How I Increased my SaaS Conversion Rate by 700%

Want to transform your whole business in just a few simple steps? Well this might not do that, but it's an interesting story...
by Jon Yongfook ·


    First of all, apologies for the title. This is not a step by step guide of how to 7x your profits overnight.

    This is a story about the importance of good data, and how forgetting the basics can potentially lead you to make bad business decisions.

    Poor Conversion Rate

    Bannerbear has always had a low-ish conversion rate from signups to paying customer. But at some point in mid-2022 it settled down to 2% and we even had a 1% month recently.

    When it was hovering at 3% or 2% I never really gave this much thought, as the number of new subscribers was acceptable. I just figured we had a lot of "tire kicker" low-intent signups who were just checking out the product with no intention to buy.

    When we had a 1% month though, I started to think hmm ok something is wrong.

    This 1% month was during the July SaaS Summer Slowdown (TM) so a slow month was somewhat expected… but still, as a SaaS founder when you see a 1% conversion rate it unsettles you.

    I immediately started thinking about ways to increase this.

    Brainstorming Fixes

    I always thought we had a decent onboarding. Onboarding can always be improved and tweaked of course, but the same onboarding had got us from 0 to $50K in MRR, so it couldn't be that awful right?

    I brainstormed in public on Twitter about possible fixes, got some mixed responses from people saying things like:

    • Your onboarding is trash
    • Your onboarding is fine
    • Have you considered tearing everything apart and starting again

    But one reply in particular got me thinking.

    Monica Lent asked a simple question:

    Are you sure these signups are "qualified"?

    Meaning, am I certain these are high quality or high intent signups?

    It raised a doubt in my mind for the very first time… what if a big chunk of my signups are bots?

    Inspecting the Data

    I pulled up the console and did a query on the last 200 signups to check two things: their email, and if they had created any projects in Bannerbear (basically if they had taken the first step of the onboarding, by clicking one button).

    It was super obvious to see the pattern. Tons of gmail addresses with names from a particular country, who had zero account activity.

    So what's actually going on here?

    Jesse Hanley hypothesised that it was a form of low-level List Bombing. Because we send out an email confirmation after signing up, some nefarious actor was using our signup form (and most likely many many others) to perform email bomb attacks on unsuspecting individuals.

    Whatever the actual reason, things looked shifty and I wanted to clean this up. It was dawning on me that my actual conversion rate looked nothing like 1%.

    Bringing Down the Hammer

    With a few clicks of my Cloudflare console, I simply added a captcha to the signup form, which only displays for countries I suspected of abusing my signup form.

    The captcha was implemented on the 29th of July. Signups dropped like a rock.

    This whole time I thought Bannerbear was getting 3000+ organic signups per month, but only 60 paying customers per month - clearly some sort of onboarding issue, not getting enough people to the aha moment etc.

    In fact, it now looks like our organic signup numbers are merely a fraction of that.

    It is much less of a conversion rate issue, and simply more of a distribution issue. In order to grow, we need more high quality signups.

    Encouraging More High Quality Signups

    So just get more signups then! I think every founder on the planet wants this, so it's not exactly something simple to do. You can't just wave a magic wand and get more high quality leads into your business. But here are some things we have implemented…

    Mandatory Plan Choice

    Bannerbear has always been a "no credit card required" signup, and we intend to keep it that way.

    But previously all of the various Sign Up CTAs on the marketing site would bring you directly to the signup form.

    Now all of those CTAs bring you to the pricing page, so that you have to "choose a plan" to proceed. Credit card not required.

    I think this just helps to signal that hey btw, this is a paid product. Helps to turn away people who thought it might be some kind of free service.

    Exit Intent Modals

    We get a lot of traffic to our blog and various free tools, and now we are a bit more aggressive in asking users to sign up. Whereas previously the bottom of blog CTA was a link to join our mailing list, it's now a sign up CTA.

    There are also exit intent modals, which pop up if you go to close the tab.

    This helps to funnel a bit of that blog traffic into the main homepage, where they can actually learn and understand what our product does.

    Starting Google Ads

    We have also started a simple Google Ad campaign which is getting us more high-intent clicks for specific terms and keywords. Currently this is set to a low budget since we aren't experts at this.

    Our New Conversion Rate

    So what is our new conversion rate?

    It's around 7% to 8%.

    Note that the screenshot below was taken 10 days into 2023-08

    If we assume that the signup and conversion rate will remain stable through the month, then factoring in the other 21 days of the month we should be looking at 800+ signups and 57 new customers.

    In other words, the same number of absolute new customers as before - or perhaps better, if you want to account for the SaaS Summer Slowdown.


    So if we haven't significantly moved the needle in terms of new customers, what was the point of all this?

    Firstly I've learned that our true trial conversion rate isn't nearly as bad as I thought it was. It still needs improving, but it's not as completely broken as the data was previously showing.

    Secondly I need to place more emphasis on distribution. I thought we had things all figured out and were getting plenty of signups per month, but those were mostly bots. So I need to look for new opportunities to drive qualified signups and not just rely on our organic traffic.

    Thirdly I should be taking periodic looks at the basic data. If I had a dashboard showing onboarding steps completed I would have noticed something was off almost 10 months ago when our signup rate started to increase.

    Running a SaaS is a journey of learning!


    Now that the above experiment has run a full month, I have the updated data. Above I mentioned on day 10 that if the new rate of signups and conversions remained constant through the month, we should be looking at 800+ signups and 57 new customers.

    The result at the end of the month? 754 signups and 64 new customers.

    One of our best months ever - a great outcome!

    About the authorJon Yongfook@yongfook
    Jon is the founder of Bannerbear. He has worked as a designer and programmer for 20 years and is fascinated by the role of technology in design automation.

    7 Lessons Growing a Bootstrapped SaaS to $50K MRR

    In this blog / video I share a few tangible lessons that I have learned, specifically targeting technical founders who are bootstrapping their own ventures.

    How My Startup Survived a DDoS Attack

    Bannerbear was DDoSd but we survived - here's how the day went and how we ultimately mitigated the attack

    3 Reasons Why I Think 50% Coding 50% Marketing is the Best Framework for Solo Tech Founders

    Almost from the beginning I have employed a 50:50 split between coding and marketing for my SaaS product Bannerbear. Here's why I think it works.

    Automate & Scale
    Your Marketing

    Bannerbear helps you auto-generate social media visuals, banners and more with our API and nocode integrations

    How I Increased my SaaS Conversion Rate by 700%
    How I Increased my SaaS Conversion Rate by 700%