Case Study
How we scaled signups for a b2c SaaS startup
Muse is a SaaS startup that offers a super easy to use platform to creators for selling their digital products. They are direct competitors of brands like Gumroad, Podia and Stan. When we started, Muse was already live for some months and had been generating traffic to their own landing pages coming from Google ads.
They asked for our help with finding product-market fit and generating more leads that turn into paying users.
We started off with doing 2 separate audits for Muse, to fully understand their current online marketing setup. Also we like to do audits for ourselves to understand the product/ service in an even better way.
1
Google Ads account audit
Some takeaways from the Google ads audit that we did for Muse:
Some campaigns were using the word ‘free’ in keywords and in ad copy, which of course suggests that Muse is free to use. Since Muse is a freemium platform, it is completely true that it can be used for free, however it will probably not generate much revenue by advertising on that term.
All of the campaigns were tracking landing page button clicks as if those were real sign ups. However, a sign up still had to be done through a form after clicking the button. After closer analysis we saw that there was a >30% drop-off between clicking the sign up button on the landing page and actually signing up. So this ‘conversion’ metric that the campaigns were using to optimize and also the CPL (cost per lead/ sign up) shown in Google Ads reports was not accurate. It gave the impression of leads being generated at a much lower price per lead.
2
Tracking audit
For us this was the most interesting audit that we did for Muse. It really made us go deep into their MarTech stack as well as into their tracking code and how it all was connected and relating to each other. We looked into all of the tools they were using for tracking conversions: Google ads, Google analytics, Google tag manager and Mixpanel.
We delivered a very detailed tracking audit report with many recommendations. These are the key takeaways that are worth mentioning:
About a month before we started there was a major drop in conversions from one day to another. This drop couldn’t be explained at first and had to be further investigated.
We noticed that there were quite large differences between numbers of conversions registered in Google Ads and those in Mixpanel, where they should be exactly the same. We found out that the reason for that was UTM parameters got mixed up between Google ads campaigns, resulting in not being able to compare conversion stats in Google ads and mixpanel.
Conversion tracking code was not added to some buttons on landing pages and therefore resulting in a lot less conversions tracked
There was both Google Ads and Google analytics tracking code present on pages, next to also GTM containers and custom tracking code on buttons. This made understanding the metrics more confusing
1
Switching to Meta Ads
About a month after we started setting up new Google ad campaigns for Muse we found out that we were hitting a traffic ceiling. We weren’t able to get more relevant traffic unless we would allow a further increase in cost per lead (CPL). Also after doing optimizations we weren’t able to decrease the CPL below target. This showed that we hit a ceiling and wouldn’t be able to scale with Google ads as much as we were hoping.
We then suggested testing Meta ads as this is a platform that many creators (Muse’s target audience) are actually using (Instagram, Facebook). This turned out to be the right decision and delivered great results from the start. We started with testing a lot of different creatives for the top-of-funnel campaigns that we were slowly scaling up. Later on when we gathered enough data we added campaigns based on look-a-like audiences. After a while we also added retargeting / upsell campaigns to generate paying users out of the non-paying sign ups.
2
Fixing the conversion tracking setup
First we decided not to primarily use Google analytics or Google Ads built-in conversion tracking system, because of the inaccuracy that occurs at a rapidly increasing rate these days. Because of European regulations, popular browsers are making it much harder to track conversion based on cookies only, sometimes tracking only 60-70% of actual conversions. Therefore we decided to create an advanced tracking setup for Muse, based on cookieless server-to-server conversion tracking that also posts back conversions into Google ads and Meta ads (for reporting and optimization purposes within those platforms).
To fix tracking button clicks instead of real conversions we set up a s2s conversion trigger coming from the product itself (after confirming the signup by email), so that we were sure to only measure real and confirmed sign ups. We then set up sign ups as primary conversions and landing page button clicks as secondary, to be able to have insights in the drop off in between those two.
We implemented built in {campaignid} tokens into the Google ad campaign UTM parameters used in landing page URLs, instead of hard coding them into the URLs. This way the campaign IDs couldn’t be mixed up anymore.
Of course after fixing the tracking issues, it turned out that Muse was generating a bit less sign ups compared to what they thought. That also meant the real cost per lead was higher than expected. Although these numbers were not ideal, they at least were accurate now, which gave them even more reason to try something other than Google Ads.
Switching to Meta ads was a good choice because it lowered the CPL by ±50% and we doubled the amount of leads. Also by making changes in copy and the overal ‘message’ Muse is sending out, we managed to generate more paying users out of the ‘free plan’ sign ups.