What we learned running Meta ads across five different industries this year

This year we ran active Meta ad campaigns across retail, healthcare, hospitality, professional services, and a subscription product, and the single biggest lesson we took away wasn't some clever platform trick. It was how differently the exact same underlying platform genuinely performs depending on the actual purchase decision sitting behind it, and how often that fairly obvious point gets ignored in favor of generic, one-size-fits-all "best practices" advice.

For our outdoor retail client specifically, creative fatigue set in fast, often within roughly ten days of a new ad set first launching, as the same audience saw the same creative repeatedly. What actually worked was maintaining a steady weekly cadence of genuinely new creative variations, rather than investing in fewer, more polished, longer-running ads. Volume and speed of iteration consistently beat individual ad quality on its own in this specific category.

A dental practice client asking for an immediate booking commitment in the very first ad a complete stranger ever sees converts noticeably worse than a proper sequence: an educational post first, followed by a genuine patient testimonial, only then finally a booking offer targeted specifically at people who'd already engaged meaningfully with the earlier content. Healthcare decisions simply carry more real hesitation and consideration than a retail purchase does, and the campaign structure needs to genuinely reflect that difference rather than reusing a retail playbook.

For our boutique hospitality client, heavily polished, professionally agency-shot creative actually underperformed against genuine, slightly rough guest photos and short, unscripted guest video content. People trust hospitality content that visibly looks like it came from an actual real guest considerably more than content that looks like a produced brochure, which ran directly against our initial creative instinct going into that specific account.

Every single account that meaningfully outperformed our initial expectations did so because the campaign structure genuinely matched how people actually make that specific category of purchase decision in real life, fast and emotional for retail, slow and trust-based for healthcare, social-proof-driven for hospitality. There's no single universal Meta ads playbook that works everywhere. There's effectively a distinct playbook per category of underlying decision, and figuring out which one genuinely applies to a given client is most of the actual strategy work involved, more than any specific tactic.

Despite all the differences described above, one pattern held steady across all five accounts without exception: creative that named a specific, concrete detail (a real price, a real location, a real named benefit) consistently outperformed creative built around a vague, general claim, regardless of the industry or the specific audience being targeted.

That single finding has become a standing rule in our own creative review process now, applied before any campaign launches, specifically because it's the one lesson from this whole year of testing that generalized cleanly across every category we touched.

Our professional services client, a mid-size accounting firm, taught us that Meta ads for high-trust, high-consideration services need a genuinely longer runway before judging performance than almost any other category we ran this year. Early campaign data looked mediocre against a direct-response benchmark, cost per lead considerably higher than retail, click-through rate lower than hospitality, and it would have been easy to pull the plug within the first month based on those numbers alone.

Staying the course past that initial period, and tracking actual consultation bookings rather than raw click metrics, showed a genuinely strong return once the full, longer sales cycle played out. We now set explicit, longer evaluation windows before drawing conclusions on any professional services account, specifically because of what this campaign taught us about judging this category too early.

Running ads for a subscription product surfaced a lesson none of the other four verticals fully prepared us for: optimizing purely for initial signup volume, without any regard for which specific audiences actually stuck around past the first billing cycle, produced a growing subscriber base that looked healthy on a dashboard and was quietly churning at a rate that made the whole acquisition effort unprofitable within a few months.

We started feeding 60-day retention data back into audience refinement for that account, deliberately deprioritizing audiences that signed up cheaply and churned fast even when their initial cost-per-acquisition number looked excellent. That single change improved the account's actual long-term profitability more than any creative or targeting tweak we made across the entire year of work on it.

Before this year of running genuinely varied verticals side by side, our new-client pitches leaned more heavily on generic platform expertise, a fairly standard "we know Meta ads" positioning that most agencies use interchangeably regardless of industry. We've since rebuilt our discovery process to ask, upfront and explicitly, what the client's real buying decision actually looks like emotionally and practically, before proposing any specific campaign structure at all.

That single change in how we start a new engagement has noticeably shortened the time it takes a new account to find its actual winning structure, because we're no longer starting from a generic template and iterating our way toward the category-specific approach that this year's direct experience across five verticals taught us to expect from the very beginning.

Even as the winning creative and structure differed sharply by vertical, the actual process for finding that winning approach stayed identical across every account: a tight weekly review cadence, a clear single primary metric agreed with the client before launch, and a genuine willingness to kill an underperforming approach quickly rather than let sunk cost keep it running past the point the data had already made clear.

That consistency in process, layered underneath very different category-specific findings, is probably the more durable lesson from this year of work than any single vertical-specific insight, since the specific tactics will keep shifting as platforms and audiences change, and the underlying discipline of testing well is what actually keeps working regardless.