Stanford.edu

Tests Summary Summary Widget Summary Table

Latency Summary Latency Test Overview Latency Graph Latency Details Compare Latency Stanford rating table Latency Importance

DNS Summary DNS Speed Tests DNS Graph Compare DNS Stanford Rating Table

Page Speed Summary Page Speed Summary Page Speed Load Objects Page Load Tips

Site Speed Laboratory Summary for this website

Summary Widget

Website Latency (Ping Test)

Latency History Graph

How to read the graph's data

Compare Latency Statistic

Compare latency

vs

Our Servers

Usa

NY, New York
Florida, Tampa

Germany, Frankfurt
Canada, Montreal
United Kingdom, Gloucester

Dynamic of the Latency Changes for the website from various locations.

Why the low latency is important with this site as an example

To display this website browser needs to download 68 objects. Though modern browsers use parallel connection now, we'll simplify the conditions and calculate the time spent by the browser for these sites in a single-threaded mode. Time required to establish HTTP connection only: + preparation of the webserver response. Number 3 in the formula is a three phase of a TCP connection: sending a SYN request from the client to the server, sending ASK + SYN from the server to the client and the ASK sending from the client to the server. Though we can not calculate the time for a webserver response, the total time for sending and receiving requests for each of the object will be: Of course, modern browsers use several simultaneous connections to download content from a domain (from 2 to 8 at the time of writing this document). And in some browsers the number of connections can be set manually (Opera). But, if you impose these data on the first site Carnegiefoundation.org from the latency table for the city Stanford, it turns out that the difference in time only for the sending and receiving requests to the web server could be: 1.84 sec. Sure, making such calculations must always take into account the audience site, and always make an accent on the fact that the server should always be so geographically closer to the primary audience as it possible. Also we recommend to pay attention to an article about importance of the latency on our site.

DNS Lookup Speed

DNS History Graph

Compare DNS Speed

Compare DNS Lookup

vs

Our Servers

Usa

NY, New York
Florida, Tampa

Germany, Frankfurt
Canada, Montreal
United Kingdom, Gloucester

The table below shows the dynamic of the DNS Lookup speed changes for the website from various locations.

Page Speed Test

   

Page Speed

Page Speed is a key factor in the website speed estimations (along with DNS lookup speed and website latency). Speed of the page loading, sophisticated subdomains usage for static content (images, css), optimized code (html, js), hosting on fast servers with good latency to its audience leads to good results and, ultimately, to a good website profitability.

While conducting tests we emulate a visit your website using the browser Firefox 3.5.6. Test results of the page loading listed in this table:

Ways to improve website performance

Once the document (website's main page) has been downloaded, we analyze it. It allows us to give webmasters some recommendations on how to optimize page load speed. We analyze the internal and external factors.

Reduse DNS Resolve

Browsers must resolve site's hosts to display all of the elements of the Stanford.edu's page. Every domain name resolve requires a time, which precedes the load from a specific domain. You can speed up page loading by reducing the number of hosts, from which elements of the page are loaded.

Enable GZIP Compression

This is a good practice to give the user content in compressed form. We've detected 104.75 Kb of content that could be compressed. By compressing content you can achieve the size of ~ 26.19 Kb (saving 78.56 Kb). This would accelerate the page loading speed and would positively impact the bandwidth savings in Stanford University.

Keep Components under 25K

Users often visit web site using a mobile phone. Even the most modern and advanced gadgets have some limitations. For example an iPhone browser do not cache page components bigger than 25K. So you should strive to ensure that scripts, CSS-styles, etc. are less than 25K.

Alternate text for the images.

As you may already know, search engine that indexes your pages, do not have eyes. And all content you are trying to display on your site is just HTML & CSS code, images, Flash, etc for the robots. It is It's extremely important to describe each of the objects more datailed using ALT, TITLE and other tags.

Our robot has determined that the stanford.edu page has 6 images, which has no tag ALT. Tag ALT is a text field, that specifies an alternate text for an image. But not only. Search engines consider this tag as a supplemental result for assessing the content of your page entirely. Do not ignore it!

For example your page has this code:

96.

<img src="/images/nav/thin_bar_white.jpg">

We recommend you use rich img code in that place:

96.

<img src="/images/nav/thin_bar_white.jpg" alt="business">

P.S. We used keyword for this image that seemed to us most important for your page. But you can go ahead and give to each of 6 images an ALT text. This problem requires responsible approach. No need to put the same text at alt tag of the all images. Imagine that the pictures do not show and try to make sure that users understand what is shown on the picture based on the alt tag

Specify image dimensions

The browser needs to know the actual dimensions of image (width and height) to quickly render webpage.

We have detected that the 32 images has no width and height.

If you do not specify the width and/or height, then in order to correctly display the image browser needs to download it and only then put on the screen. Otherwise, there is a chance browser will display elements under the picture inappropriately as well as the possible presence of "jumping" effect of the page content (until the page load is finished).

We provide below example of how to define the image http://stanford.edu/images/icons/download.gif dimensions:

315.

<img src="http://stanford.edu/images/icons/download.gif" class="imagewide" width="IMAGE_WIDTH_HERE" height="IMAGE_HEIGHT_HERE">

Using H1 tag

Show main content of the page clearly. One of the most important internal website ranking factors is the H1 tag. It's a title of the document and the most powerful keyword, actually. Our robot did not find the h1 tag on the page.

Use the TITLE for A tag

As we have already mentioned, search engine bots can not see the site like a people do it. They more focused on the descriptions of certain page elements. That's why we strongly recommend use TITLE operator in the "A" tag. This operator not only describes the page referenced by the page more detailed, but also takes part in the page describing. For example, you have the link on your website:

49.

<a href="#content_main">Skip to Content</a>

We suggest you to remake it in this way:

49.

<a href="#content_main" title="education">Skip to Content</a>

Of course, you can (and should) use more proper phrase instead of the "education" that describes the link #content_main.

Number of links on the page

Try to avoid a large number of links on the page. Our robot have found 192 links on this page. Google recommends to keep the number of links fewer than 100. More information about this topic you can find in Google Webmaster Guidelines and in the post "How many links per page?" by Matt Cutts (head of Google’s Webspam team)