If your Meta ads have been quietly underperforming and you can’t work out why, the length of your ad copy might be the problem, not the offer. Most builders write one paragraph of ad copy, run it everywhere, and assume that’s enough. In 2026, that habit is costing leads.
Here’s what’s actually happening behind the scenes on Facebook and Instagram, and what to do about it.
Why Meta ads work differently now
Meta rolled out a new ad delivery system called Andromeda. It changed how ads get shown to people. Instead of you picking an audience and hoping the right people see your ad, Meta now tests many versions of your creative and routes each version to whoever it predicts will respond to it.
Practically, this means the old approach of writing one paragraph of copy and running it as-is gives Meta almost nothing to work with. The builders getting the best results in 2026 are the ones feeding the system genuine variety, not the ones writing the “perfect” single ad.
The length mistake most builders make
A recent study of close to 44 million Meta ads by AdSpyder found something worth paying attention to. Ads under 50 characters (think one punchy line) performed best over time. Ads between 250 and 500 characters (a genuine, detailed story) came in second. The worst-performing length? Anywhere in between, the “medium paragraph” that most builders write without thinking twice about it.
Term: The messy middle. This is copy that’s too long to work as a quick hook and too short to build a real case for why someone should enquire. It’s also where most builder ad copy currently sits.
What to do instead
- Write one genuinely short line for every ad. Only around 125 characters show on mobile before someone has to tap “See More,” and almost nobody taps. Your first sentence needs to work as the whole ad, on its own, every time.
- Save the long version for your best-performing campaigns. If a campaign isn’t generating enough enquiries in a given month, a second copy variant won’t fix that. It just splits an already thin result in half. Put the long-form effort into the campaign that’s already working.
- Test your headline, not just your body copy. Meta lets you load up to five headline options per ad. A short, benefit-led headline (“Move-In Ready This Spring”) consistently beats a feature-dump headline (“Award-Winning Four-Bedroom Design With Premium Inclusions”). Most builders never touch this.
- Fix your tracking before you add more creative. None of the above matters if Meta can’t accurately see which enquiries came from which ad. Clean conversion tracking (properly set up Conversions API, not just the pixel) is what lets the system learn fast. A builder running average ad copy with clean tracking will usually beat a builder running great ad copy with messy tracking.
A quick way to check your own account
Pull your last quarter of ad performance and sort by enquiries per month, not total spend. Any campaign generating fewer than roughly 3 to 5 enquiries a month doesn’t have enough volume to fairly test two copy lengths against each other. Put one strong, hook-led version on those campaigns instead, and save the short-versus-long testing for the campaign carrying your real volume.
In our work with builders, this single change (stop testing thin campaigns, concentrate testing on the campaign with real volume) is usually the fastest way to make an existing ad budget work harder without spending an extra dollar.
The bottom line
You don’t need to choose between short and long ad copy. You need to know where each one earns its place. Every ad gets a short, standalone hook. Long-form copy is a strategic choice for your highest-volume campaign, not a rule applied everywhere. And headline testing is free performance most builders are leaving on the table.
Want a second set of eyes on where your own Meta account stands? Get in touch with the Sea Salt Marketing team or subscribe to our monthly newsletter for more builder-specific marketing breakdowns like this one.
Source: AdSpyder, “Ad Copy Length Best Practices: Data From 211M+ Ads”
