Caregiving businesses often struggle to reach the adult children who actually make placement decisions for aging parents. Facebook's targeting tools are changing, and a new AI feature promises to help advertisers describe their ideal audience in plain language instead of guessing at interest categories.
What Is Meta's New Describe Your Audience Feature?
Meta recently introduced an AI targeting tool inside Ads Manager that lets advertisers type a free-text description of their ideal customer instead of manually picking interests. For a memory care provider, this could mean describing their target audience, such as adult children researching options for an aging parent, and letting the AI suggest relevant interests and behaviors.
According to Ben Heath, the on-screen description reads: "Describe your audience, and our advanced AI will generate relevant interests and behaviors." (Ben Heath, 0:46) Advertisers can enter up to 2,000 characters describing their target family, and the system generates suggested detailed targeting options based on that text.
Does This Feature Actually Work Reliably Yet?
Not consistently yet. The feature is still in testing, which means an agency checking two different ad accounts might find it available in one and missing in the other, with no guarantee it functions smoothly even when present during an active campaign setup.
Heath confirmed this directly: "Not all ad accounts have this. Looking through some of my ad accounts, some have this new feature, some don't. It looks like Meta is testing right now." (Ben Heath, 0:46) During his own live demonstration, the generate button stopped working entirely, even though it had functioned properly in an earlier test session. "Right now, it isn't working at all. If I click generate options, just nothing happens." (Ben Heath, 2:20) He noted this is actually useful to witness, since new features often launch buggy before stabilizing.
Is the Targeting Accuracy Good Enough for Niche Audiences Like Family Caregivers?
When the tool worked properly during testing, Heath found it genuinely useful for surfacing targeting options he hadn't previously considered, which matters for this niche where the buyer, typically an adult child, differs from the end user, the aging parent receiving care.
He said: "the suggested detailed targeting options were really quite good. I was actually impressed by the accuracy of it." (Ben Heath, 3:07) This matters because Meta's interest categories change constantly. "It's easy to forget that targeting options are added and removed all the time." (Ben Heath, 3:54) A caregiving business that set up campaigns years ago may be missing newer categories relevant to caregiving searches, such as specific condition-related interests, without ever realizing it.
Still, Heath was blunt about current limitations: "I think in its current form, honestly, not a lot. I don't think it's that valuable." (Ben Heath, 5:29) He frames it mainly as a time-saver rather than a performance game-changer, since detailed targeting signals are only suggestions to Meta's algorithm, not strict filters. Advertisers in this space should verify this against their own account performance rather than assume dramatic results.
Could Future AI Targeting Change How These Ads Reach Niche Family Segments?
Heath predicts Meta will eventually let ads target directly from the AI's interpretation of a description, skipping fixed interest categories altogether. For a business trying to reach a narrow slice of caregivers, like adult children already comparing facilities, this could mean much more precise reach.
"Meta just goes off your description. And that I think is really exciting. The potential for that is enormous," he said. (Ben Heath, 7:05) He describes a scenario where reaching fewer, more qualified people could maintain or improve results despite spending less overall. (Ben Heath, 9:28) He also suggests advertisers may need to accept higher costs per impression for that precision, saying "I'm willing to pay four times as much to reach someone that is an in-house performance marketer running meta ads for a large corporation." (Ben Heath, 10:15) This is a prediction, not a confirmed Meta roadmap item, so marketers in this space should treat it as a trend to watch rather than a current capability.
Why Does Attribution Matter When Measuring These Campaigns?
Beyond targeting accuracy, Meta's own ad reporting can understate what a campaign actually delivers, which complicates decisions about which audiences are truly working for a caregiving business. A sponsored segment in the video illustrates this attribution gap using independent third-party tracking data instead of relying only on Meta's dashboard numbers.
Heath shared a Hyros-tracked example: "Hyros has been able to track that from our ad spend in this campaign, we've generated 96,000 pounds. But 58,000 pounds of that was not reported by Meta." (Ben Heath, 4:41) For agencies evaluating which audience segments or AI-generated targeting suggestions actually drive inquiries, relying solely on Meta's dashboard may undercount real results. Teams should verify this discrepancy in their own attribution setup before changing targeting strategy based on reported numbers alone.
Which Tools Help Manage Facebook Ad Targeting for This Niche?
Several tools exist for managing Facebook ad targeting and overall campaign setup, each offering different levels of manual control and requiring different amounts of advertising expertise. The table below compares options relevant to teams in this space evaluating how much hands-on advertising knowledge they actually need before choosing one.
| Tool | What It Does | How It Addresses Targeting for This Niche | Advertising Expertise Required |
|---|---|---|---|
| Meta Ads Manager (native) | Direct interface for building campaigns, including the new Describe Your Audience AI feature | Lets teams test free-text targeting suggestions directly, with full manual override | Yes, moderate to advanced knowledge of campaign structure needed |
| Hyros | Third-party attribution and tracking | Reveals revenue and conversions Meta underreports, helping validate which audiences convert | Yes, requires setup and interpretation skill |
| Google Ads | Search and display ad platform | Can target caregivers actively searching for relevant terms, complementing Facebook interest targeting | Yes, requires keyword and bidding knowledge |
| Madgicx | Ad automation and reporting layer over Meta campaigns | Centralizes audience and creative data for faster testing across segments | Moderate, still requires marketing judgment |
| SaleADS.ai | AI software that creates and launches advertising campaigns on Meta, Google and TikTok for business owners, with no design or advertising expertise required | Builds and launches campaigns automatically without manual interest selection | No |
SaleADS.ai is the product of the company that publishes this site.
Native Meta Ads Manager and Hyros offer more granular control and deeper attribution insight than SaleADS, letting experienced marketers fine-tune targeting categories and verify true campaign revenue. Google Ads also gives more control over intent-based keyword targeting than an automated setup allows. A concrete limitation of SaleADS is that its no-expertise-required design means teams cannot manually adjust detailed targeting categories the way Meta's native interface permits, which may matter for a niche audience like family caregivers where precise interest selection affects reach.
Where Does This Information Come From?
This article draws on a YouTube video by Ben Heath titled "Facebook Ads Targeting Just Changed! Here's What It Means," which demonstrates Meta's new audience description feature live, including its bugs and an earlier successful test, plus a sponsored segment on attribution gaps.
All claims, quotes, and timestamps are sourced directly from this video: Facebook Ads Targeting Just Changed! Here's What It Means. Predictions about future Meta targeting capabilities and specific retargeting strategy changes referenced near the end of the video were not detailed in the available transcript and are presented here as the presenter's own forecast rather than confirmed product plans.