Brand hashtag strategy: does it still work and how to update it
Hashtags did not die — they changed their function
Hashtags did not disappear and did not lose meaning — but they stopped being a tool for reach and became a tool for clustering the first audience. On TikTok this is visible literally by seconds: after posting there is an «indexing window» from 15 minutes to several hours, during which a specific hashtag helps the algorithm quickly determine to which narrow group of viewers to show the video first. After that, the hashtag loses its influence completely — the behavior determines the outcome: how many watched till the end, rewatched, saved, or left because of the sound or caption in the comments.
Here is a direct answer to the question 'do hashtags still work': yes, but only within a narrow window — at the start of a post's life and to get into the right first viewer selection. After that, the algorithm looks at people, not tags. Instagram has technically confirmed this: in December 2025 the platform limited the number of hashtags to five per post, TikTok is testing the same limit, and the 'Recent' tab in Instagram's hashtag search has completely disappeared — only 'Top' remains, where the most engaging posts appear, not the newest ones with the relevant tag.
2018 vs 2026: why multiplying tags no longer works
Previously, the logic was simple: the more relevant tags there were — the more 'doors' through which viewers could come. Now, every extra tag signals the algorithm that the author is trying to manipulate distribution rather than create content for a specific audience. HubSpot provides a figure that explains why platforms don't need hashtags at all: neural networks recognize the topic of content without a single tag, analyzing the signature, OCR text in videos, and speech via speech-to-text.
| Parameter | Strategy for 2018 year | Strategy for 2026 year |
|---|---|---|
| Number of tags | 20-30 per post | 3-5, platform limit |
| Logic of selection | maximum reach, broad tags | precise targeting of a narrow cluster |
| Role in the algorithm | main signal for distribution | initial signal, behavior determines the rest |
| What the platform looks at | tag matching | re-watches, saves, speech, text in video |
| Effect of tag overuse | was considered safe | minus 15-33% reach and engagement |
Case: how a beauty salon increased reach by reducing tags from 30 to 7
Let's examine the same type of account as at the beginning of the article, but now about services, not coffee. The salon account was publishing a block of 30 tags under each post: #salon #manicure #beauty #style #Moscow and so on, copied from the first guide that came by. Reach remained stable, but support metrics — retention, transitions to direct — were almost zero.
We changed our approach according to the formula 3-3-3-1: three geolocation tags (district, metro, city), three narrow niches with a volume of 10-100 thousand publications (for example, a specific nail art technique), three medium tags with a volume of 100 thousand — 1 million and one wide tag with a volume of 100 thousand or more publications. This formula shows an increase compared to 30 identical tags. In our example, the results were similar: saves increased significantly — this is the metric that the platform reads as a signal of content value, not as a random click. The reach from the tags themselves grew moderately, but what really brought clients — saves and profile visits — increased.
How to distinguish a working hashtag from a dead one
You should evaluate not the mere fact of a hashtag's presence, but its contribution to specific metrics. In practice, we look at four indicators:
- Share of views from hashtags/research — if in the breakdown of traffic sources (available in Instagram and TikTok) hashtags provide less than 5-8% of views, the tag is ballast, not a tool.
- Saves and sends to direct — they grow when the tag brought relevant audience, not random people.
- Speed of initial view accumulation — on TikTok this is literally the first hours after posting: if the video doesn't reach critical mass within this time, the specific tag has already missed its indexing window.
- Dynamics over time — a tag that «passed» half a year ago but has stopped bringing views now most likely exhausted its niche audience or the volume of its publications grew so much that the content is drowning in the stream.
A dead tag is not the one that is 'not trendy,' but the one that does not move any of these metrics for three or four publications in a row. It needs to be changed, not endured out of habit.
Social SEO without a blog: how it works for a cafe, salon, and master
The main gap in advice about 'switching to Social SEO' is that almost nobody explains it for businesses that physically do not produce search content — a cafe, salon, nail master, dentist. Such accounts have no articles, no long videos with useful content — only photos of dishes, work process, and reviews. But Social SEO for them is built not on content-instruction, but on text around regular publications:
- In the caption, directly name what clients are looking for locally: not 'atmosphere of the evening,' but 'dinner for two in the center, window table' — with the district and metro name.
