Hashtag Science — Hashtag Science Methodology
Specialized knowledge used by the hashtag-analyst agent when designing hashtag strategies.
Why Hashtag "Science"
Hashtags are not applied by intuition. They involve calculating discoverability, competition level, and targeting accuracy to design optimal combinations.
Hashtag Pyramid Strategy
/\
/ 1 \ <- Large (1M+ posts)
/------\ 2-3: Discovered with luck
/ 2 \ <- Medium (100K-1M)
/----------\ 5-7: Core battlefield
/ 3 \ <- Small (10K-100K)
/--------------\ 5-8: Targeted reach
/ 4 \ <- Niche (under 10K)
/------------------\ 3-5: Community entry
Role by Tier
| Tier |
Post Count |
Role |
Strategy |
| Large |
1M+ |
Category indicator |
#marketing, #startup — max 3 |
| Medium |
100K-1M |
Core reach |
Target's actual search terms — 5-7 |
| Small |
10K-100K |
Competitive advantage |
Specific topic terms — 5-8 |
| Niche |
Under 10K |
Community |
Industry-specific terminology — 3-5 |
| Brand |
Proprietary |
Tracking/Campaign |
#BrandName, #CampaignName — 1-2 |
Platform-Specific Hashtag Strategies
Instagram
| Item |
Recommendation |
| Count |
15-25 (use 20-25 of the max 30) |
| Placement |
Bottom of caption or first comment |
| Composition |
Follow pyramid ratios |
| Prohibited |
Banned hashtags (shadowban risk), unrelated large tags |
| Rotation |
Change 30%+ per post (repetition = spam detection) |
TikTok
| Item |
Recommendation |
| Count |
3-5 (caption character limit) |
| Placement |
Directly in caption |
| Composition |
1-2 trending + 2-3 niche |
| Special |
#fyp #foryou have no effect (officially denied) |
| Strategy |
Trending sound + niche hashtag combination is optimal |
LinkedIn
| Item |
Recommendation |
| Count |
3-5 (excessive use flagged as spam) |
| Placement |
Bottom of post |
| Composition |
2-3 industry keywords + 1-2 role keywords |
| Strategy |
Use followable hashtags (#marketing, #startup) |
X/Twitter
| Item |
Recommendation |
| Count |
1-2 (3+ decreases engagement) |
| Placement |
Naturally within tweet body |
| Composition |
1 trending + 1 niche |
| Strategy |
Natural keyword insertion is more effective than hashtags |
Hashtag Research Process
Step 1: Seed Keyword Collection
| Source |
Method |
| Competitors |
Collect hashtags from competitor accounts' popular posts |
| Autocomplete |
Enter keywords in Instagram/TikTok search -> collect autocomplete suggestions |
| Related Tags |
Collect "related hashtags" from hashtag pages |
| Community |
Observe hashtags used by the target audience |
Step 2: Hashtag Analysis Matrix
| Hashtag |
Post Count |
Tier |
Relevance |
Competition |
Final Score |
| #[tag] |
XXX,XXX |
Large/Med/Small/Niche |
1-5 |
1-5 |
Weighted sum |
Final Score = (Relevance x 0.5) + (Optimal Competition x 0.3) + (Trending x 0.2)
- Relevance: Degree of exact match with content
- Optimal Competition: Neither too high nor too low is optimal (inverted U-curve)
- Trending: Growing tags get bonus points
Step 3: Set Composition
Pre-compose hashtag sets by post type:
[Educational Content Set]
Large: #marketing #business
Medium: #marketingstrategy #digitalmarketing #contentmarketing #snsmarketing #marketingtips
Small: #marketinglearning #marketerlife #performancemarketing #growthhacking #brandingstrategy
Niche: #marketergrowth #juniormarketer #marketingpractice
Brand: #BrandName
[Product Showcase Set]
[Community Set]
[Trend Set]
Banned Hashtag Management
Shadowban Risk Hashtags
Some hashtags can trigger "shadowbans" on platforms like Instagram, limiting content visibility:
- Hashtags misused for sexual/violent meanings
- Hashtags heavily used by spam bots
- Hashtags temporarily restricted during certain periods
Prevention Strategy
- Search the hashtag before use to confirm normal visibility
- If the "recent posts" tab is empty, restriction is likely
- Change 30%+ of hashtag sets each time
- Do not use large hashtags unrelated to the content
