Quick answer: if your Facebook or Instagram leads dried up around January or June 2026, do not blame the market yet. On 15 January 2026 Meta stopped serving every ad set that still contained a deprecated detailed-targeting interest, and it did so silently. The ad set does not pause, error or email you. It just stops spending. Check your Delivery column and audience warnings first, then rebuild affected ad sets on broad targeting, because Meta has made granular interest targeting a dead end.

First, the four-step check for a dead ad set

Before you rewrite your offer or fire your marketing, spend ten minutes in Ads Manager:

  1. Check the Delivery column. Open Ads Manager at ad set level. Anything marked “Not delivering”, or “Active” but with zero impressions over the last seven days, is your suspect list. An “Active” ad set that spends nothing is the signature of this change.
  2. Open the audience section of each suspect ad set. Meta flags deprecated interests with a warning that some detailed targeting options are no longer available. If your ad set was built before mid-2025 and used stacked interests like “home renovation” plus a suburb list, this is where the body is buried.
  3. Look at the spend graph for a flatline date. Pull a date range from December 2025 to now and look at daily spend per ad set. A hard flatline on or around 15 January 2026 is this deprecation. A softer wobble from June 2026 lines up with the later consolidation round.
  4. Check what Advantage has switched on. In any newer ad set with a conversion goal, look for “Advantage detailed targeting” and “Advantage+ audience” labels. If they are on, your interest selections are suggestions, not limits, which explains why results changed even though you changed nothing.

If step 1 and 3 both point at January, you have found what practitioners have been calling zombie ad sets: campaigns that look alive in the account but have not served an impression since the cutoff. One agency audit published in June 2026 reported finding 15% to 40% of client budgets sitting in ad sets that had not delivered since mid-January. That figure is practitioner-reported rather than Meta-confirmed, but the pattern matches what we see in tradie accounts: the account looks busy, the phone is quiet.

What Meta actually changed, and when

Here is the timeline, checked against Meta’s notices as republished by its partners and by industry press on 7 August 2026:

  • From 23 June 2025: Meta began consolidating thousands of granular detailed-targeting interests into broader buckets. Niche interests for specific sports, car models, foods and music genres were merged or retired, and warning banners started appearing in Ads Manager.
  • 15 December 2025: deprecated interests became unselectable. You could no longer add them to new ad sets or saved audiences.
  • 15 January 2026: the hard cutoff. Ad sets still containing a deprecated interest stopped delivering entirely. No pause, no error state, just zero spend.
  • Detailed Targeting Exclusions: gone. Meta started removing interest-based exclusions from new campaigns back in mid-2024 and wound them down in stages; practitioner guides report the last of them were phased out by 31 January 2026. Either way, in 2026 you cannot exclude people by interest at all. Custom audience exclusions and account-level audience controls are what remains.
  • June 2026: practitioner coverage reports a further consolidation round, merging more of the remaining interest categories. We could not find a standalone Meta announcement for this wave, so treat the date as reported rather than official, but plenty of advertisers saw ad set warnings and delivery changes that month.
  • Advantage detailed targeting is now the default, and for conversion goals it is mandatory. Meta’s own documentation, mirrored in partner help centres, confirms that when your performance goal is conversion-based, Advantage detailed targeting is applied automatically and cannot be turned off. Your interest picks are treated as a starting point that Meta is free to expand past.

None of this was hidden, exactly. Meta showed warnings in Ads Manager through late 2025. But if you are a plumber running your own ads between jobs, “I will deal with that banner later” is how an account quietly enters February with a third of its budget doing nothing.

How to fix it: rebuild broad, and let the creative do the targeting

Do not try to patch a deprecated ad set by swapping interests one for one. The replacement buckets are so broad that you inherit all the downside of broad targeting with none of the upside of a clean signal. Rebuild instead:

  1. Duplicate the campaign and start the ad set fresh. Set location to your real service area, a sensible age range, and nothing else. This is the direction Meta is forcing, and its delivery system genuinely performs better with room to move, provided the next two steps are done properly.
  2. Make the ad do the targeting. With broad audiences, your creative is the filter. “Hot water system dead? Darwin plumber, same-day replacement” self-selects the right people better than any interest bucket ever did. Say the trade, the place and the problem in the first two seconds.
  3. Feed it real conversion signal. Broad targeting only works when Meta knows what a good lead looks like, which means your form or call events must actually fire. If your tracking is half-wired, broad targeting will happily find you cheap clicks from the wrong suburbs. This is the same discipline that decides whether your website turns visits into leads.

When Advantage+ is fine for a tradie: you serve a wide metro area, you can handle whatever job types come in, and you have steady conversion volume for the system to learn from. Let it run.

When it wastes budget: you only want one high-value job type, your service radius is tight, or you get a handful of leads a month. Advantage will cheerfully spend on look-alike work you do not want, and with exclusions gone you cannot fence it off by interest. In that case keep the campaign narrow with geography and creative, or put the budget into search instead, where intent does the filtering. Our Google Ads vs Meta ads comparison for tradies covers when each platform earns its keep.

The honest bit

DIY Meta ads keep getting harder, and that is by design. Meta is deprecating the manual levers because its automation performs better on average across millions of advertisers, and “on average” is doing a lot of work in that sentence for a sole trader with a $30-a-day budget. Full transparency: running Meta ads for tradies is part of our Lead System, so we benefit if you decide this is no longer worth your evenings. But the free advice stands on its own: run the four-step check today, kill the zombie ad sets, rebuild broad with creative that names your trade and your town, and make sure conversions are tracked before you spend another dollar.

Timeline details were checked on 7 August 2026 against Meta partner help documentation (Brandwatch, Sprinklr) and industry coverage from Social Media Today and MediaPost. Where a detail rests on practitioner reports rather than a Meta announcement, we have said so in the text.