Recently I have been searching for different things in google search engine,

  1. bower save flag
  2. manage bower cache

The first results for these two different queries were same page https://bower.io/docs/api/ but the descriptions below them were:

  1. When --save flag is used, all additional endpoint are saved to dependencies in bower.json . Bower recommends to always use --save flag to achieve ...

  2. cache; help; home; info; init; install; link; list; login; lookup; prune; register; search; update; uninstall ... bower cache []. Manage bower cache ...

I went down to both links because by reading their descriptions it became evident that the information I am looking for is certainly in that page. Actually the https://bower.io/docs/api/ page has no meta description tag. If they had instead a description tag with a fixed content then for both of my queries I would see same description which most probably won't contain the relevant keywords save flag and cache and I would have the impression that the page might not have information about those keywords.

So by not having description tags a webpage can actually offer users the most relevant information below their title in search results. That is if a page has 1000 lines then we can have 1000 different descriptions shown in search results. Isn't this good? Wouldn't this improve the chances to convey more relevant info to each user in the search results?

  • It all depends upon the description meta-tag and the search query. A well crafted description meta-tag should work for many of the search queries, however, omitting a description meta-tag leaves Google to make one up. We are talking about Google here. Do you want Google always making up your SERP snippet? I don't! I would imagine that for some pages, it may be better not to include the description meta-tag, however, I do believe I would always include one anyway.
    – closetnoc
    Feb 9, 2017 at 16:53

3 Answers 3


Google customises the description that appears in the SERPs based on what you search for, regardless of whether you've included the meta description in the page source or not.

If the content of the page is dynamic and changes often then it maybe difficult to create a suitable meta description so it is probably best omitted in this case. Otherwise, the meta description can provide a helpful default and Google isn't the only service that may (or may not) use it.

  • So even if I use some content in description tag then google would show some other description as per user query? Then what is the benefit of meta desciption tag at all?
    – user
    Feb 9, 2017 at 15:05
  • Google can still show the contents of the meta description if it is relevant to the search and there are other web services that do use the meta description. Just to clarify, the meta description has never (at least, not for a long time) influenced ranking.
    – DocRoot
    Feb 9, 2017 at 17:08
  • So meta description content acts like other content in body tag of webpage. Google chooses, based upon query, whether the description meta content is relevant or body tag's content is relevant.
    – user
    Feb 10, 2017 at 5:49

Google almost always chooses a description to use in the search results that contains the search terms. If you meta description doesn't contain at least some of the words that the user searched for, Google isn't going to choose it in that case.

Leaving the meta description blank (or omitting it) will cause Google to choose a sentence from the body of the page all the time. Google will use the meta description when you have written one and it matches the search query.

There are some big advantages to having your meta description used in the search results. That is your one chance to convince the user to click on your site as opposed to the other sites in the search results. Crafting a custom message for search users can dramatically change the click through rate (CTR) from the search results to your site. A good meta description will:

  • Use keywords users most often search for to pull up the page
  • Contain a "hook". A hook is something about your content that could make it attractive. It often contains a stat or number. Some examples:
    • A stat about the number of items available: "4,300 blue widgets to choose from"
    • How fresh you content is: "3 new reviews" or "Updated 2017"
    • How many people use it: "95 bookings this month"
  • Contains a call to action (CTA). The CTA lets people know what they are expected to on your site. This both lets users know that they can do the action and makes the visitors you do get more likely to be the ones that are interested your offering. Examples:
    • "Buy a blue widget"
    • "Book your blue widget"
    • "Open a blue widget account"
    • "Sign up for the blue widget newsletter"

Leaving the meta description blank leaves out the possibility of this opportunity.


From: https://moz.com/learn/seo/meta-description

Not a Google Ranking Factor

Google announced in September of 2009 that neither meta descriptions nor meta keywords factor into Google's ranking algorithms for web search. Google uses meta descriptions to return results when searchers use advanced search operators to match meta tag content, as well as to pull preview snippets on search result pages, but it's important to note that meta descriptions do not to influence Google's ranking algorithms for normal web search.

Also AFAIK meta tag keywords was also "deprecated" by google, because there is a lot of bad practices around it.

  • No no no, I am not interested in ranking of a webpage. I am interested in how google chooses whether to show meta description in search result or webpage's actual body content.
    – user
    Feb 10, 2017 at 5:50
  • that can be hard to tell, because most of google's algorithms are unknown. It may have more to do with how you generate the content when the crawler hits your site and how far are your content from the description tag (h* tags relevance). Also, as said in the extract, google only uses the description tag when you use the advanced search parameters to search for it, most of the time it should be able to use the h tags to get a more accurate description.
    – LordNeo
    Feb 10, 2017 at 12:42
  • How can crawler determine the way content is generated? It could be generated from php, nodejs, ruby or some other language may be even cold fusion.
    – user
    Feb 10, 2017 at 12:43
  • hitting same website under different conditions (post, get) and checking for changes, using different datacenters, etc. All that can tell if the website changes it's content depending on the search/user/geoloc.
    – LordNeo
    Feb 10, 2017 at 12:44

Not the answer you're looking for? Browse other questions tagged or ask your own question.