Meta Ads for SaaS flopped at €6k spend - any advice?

nightowl

I've been running Meta ads for a SaaS product - an AI SEO tool aimed at small businesses and Shopify stores. After spending around six thousand euros over three months, the campaigns just hit a wall. At first we optimised for 'Start Trial' - got a decent number of trial starts, but many of those cards bounced when the subscription hit. Then a Meta marketing pro advised switching to 'Subscribe/Purchase'. That worked for a short burst, then performance fell off a cliff - 'Low Results' everywhere. Switched back to trial optimisation, and now that's barely firing either. Latest week: about €450 spend, two trial starts, one paying customer. All told, Meta has brought in maybe eight customers, most stuck around for a month or two. Not dead, but clearly broken.

Meta support now suggests something different: run a Traffic campaign (€40/day) with ad sets for landing page views and link clicks, plus an Awareness campaign (€25/day) optimised for reach. Don't touch anything for two weeks, then improve creatives. Their reasoning: cheaper traffic feeds the pixel more data, which eventually leads to better conversions.

To me this feels like the classic 'buy cheap traffic to warm the pixel' advice that works for impulse-buy ecom, not for a SaaS product that needs 3-5 months to show value. When your audience is price-sensitive small businesses, cheap clicks bring people who aren't ready to wait for an ROI. The macro picture is tough - inflation is squeezing those same businesses. Pushing them through a bottom-funnel offer with low-intent traffic seems backwards.

I'd rather fix the funnel tracking, improve the trial-to-paid conversion, and keep optimising for Purchase with better creatives. But with a daily budget of only €55-65, splitting it across traffic and awareness feels like spreading butter too thin.

Would any of you test that Meta support setup, or would you double down on purchase optimisations? What signals would you look at before deciding?

localpack

I’d be cautious shifting fully to Traffic or Awareness for SaaS. You’ll get cheaper clicks but weaker intent. Better to fix funnel tracking, improve trial-to-paid conversion, and keep optimizing for Purchase with stronger creatives.

analyticsjunkie

Oh, the classic "let's throw Traffic at it and pray" move. Cheap traffic doesn't teach Meta to find buyers - it teaches Meta to find people with idle thumbs. For a SaaS with a paid trial and delayed payoff, that's basically lighting budget on fire while looking hopeful.

The fact Purchase optimisation worked at all before tanking is the real clue here. Sounds less like "Meta can't find buyers" and more like a signal quality issue or a scaling wall. Did anything change in the funnel right before performance nosedived? Creative fatigue? Trial flow friction? Because that drop-off pattern screams platform learning reset, not a sudden shift in audience intent.

trafficcop

Have you checked whether the real issue is actually signal quality and optimization mismatch rather than Meta simply “not working” for your SaaS?

Honestly I do not think moving heavily toward Traffic and Awareness optimization is the right long term direction for a product like yours. Meta is extremely literal. If you optimize for link clicks, landing page views, or reach, Meta will absolutely find cheap traffic, but that traffic is usually weak for a SaaS subscription product that requires trust, onboarding, delayed value realization, and long term retention before users fully understand the benefit.

Worked with a B2B AI automation SaaS where Meta initially optimized for cheap free trials and scaled into low quality users. CTR and signup volume looked healthy, but retention collapsed because Meta learned toward curious clickers instead of serious business owners. When they switched directly to Purchase optimization, results briefly improved but volume became unstable due to insufficient conversion density. We rebuilt the account around a deeper onboarding activation event while simplifying the campaign structure and tightening the signal quality. Within around 6 weeks CAC dropped from roughly €420 to €170, trial to paid conversion improved from around 6% to 18%, and retention increased significantly because Meta finally learned from qualified users instead of shallow trial behavior.

are the customers who stay subscribed actively using the product and getting measurable SEO improvements, or is onboarding and delayed value realization still causing churn afterward?

cpaoptimizer

Meta support usually does not give very advanced strategic advice. Their main goal is often to increase platform spend, so I personally would not blindly switch from Sales optimization to Traffic + Awareness for a SaaS subscription product.

Cheap traffic does not automatically mean good users. In many cases Traffic campaigns just bring low-intent visitors and vanity metrics while burning budget.

I think the bigger issue is probably:

ad creatives/hooks
ICP clarity
landing page conversion flow
trust and positioning
onboarding experience

Especially for SaaS, your creative needs to immediately explain:

who this is for
what painful SEO problem it solves
why it’s different from agencies/tools
and what realistic outcome users can expect

Also your landing page needs strong action flow because asking users to trust an AI SEO tool with a paid trial is already a harder conversion compared to ecommerce.

Since you already got some paying users that stayed 1–2 months, Meta is clearly not “dead.” That means there is demand somewhere. I’d focus more on improving creatives, hooks, onboarding, and ICP messaging instead of optimizing for cheap traffic signals.

Personally I would still stay closer to Sales/Subscribe optimization rather than Link Clicks or Reach campaigns for a SaaS product, because low-quality traffic usually won’t bring good signup signals anyway.

datanerd

Ignore that advice from Meta support - Traffic and Awareness campaigns will do nothing for SaaS sign-ups. Stick with Purchase or Trial conversion optimisation and sort the card failure issue first by testing a different payment flow. That's the real bottleneck, not the optimisation objective.

nightowl

with that daily budget, I'd treat it like a precision instrument rather than a broad brush. Focus on one conversion event, let the algorithm fully learn, then scale slowly. what's your current frequency looking like?