Inside GenPPT’s A/B Testing Strategy That Improved Conversion Rate by 136%

GenPPT increased its conversion rate from 3.85% to 9.09% by restructuring how users encountered its paywall.

No new features. No growth hacks. Just one change: replacing an upfront decision with a low-friction trial.

The Old (Complex) Paywall Blocked Conversions

The original pricing page forced users to make a complex decision:

• $9/month Pro plan (15 presentations/month)

• $19/month Growth plan (unlimited)

• $39 one-time plan (3 months access)

Each plan used a different logic — per month, per credit, or per term. The old pricing had a choice overload bias.

Users dropped off due to decision fatigue. There was no free trial. Cold users had to commit with no experience. For first-time visitors, it felt much more like a gamble.

The Newer (Easier) Paywall Did Less, Better

The trial offered clarity:

• 3-day free trial of the full Pro plan

• Card required, billed after trial unless canceled

• Headline: “Free for 3 days. $9/month after. Cancel anytime.”

The CTA shifted from “Start Creating” to “Try Pro for Free,” reducing pressure. One plan, clearly explained. No pricing grid. No overload.

A/B Testing Revealed the Problem

Using PostHog, an A/B Testing SaaS, GenPPT ran a clean split test:

• Control (no trial): 1 purchase / 26 users → 3.85%

• Variant (3-day trial): 8 purchases / 88 users → 9.09%

A 136% improvement — purely from reframing the paywall. No more friction across the funnel.

Users can try the product before paying or using CC. More trials means more activation loops / emails.

The test required no new features. Just a restructured flow. It was tracked end-to-end with PostHog.

Ilias Ism, the owner of GenPPT, covered how he uses PostHog for A/B testing in this hour long YouTube tutorial.

Top Tweets of the day

1/

Every CEO says a lot of shit when the market is good and they have virtually no competition.

You'll see their principles change when they start making less money.

Bezos did it with ads. Speigel did it when Snapchat’s revenue started tanking.

2/

OpenAI's Codex is a mind-blowing workflow. They took Claude Code's Terminal Based Agents to a whole another level with a GUI. OpenAI is going for the consumer while Claude is tackling purely developers.

Steal the good ideas from your competitors and one up them.

One of OpenAI's founder Greg Brockman has even said 2025 is the year of agents. And I've been hearing a lot of noise on Claude Code being the best coding agent at the moment for $100 Max plan.

Expect this to happen in other fields.

Coding is based on input and output. Whatever you tell the machine, it will do it exactly as you said. So its easy to test and get feedback.

Writing tests are hard for other things like marketing or design since these are creative fields. There is a formula but not many people have formulated it. Some people who have come close to formulating marketing are in 2 books: Marketing Memetics and $100M Offers.

So whatever happens in coding will be followed up in other industries. The smart play would be to use the knowledge of Agents in Coding industry and make a product where there is no competition like Medical, Sales, or NSFW industry.

3/

Watch what people do, not what they say. They show you their real priorities with their actions. Words are just a filler to earn cool social points.

Rabbit Holes

1. How to Get $300 in Free AI Credits Using Google Cloud by Kilo Code

2. How to Think Like Rick Rubin. 20-minute crash course in creativity by Shaan Puri

3. Claude Code: Anthropic's Agent in Your Terminal by Cat Wu and Boris Cherny

What’d ya think of today’s newsletter? Hit ‘reply’ and let me know.

Do me a favor and share it in your company's Slack #marketing channel.

First time? Subscribe.

Follow me on X.

More Startup Spells 🪄

1. Quittr’s $100 Meme Ad vs. OnlyFans: How a Porn Recovery App Pulled in $21,758 (LINK)

2. Reverse-Engineering Distribution (via Open Analytics) (LINK)

3. LogRocket’s Winning Global Outsourcing Playbook: 450 Developers, $350 per Article, 3M Views Per Month (LINK)

4. JSAUX's Influencer Seeding Strategy (LINK)