How my $670 ad experiment can save your marketing budget

Last Updated: Fri Feb 23 2024

LinkedIn, Reddit, Google Ads, Facebook have all happily taken my money - what did I get back?

For context we (Supatool) are a business SaaS product gearing up for a launch later this year. So I wanted to test out some ad copy, see what the response to ads was, and what features people clicked on the most.

Now, I know organic marketing is the best for stable growth (and we have a healthy organic waitlist), but as 70% of the top SaaS products market through ads I'm assuming it works for them. So I thought, why not give it a go and see what sort of response we get for our waitlist/lead generation.

I hadn’t originally intended to spend $670, but here we are. So sharing what I found with any other founders thinking about ads.

I went with four platforms:

LinkedIn since we have a business product
Reddit, I was curious.
Google Ads the old favourite (I was genuinely astonished by the conversion rate)
Facebook, lots of eyeballs

TLDR;

LinkedIn: $144.50, 0.63% CTR, $72.25 CPC
Reddit: $44.61, 2.60% CTR, No conversions
Facebook: $237.65, 1.8% CTR, $6.32 CPC
Google: $245.63, 8.84% CTR, $5.46 CPC (at a whopping, 35.71% conversion rate)

The setup

I gave them all the same daily budget, although some did demand more. Google initially said my campaign’s performance would be ‘hurt’ unless I had a daily budget of $140, but it worked out pretty well in the end with a fraction of that.

In order to evaluate lead quality on sign up we sent people a ‘please verify your email’ email. About 70% of Google Ad leads verified their email, neither of the LinkedIn ones did (and they both weirdly appear to be personal email addresses), and 50% of the Facebook ones did.

LinkedIn Performance

On LinkedIn I had high hopes. It presents itself as the B2B ad platform. I set up both the form based ads and display ads with images we had made. Initially things were looking good, impressions and a high CTR. A little over 1% at first, according to their marketing docs the global average is fixed between 0.44% to 0.65.

However, no conversions. Nadda.

Fake Linkedin Clicks?

I checked our analytics platform and something was irking me. All the LinkedIn ads clicks lasted between 6 and 6.1 seconds exactly. They were nearly all android, and despite being on the page for 6 seconds had absolutely no interaction with the page at all.

And what was weirder, they all seemed to have a mouse icon in our session replay tool. Normal android devices I’ve tested, and I’ve tested probably a dozen, don’t have this mystery cursor.

This is just me guessing, but it looks like it’s on an emulated android device. They all have the referrer https://aax.amazon-adsystem.com/ and are from either New York, or Chicago. They do all have the same GeoIP Latitude & Longitude too but given the way GeoIP works that could be normal.

Eventual Success

Anyway, once I unchecked “LinkedIn Audience Network” as a placement option we stopped seeing people referred from there. And although clicks did plummet, we finally got conversions. So in fairness to LinkedIn, perhaps I need to re-run this with that unchecked from the beginning.

The cost per conversion when the campaign was only on LinkedIn was ~$30. And clicks and impressions were noticeably lower, but conversions are the thing we care about!

Reddit Performance:

You may notice that Reddit has a noticeably lower spend than the others. $44 vs an average of $200 for the others. This wasn’t on purpose. I gave it the same amount of time and budget as the others but it just didn’t spend it.

It did have by far the lowest cost per click at 19 cents but 230 odd clicks later no tracked conversions. So Reddit may be useful for brand awareness, but the targeting options are pretty lacking.

The Bad

Also, the ads do just seem to blend into areas of the page the user wouldn’t really be looking at. Even on the ‘see your ad’ test pages I actually had to ctrl+f Supatool to see where the ad was. I had scrolled past my own ad without realising!

I will also say I lost a bit of patience with Reddit because the ad creation experience was more than a little irritating.

You can, like Google, get it to auto create ad versions but Reddit then makes these individual ads you have to control.

And part of Reddit’s setup is providing a destination url and a ‘display url’, i.e. where they go and the shorter url of what they see on the ad. The ad creator will automatically generate the display url for you, just your domain rather than the full landing page.

I say again, it does this automatically.

Yet my first slew of ads were all banned because the destination and display urls were different.

It made them different!

So I then had to go through 13 ads and manually make the change. If you have carousel ads, then you have to change the destination & display url on each image in the carousel…

Reddit also then banned several of the ads, and I’m still not sure why. The email they sent gave three reasons - none of which were true. I did get an email from a Reddit Ad Account manager and I sent the problem I was having to her, she then forwarded it on to a ‘Client Partner’.

The Good

However, I got around the problem by creating literally the exact same ads again, magically they were approved. So 🤷🤷🤷

Facebook Performance

I was pretty sceptical about Facebook ads. I associate Facebook & Instagram with B2C products but given they have over three billion active users I thought it couldn’t hurt to try.

I was impressed that they had a Form lead option, where instead of just a normal ad you see a form. Not quite as advanced as Google’s, which I’ll discuss last, but still it’s what we got all bar one of the leads from.

Targeting options were pretty great too, could dial it down to interests which did seem to work as people gave us their email address.

The Results

Although as noted above only a little over half of people verified their email address. I think this could be because Facebook likes people to stay on the platform, and generally they will, so by the time they come back to their email they might not remember why you’re asking them to verify their signup.

Time will tell how good quality these leads are, we’ll only truly know when we have a public product, and can see if they sign up and pay.

But what I did like about Facebook’s ad builder is it didn’t try and get you to spend big bucks coming out of the gate. I plugged in the campaign settings and it suggested an initial daily budget of $6.31. Obviously this would ramp up over time but unlike Google below, it seems like Facebook is a relatively inexpensive ad platform to start on and grow into.

Google Performance

I was pretty certain we’d get conversions with Google. I just wasn’t expecting as many as we got. The average conversion rate was 35.71%, and it was 38.4% for the Form ads.

I did spend the most with Google, $245, but we got by far the best results. The targeting options were, as you’d expect, pretty phenomenal.

If Google had it's way I'd have spent $140 every day of this campaign, any less it said and it would 'hurt performance'. Now I'm not adverse to spending money on ads, this post wouldn't exist otherwise, but $140 a day from the get go seemed a little too steep. So I ignored Google's protestations and entered $25 per day.

Fair play to Google they were the quickest to get conversions, and suggest changes to optimise the campaign. Annoyingly one of the suggestions was ‘spend a lot more’ but that is to be expected.

And if I was running this campaign where conversions generated sales, rather than just sign ups, I can very much see this as a viable option.

Conclusion

I can’t really say why the Google campaign was so much more successful than the others right now, I’m still drilling down into the numbers.

Obviously there could be some user error in some of the campaign setups, but I spent about 2 hours setting up each campaign with the different assets and copy. I read and followed their instructions/best practices.

I’m happy with the results, we got waitlist signups, but for a bootstrapped company it’s not viable as a long term launch strategy. I always knew this, but the campaign has given me a lot more confidence about our launch plan, and feature roadmap.

Worth it?

For that confidence boost/feeling of being on the right track it was worth it. I do think I could have probably spent a third of what I did and still got actionable results. But we live and learn, and nearly all of those platforms offer some sort of sign up credit so it hasn’t actually cost me the full amount out of the company bank account.

I’ll post a deeper dive with more numbers and stats etc when I’ve gone through all the data. But let me know if there’s something specific you’d like me to include in my post about the process or results.