- Articulate key words out loud in videos — TikTok and Reels recognize speech via speech-to-text, and this works as alternative text, but for oral delivery: the master mentions techniques, services, materials out loud.
- Fill in alt-text for photos on Instagram — most accounts don’t use this function at all, yet it directly participates in the platform’s search results.
- Use geotags and local names in post and story texts, not just in hashtags — this is the same signal, but more resistant to algorithm changes.
Geolocation hashtags essentially become secondary to mentioning the district directly in the text: they confirm what has already been said in words, rather than replacing them.
Different strategy for the platform and type of business — and how to update it without a complete overhaul every year
The mistake of most guides — a universal set of tips for all platforms at once. In practice, the logic differs significantly:
| Platform | What to do with hashtags | Where to shift the focus |
|---|---|---|
| 5 tags, the formula 3-3-3-1, betting on 'Top' | alt-text, caption with keywords, saves | |
| TikTok | 3-5 tags only in the first hours of the video’s life | speech in frame, OCR text on video, re-watches |
| YouTube | are practically useless — the platform ignores more than 60 tags per channel | name, description, subtitles |
| X (Twitter) | removed from ad formats | tweet text, links, reactions |
| priority reduced by algorithm | post text, professional terms in first lines |
There is no universal scheme that will survive unchanged through one year and through two — platforms will continue to reduce the role of tags in favor of behavioral signals. But there is a working principle of adaptation that does not depend on a specific algorithm: treat the hashtag not as a magical amplifier, but as precise targeting for the first group of viewers, and once a quarter check its contribution against real metrics — views from the 'research' section, saves, and speed of initial impressions. If the platform cuts the hashtag limit — this is not a loss of the tool, but a signal that it's time to invest in text, speech, and alt-descriptions, which are harder to become outdated because they are read by people, not parsed for string matches.
Hashtags did not die — they changed their function
Hashtags did not disappear and did not lose meaning — but they stopped being a tool for reach and became a tool for clustering the first audience. On TikTok this is visible literally by seconds: after posting there is an «indexing window» from 15 minutes to several hours, during which a specific hashtag helps the algorithm quickly determine to which narrow group of viewers to show the video first. After that, the hashtag loses its influence completely — the behavior determines the outcome: how many watched till the end, rewatched, saved, or left because of the sound or caption in the comments.
Here is a direct answer to the question 'do hashtags still work': yes, but only within a narrow window — at the start of a post's life and to get into the right first viewer selection. After that, the algorithm looks at people, not tags. Instagram has technically confirmed this: in December 2025 the platform limited the number of hashtags to five per post, TikTok is testing the same limit, and the 'Recent' tab in Instagram's hashtag search has completely disappeared — only 'Top' remains, where the most engaging posts appear, not the newest ones with the relevant tag.
2018 vs 2026: why multiplying tags no longer works
Previously, the logic was simple: the more relevant tags there were — the more 'doors' through which viewers could come. Now, every extra tag signals the algorithm that the author is trying to manipulate distribution rather than create content for a specific audience. HubSpot provides a figure that explains why platforms don't need hashtags at all: neural networks recognize the topic of content without a single tag, analyzing the signature, OCR text in videos, and speech via speech-to-text.
| Parameter | Strategy for 2018 year | Strategy for 2026 year |
|---|---|---|
| Number of tags | 20-30 per post | 3-5, platform limit |
| Logic of selection | maximum reach, broad tags | precise targeting of a narrow cluster |
| Role in the algorithm | main signal for distribution | initial signal, behavior determines the rest |
| What the platform looks at | tag matching | re-watches, saves, speech, text in video |
| Effect of tag overuse | was considered safe | minus 15-33% reach and engagement |
Case: how a beauty salon increased reach by reducing tags from 30 to 7
Let's examine the same type of account as at the beginning of the article, but now about services, not coffee. The salon account was publishing a block of 30 tags under each post: #salon #manicure #beauty #style #Moscow and so on, copied from the first guide that came by. Reach remained stable, but support metrics — retention, transitions to direct — were almost zero.
