Perhaps the easiest way to start our discussion about how algorithms affect reputations is with what you, gentle reader, already know: Search.
Everyone uses search engines but not everyone understands how and why they work.
The algorithms that power search engines are very literally using reputation signals to decided what are the most appropriate links to give you after you type (or increasingly dictate) in the box and hit Search.
What Is A Search Signal?
What, exactly, is a signal, you ask?
The dictionary definition is “a gesture, action, or sound that is used to convey information or instructions.”
Signals in the context of search are not that much different. According to Google, its “algorithms rely on more than 200 unique signals or ‘clues’ that make it possible to guess what you might really be looking for.”
1) Signals In Search Queries
Think about how you search. You go to your favorite search engine and type or dictate your query into the search box and hit search. The keywords you enter into the search engine are a signal that tells Google or Bing what you want.
Depending upon the type of keywords you use, your search query may not only tell the search engine what you want but it may also betray your intentions. Let’s take a look at several related search queries to illustrate.
If I search for “guitar,” that doesn’t give Google much of a signal to understand what I want. It will therefore return a variety of links on the Search Engine Results Pages (SERPs), including:
- Local businesses related to guitars,
- A Wikipedia page about guitars,
- Images of guitars,
- Online retailers that sell guitars,
- The Guitar World publication,
- Recent news about guitars,
- Guitar lesson and training sites, and
- Related guitar searches.
If I refine my search by adding a brand name, I will give Google another signal about what I want. If I search “Gibson Guitars,” now I get information about the Gibson Guitar Corporation as well as:
- Product Listing Ads for Gibson guitars,
- The Gibson company website and Twitter account,
- The Gibson company Wikipedia page,
- Images of Gibson guitars, and
- Retailers that sell Gibson guitars.
I’ve given Google another signal from which it can give me more precise search results (Gibson guitars rather than guitars generally). My search query now more accurately reflects the information I want but it does not betray the intent of my search.
Finally, if I enter a multi-keyword search like “Gibson Les Paul Custom Tobacco Sunburst Prices,” I will give Google a search query that indicates precisely what I am looking for and betrays my intent to buy a Gibson Les Paul Custom tobacco sunburst guitar.
Therefore, Google SERPs displays a list of prices and the links to the retailers from whom I can buy such a guitar in a sidebar Product Listing ad. With the exception of a Wikipedia page and a YouTube video, I can buy a guitar from any of the sites Google links to.
As you can see, the keywords we enter into the search box are signals in and of themselves.
The above search behavior can also provide a reputation signal to Google that I am a guitarist.
2) Signals In Search Behavior
Now, think about how you search. You have no doubt searched for something before, clicked on a link that looked like what you wanted and visited the page only to find that it wasn’t what you wanted.
After seeing that the page wasn’t what you were looking for, you probably clicked the back button on your browser and went back to the SERPs to look for a better link. Reviewing the page and returning to the search results probably took all of three seconds.
That behavior is called a “bounce” because you are bouncing from the page on a site without visiting another page on the site or simply by spending very little time on that page.
Now think about all the things Google knows about that process:
- It knows the keywords you entered,
- It knows the link you clicked on,
- It knows you spent three seconds on the page, and
- It knows you then immediately returned to the SERPs and began scrolling through the results.
From that process, Google can deduce that the link you clicked on was not relevant to the keywords you used in your search.
Your own behavior has just sent Google a signal about the link you clicked on that has implications for that page’s reputation.
If enough of Google’s users who searched for the same thing you did and behaved the same way you did by bouncing from that page after clicking on that link, Google will likely not include that link in search results for that query in the future.
3) Search Signals In Web Page Content
Generally speaking, search engines like Google work by using automated programs that traverse the web, find new content, suck all that content into a massive database, categorize the content, and apply weighting factors on different elements of that content. These programs are called Bots, short for robots.
You will notice that more often than not, the keywords you use in a search query are included in the headline of the page you clicked on. That, combined with a bunch of other factors, is a signal to Google that that web page is relevant for the search query you entered.
There are, therefore, signals that producers of content can implement that will give them a better chance of getting visibility for their content in search results and therefore earning website traffic from search engines. This practice is called search engine optimization, which I will address in-depth in subsequent posts.
4) PageRank: Links As Signals
The first search engine, Archie, was launched in 1990. Google was launched in 1998 and, with no advertising budget, quickly became the most popular search engine, which it remains to this day.
Google figures out that the glue that holds the internet together, the hyperlink, had value as a relevance signal. So it started examining the linking structure between sites and assigned values to those links.
Thus, Google’s PageRank algorithm was born (named after Google co-founder Larry Page).
Google treated links as votes but unlike democracies, all votes were not the same. Some links carry much more weight than others.
A link from the New York Times, for example, is considered more valuable than a link from the Kansas City Star. As a national outlet and as “the paper of record,” the New York Times has more authority than the Kansas City Star.
Topical relevance is a factor as well.
Link from an article about online reputations published by the Kansas City Star to this blog post, for example, would be a strong signal to Google to give this page more visibility in search results about online reputations because:
- The topic of both the Star article and this post are very similar, and
- As a newspaper website, kansascity.com is given significantly more authority than other sites because the content published on it is presumed to have been vetted by journalistic standards.
Such links are positive signals that tend to boost a website’s rankings in search results, but links can also send a negative signal.
Once website owners figured out how Google valued links, they began devising schemes to try and boost their rankings in Google search results.
One prominent scheme that was popular for a time was the development of microsites whose sole purpose was to link to the site for which the owners were hoping to increase their search visibility.
Because these microsites often hosted content that was of dubious value in and of itself, and because the pattern of linking made clear the sole intent of the sites was to boost search rankings, Google penalized the sites employing such tactics by suppressing the site’s visibility or banning the site altogether from its results.
Thus, links can have both a positive and a negative effect on your online reputation.
In my next post, I’ll discuss the evolution of search engines to explain how they work in order to set the foundation for understanding the role they play in shaping our reputations.