Meta Changed How Your Ads Are Measured — Here's What to Check in Your Account

If you run Meta ads for your small business and your numbers looked strange this week — either better or worse than expected — you’re not imagining it. Meta rolled out a significant attribution update in March that changes how clicks and video views are counted. Before you make any budget decisions based on what you’re seeing, read this first.

What Meta Changed

Two changes hit most ad accounts this month, and they’re worth understanding separately.

First: Link clicks and non-link clicks are now categorized separately. Previously, Meta’s click metrics lumped together a range of actions — clicking a link, clicking to expand an image, clicking on a profile, clicking “See More” on an ad. Now these are tracked in distinct categories. If you were using total clicks as a proxy for intent or traffic, your numbers have shifted — not because performance changed, but because the measurement bucket changed.

Second: The video engaged-view threshold dropped from 10 seconds to 5 seconds. Any campaign using video-view optimization or reporting on video views will now count a view at the 5-second mark instead of 10. Reported view counts will go up — often significantly — but this reflects a lower engagement bar, not a performance improvement.

According to Digital Trainee and confirmed across multiple ad management platforms, these changes affect how ROAS is calculated in accounts using certain attribution windows. Expect apparent performance shifts that are measurement artifacts, not real changes in what your ads are actually doing.

Why This Creates Confusion (and How to Read Your Numbers)

Here’s the practical problem: if your Meta reporting dashboard shows a ROAS jump, your first instinct might be to scale budget. If it shows a drop, you might want to pause campaigns. Either reaction could be wrong right now.

The right move is to look at your downstream business metrics — actual sales, form submissions, calls, or revenue tracked through your CRM — and compare those to the same period last month. If your real-world results are consistent with historical patterns, the change in Meta’s reported numbers is a measurement shift, not a performance shift.

A few specific things to check in your account:

Review your attribution window settings. Go to your Ads Manager columns and confirm which attribution window your campaigns are using (1-day click, 7-day click, 1-day view, etc.). The new categorization affects these windows differently, and you may want to normalize your reporting window to make accurate comparisons.

Pull a date comparison. Look at link clicks (not total clicks) for your campaigns March 1–15 versus March 15–25. If you see a dramatic difference, the categorization change is likely the cause, not campaign performance.

Don’t optimize against video views alone. With the threshold now at 5 seconds, video view volume is an even less reliable indicator of meaningful engagement. If video performance matters to you, focus on 25% and 50% video completion rates instead.

The Genuinely Good News: Natural Language Audience Targeting

Buried in the same update is something that should actually excite small business advertisers: Meta has launched a natural language AI audience targeting tool.

Instead of building audiences by manually stacking demographic filters, interest categories, and behavior signals, you can now describe your ideal customer in plain English. “Women who own small service businesses, interested in productivity and local marketing, age 35–55” — and Meta’s AI builds the targeting audience from that description.

This is a meaningful shift for small business owners who don’t have a media buyer or don’t want to spend hours in Ads Manager. Meta’s AI has been shown to outperform manual interest stacking in a growing number of campaign types, particularly for smaller budgets where the algorithm needs help finding the right audience quickly.

To try it: in your ad set audience settings, look for the “Describe your audience” or AI audience option (rollout is still in progress — it may appear as a beta feature in your account). It’s worth testing against your current best-performing manual audience.

What About the Instagram DM Privacy Change?

One more Meta item worth knowing: Fortune reported that Meta is reversing end-to-end encryption on Instagram DMs, with the change taking effect May 8. This means Instagram private messages will again be subject to Meta’s automated content scanning and AI moderation.

If you use Instagram DMs for customer service conversations, appointment booking, or sales follow-ups, your messages will be visible to Meta’s systems after May 8. This isn’t necessarily a reason to stop using DMs for business — but it’s worth knowing, and it’s a reason to consider moving sensitive customer conversations (pricing discussions, medical or financial information) to a more private channel.

Your Action List

1. Compare real results, not just Meta metrics — check actual leads, sales, or revenue against prior periods before drawing conclusions
2. Normalize your reporting columns — switch to link clicks (not total clicks) as your primary click metric going forward
3. Use video completion rates over view counts — 25% and 50% completions are more meaningful signals than raw view totals
4. Test the natural language audience builder if it’s available in your account
5. Note the May 8 Instagram DM change — adjust your customer communication approach if needed

The bottom line: don’t make budget cuts or scaling decisions based on this week’s numbers in isolation. Let the dust settle, establish new baselines with the updated measurement framework, and then optimize from there.