Keyword density and LSI keywords are very close but both have different meanings altogether. You can easily rank very high in search engines if you follow simple rules which are listed below.
Step 1 : Research on a unique topic. This is the best way to always rank higher in search engines.
Step 2: Write the content organically/naturally targeting your readers. The best way to write the content which can reach users is storytelling methods. Everyone like stories and if your article involves story telling then it’s the best way to convey the information to users.
Step 3: Optimize your content for specific targeted keywords. Just follow the simple SEO optimization technique and optimize your article.
Step 4: Use LSI keywords in your articles to sound natural and organically. The best tool available for free over the internet is LSIGraph and Google Search. You could also use Google keyword planner tool to find out LSI keywords. Just type your targeted keyword in the search and get the complete list of LSI keywords which you need to target. Place the LSI keywords wherever applicable without changing the meaning of the sentence and it should sound natural when it is read. Do not spam with a huge list of keywords.
Step 5: Use the LSI keywords wisely, the best place to use LSI keywords in the article are :-
- Use LSI in MetaDescription
- Use LSI keywords in Image Alt text or in Image Caption
- Target multiple keywords if you have more than 1 image in the article.
- Use LSI in Video Title if your article has video.
Step 6: How many LSI keywords to use? Do not over optimize the LSI keywords, it will have a negative impact on SEO ranking. Depending on your content length and only if it sounds natural while using the LSI keyword I would recommend to use it. On an average, you could use 4-5 LSI keywords if your blog content length is 600+ words.
If you follow the above guidelines properly then definitely you would rank higher in search engine over the time. Also, one more best place to find the LSI keywords is QUORA. Search quora for the targeted keyword and you will find many results and LSI keywords which people are asking for answers in QUORA.