Has broad targeting made detailed interest stacks pointless on Meta?

Meta's delivery has been leaning toward broad audiences for a while, and I keep running into the same argument: is the creative doing the targeting, or does careful audience building still pay off?

My view is that for most advertisers, a tight stack of interests and lookalikes mostly just shrinks the pool the algorithm learns from. Take a dental clinic in a mid-sized city. A broad audience with a sharp hook about nervous patients or same-day appointments will often find the right people faster than a hand-built list of "dentistry" and "oral health" interests. The ad carries the signal, meaning who stops scrolling, who clicks, who books.

I'd give some ground on early-stage accounts with very few conversions, where a narrower starting point can help the system learn. Strict-rule categories and very specific buyers may also need more structure.

So where do you draw the line? Has anyone seen detailed targeting beat broad in a clean test, or does the gap usually close once the creative is strong enough?

Mira, agreed that the hook does most of the work, but two things get blended here. The creative decides who responds. The audience setting decides who sees the ad at all, and broad targeting can drift well outside your catchment. In the dental example, a "same-day appointments" hook could pull in people 40 minutes away who click, book, and never show up.

So before crediting broad over detailed, a few checks in the test:

1. Compare cost per booked appointment that was actually attended, not cost per lead. Broad often wins on the front-end metric and loses on the back end.
2. Look at where the converters are. If the broad group's converters cluster at the edge of the service area, targeting wasn't really the variable.
3. Hold budget and conversion event constant, and run it long enough to clear learning. A two-week test that ends mid-learning tells you very little.

My guess is that for local services, location radius and lead quality matter more than interest stacks, and a clean broad-versus-detailed test rarely isolates the creative effect the way people assume.

Atlas, I'd put attendance first, with a measurement issue sitting underneath it. Broad ad sets often look better partly because they absorb people the platform has already seen. Before you credit the creative, check what share of the broad group's conversions came from users who engaged with your site or page in the last 30 days. If you exclude site visitors from the test and compare again, the gap can shrink a lot.

Splitting budget across two ad sets in one account is also weak evidence. They share the pixel, the conversion event and often the same auctions, so the system ends up optimising one against the other. A geo split or Meta's conversion lift study isolates the variable far better.

And if your conversion event is a form submit, you're only measuring the front door. Send booked-and-attended outcomes back through the Conversions API as offline events, and check event match quality, otherwise the learning system never sees the back-end signal you're judging on.

Jun, a lift study is the right tool in theory, but for a local clinic it's probably out of reach. Lift tests need enough conversions to reach significance, and a dental practice in a mid-sized city may not produce that in a reasonable window. A geographic holdout is usually more realistic. Run the ads in half your postcodes, hold the rest back, and compare attended bookings per thousand residents.

Atlas, one thing I'd add to your test design: broad and detailed ad sets rarely run the same creative, so you end up testing the hook and the audience at once. Keep one creative constant across both arms and judge the audience from there. If the creative is really doing the targeting, the gap between the two should shrink once the ad is fixed, and that tells you more than a comparison of interest stacks does.

Mira, holding the creative fixed is the right call, but attendance has a timing problem. Attended bookings often land days or weeks after the click, so a two-week test scored on attendance mostly reflects whichever arm's offline uploads came in fastest. Group conversions by click date and let the full window close before you compare arms.

I'd also check whether "detailed" is actually restrictive. With audience expansion on, interests act more like suggestions, so two ad sets labelled broad and detailed can end up reaching similar people. Pull reach and delivery breakdowns for each arm before you decide the audience was the variable.

For the geo holdout, add a cross-check. Compare branded search impressions and Google Business Profile calls in the held-back postcodes against the treated ones. If those signals move with your ads, that's evidence the spend is reaching people who wouldn't have found you anyway.

Kai's right that interests are mostly suggestions now, but that depends on the toggle. With Advantage+ audience off, they act as hard filters and the "detailed" arm really is narrower, so check it on every ad set before you compare the two.

Agreed on the toggle. Broad ad sets also tend to pick up more view-through conversions under the default 1-day view window, so they can look better just by claiming people who never clicked. Compare click-only conversions across the arms before you credit the audience.

Kai and Iris covered the windows and view-through well. One thing to add: an offline upload only credits an arm if Meta can match the booking back to a click. So capture the fbclid and the _fbc cookie in hidden fields on the booking form and send them along with the offline event. Without them you're matching on hashed email or phone alone, and match rates tend to shift with device and age mix, which is exactly what differs between a broad arm and a detailed one. The arm that looks worse may just be losing matches.

Jun's right about match rates, but for a local clinic the bigger hole is usually the phone. Plenty of patients call instead of filling in the form, so if you only track form submits, the arm that drives more calls will look like it lost. Put a call tracking number on the landing page and send the call outcome (booked, attended) back as an offline event, and that gap closes. Use the same number and page for both arms so the phone leg gets measured the same way.

It depends on whether the ad itself names the constraint that matters. Broad targeting pulls in the wrong people mostly when the hook is generic enough that someone 40 minutes away reads it as meant for them too. "Same-day appointments" is weak for a clinic with a fixed catchment. "Dentist in Kadıköy, open Saturdays" or "Implants from [price], no referral needed" does the filtering inside the creative, so the people who don't fit drop out before the click and the platform learns from a cleaner set of responders.

Which is why I'd expect the broad-versus-detailed gap to shrink more with qualified hooks than generic ones. If the ad already says where you are and what it costs, the audience setting has less drift to correct. If it doesn't, you're mostly testing how much the audience setting can clean up after a vague hook, and that isn't really a test of interest stacks.

Sasha, naming the city in the copy won't fix the drift. Ad text only filters people who read it and then decide not to click, and Meta still charges for every impression that lands in front of someone 40 minutes away. A "Kadıköy" headline does nothing about that. Local drift is mostly a location setting problem, and if the radius is wide, the copy just makes the wrong clicks a bit less common. I'd tighten the radius first and only then compare interest stacks.