We changed our approach according to the formula 3-3-3-1: three geolocation tags (district, metro, city), three narrow niches with a volume of 10-100 thousand publications (for example, a specific nail art technique), three medium tags with a volume of 100 thousand — 1 million and one wide tag with a volume of 100 thousand or more publications. This formula shows an increase compared to 30 identical tags. In our example, the results were similar: saves increased significantly — this is the metric that the platform reads as a signal of content value, not as a random click. The reach from the tags themselves grew moderately, but what really brought clients — saves and profile visits — increased.
How to distinguish a working hashtag from a dead one
You should evaluate not the mere fact of a hashtag's presence, but its contribution to specific metrics. In practice, we look at four indicators:
- Share of views from hashtags/research — if in the breakdown of traffic sources (available in Instagram and TikTok) hashtags provide less than 5-8% of views, the tag is ballast, not a tool.
- Saves and sends to direct — they grow when the tag brought relevant audience, not random people.
- Speed of initial view accumulation — on TikTok this is literally the first hours after posting: if the video doesn't reach critical mass within this time, the specific tag has already missed its indexing window.
- Dynamics over time — a tag that «passed» half a year ago but has stopped bringing views now most likely exhausted its niche audience or the volume of its publications grew so much that the content is drowning in the stream.
A dead tag is not the one that is 'not trendy,' but the one that does not move any of these metrics for three or four publications in a row. It needs to be changed, not endured out of habit.
Social SEO without a blog: how it works for a cafe, salon, and master
The main gap in advice about 'switching to Social SEO' is that almost nobody explains it for businesses that physically do not produce search content — a cafe, salon, nail master, dentist. Such accounts have no articles, no long videos with useful content — only photos of dishes, work process, and reviews. But Social SEO for them is built not on content-instruction, but on text around regular publications:
- In the caption, directly name what clients are looking for locally: not 'atmosphere of the evening,' but 'dinner for two in the center, window table' — with the district and metro name.
- Articulate key words out loud in videos — TikTok and Reels recognize speech via speech-to-text, and this works as alternative text, but for oral delivery: the master mentions techniques, services, materials out loud.
- Fill in alt-text for photos on Instagram — most accounts don’t use this function at all, yet it directly participates in the platform’s search results.
- Use geotags and local names in post and story texts, not just in hashtags — this is the same signal, but more resistant to algorithm changes.
Geolocation hashtags essentially become secondary to mentioning the district directly in the text: they confirm what has already been said in words, rather than replacing them.
Different strategy for the platform and type of business — and how to update it without a complete overhaul every year
The mistake of most guides — a universal set of tips for all platforms at once. In practice, the logic differs significantly:
| Platform | What to do with hashtags | Where to shift the focus |
|---|---|---|
| 5 tags, the formula 3-3-3-1, betting on 'Top' | alt-text, caption with keywords, saves | |
| TikTok | 3-5 tags only in the first hours of the video’s life | speech in frame, OCR text on video, re-watches |
| YouTube | are practically useless — the platform ignores more than 60 tags per channel | name, description, subtitles |
| X (Twitter) | removed from ad formats | tweet text, links, reactions |
| priority reduced by algorithm | post text, professional terms in first lines |
There is no universal scheme that will survive unchanged through one year and through two — platforms will continue to reduce the role of tags in favor of behavioral signals. But there is a working principle of adaptation that does not depend on a specific algorithm: treat the hashtag not as a magical amplifier, but as precise targeting for the first group of viewers, and once a quarter check its contribution against real metrics — views from the 'research' section, saves, and speed of initial impressions. If the platform cuts the hashtag limit — this is not a loss of the tool, but a signal that it's time to invest in text, speech, and alt-descriptions, which are harder to become outdated because they are read by people, not parsed for string matches.
Recent articles
MegaFon, BMW, VTB and more 240+ companies. See how it looks in the showcase reel:
Our branding agency is approached to order:
Branding agency services
