient testimonials from the bottom of a landing page to right before the consultation booking form, and the hypothesis was that seeing authentic stories earlier would boost trust. It lifted consultation requests by about 14%, and what I really learned was that small tweaks in credibility placement can matter more than adding new features altogether.

1The hypothesis was simple: make our paid subscription more flexible by adding a no-obligation monthly plan option to attract skeptics. 2. We introduced a monthly plan with the ability to easily cancel, and also made the list of features in the plans more transparent and understandable. 3. This affected activation - the number of users who completed the first task increased by 30%. 4. The flexible plan attracted new customers, but many of them used the monthly plan without switching to more expensive options, so additional work was needed on upselling.

For Tutorbase, we ran an experiment around pricing transparency in the onboarding funnel. Previously, trial users had to reach out for a custom quote, so we tested a simple tiered pricing table tied to staff size right after signup. Conversions to paid accounts increased by 14%, and I learned that clarity and speed at the paywall are often more valuable to educators than highly customized pricing.

At Dirty Dough, we tested the hypothesis that lowering the friction on our franchise inquiry form would boost conversion from interest to qualified lead. By simplifying the onboarding form from 12 fields down to 4 essentials, we saw a 37% lift in completed applications, which taught me that speed of entry matters more than deep upfront data collection.

At Lusha, we hypothesized that asking for credit card info upfront during trials was cutting activation in half. So, we ran an experiment offering a card-free trial alongside the standard one. Trial-to-paid conversions held steady, but signups tripled, showing that lowering friction early in the funnel can expand the top without hurting long-term revenue.