Where hashtags still do work
The honest position is that hashtags matter much less than they did and more than the backlash suggests. Their role has shifted from distribution to classification: they are less a way to be discovered by people browsing a tag, and more a signal to the platform about what your content is about and therefore who to show it to. That is a smaller job than they once had, but it is not nothing, and it costs very little to do properly.
The mix matters more than the count
The most common mistake is using only enormous tags. A tag with fifty million posts offers effectively zero chance of visibility and tells the platform almost nothing specific about your content. A blend is more useful: a couple of broad tags for category signalling, several mid-size ones where you can plausibly compete, and a handful of genuinely niche tags that describe exactly what the post is.
- Two or three broad tags — category-level, for classification
- Four to six mid-size tags — the ones you could realistically rank in
- Three to five niche tags — specific enough that few others use them
- One branded tag, if the client has one worth building
How many, per network
On Instagram, somewhere between eight and fifteen well-chosen tags is the sensible working range — the era of thirty tags in a comment has passed. TikTok wants fewer: three to five, closely matched to the content, because the platform relies far more on what it can infer from the video itself. LinkedIn wants fewer still — three is plenty, and more looks like someone who has misjudged the room. X barely uses them at all now; one, at most, and only when it is a real conversation.
Research once, reuse for months
Hashtag research is worth doing properly and worth doing rarely. Spend an hour per client establishing sets for their main content pillars, and those sets will serve for months with minor adjustments. The mistake is treating it as a per-post activity: the marginal value of fresh tags on any individual post is tiny, and the cumulative time cost of researching every time is enormous.
Save the sets, do not retype them
Once you have good sets, the entire remaining cost should be an insertion rather than a recall exercise. Retyping tags from memory introduces drift — a branded tag spelled two different ways is a branded tag nobody can follow — and drift is exactly the failure that makes hashtag strategy look worthless in review. Whatever tool you use, keep the sets saved and insert them as a block, then trim for the individual post.
Consistency matters most across a team
The moment more than one person posts for an account, tagging drifts unless the sets are shared. This is the practical argument for keeping them in the publishing tool rather than in someone’s notes app: the set is defined once, everyone inserts the same one, and a new freelancer’s first post is tagged like your hundredth.
Know when to stop
Hashtags are a low-effort, low-return activity that becomes a negative-return activity the moment you spend real time on it. Build the sets, save them, insert them, adjust them quarterly. If you find yourself in a spreadsheet comparing tag volumes for an individual post, the strategy has stopped paying for itself and the time belongs in the content instead.