Usage statistics - View per Mobile Web Best Practice (BP)

Period

Start date: 1 April 2008

End date: 10 January 2010

Duration: 650 days

Number of Best Practices failed

Repartition per number of best practices failed
Per number of best practices failed
Number of BPs failed URIs (%) Domains (%)
mobileOK pages 9% 4%
1 BP failed 5% 4%
2 BPs failed 5% 6%
3 BPs failed 7% 8%
4 BPs failed 9% 11%
5 BPs failed 9% 12%
6 BPs failed 8% 11%
7 BPs failed 8% 11%
8 BPs failed 8% 11%
9 BPs failed 8% 10%
10 BPs failed 8% 9%
11 BPs failed 6% 7%
12 BPs failed 4% 6%
13 BPs failed 3% 4%
14 BPs failed 2% 3%
15 BPs failed 1% 2%
16 BPs failed 1% 1%
17 BPs failed <1% <1%
18 BPs failed <1% <1%
19 BPs failed <1% <1%
20 BPs failed <1% <1%
21 BPs failed <1% <1%

Best Practice(s) failed

Per Best Practice failed
Best Practice Best Practice explained URIs (%) Domains (%)
VALID_MARKUP Create documents that validate to published formal grammars. 78% 88%
MEASURES Do not use pixel measures and do not use absolute units in markup language attribute values and style sheet property values. 60% 69%
PAGE_SIZE_LIMIT Ensure that the overall size of page is appropriate to the memory limitations of the device. 57% 68%
CHARACTER_ENCODING_USE Indicate in the response the character encoding being used. 48% 60%
IMAGES_SPECIFY_SIZE Specify the size of images in markup, if they have an intrinsic size. 48% 59%
CONTENT_FORMAT_SUPPORT Send content in a format that is known to be supported by the device. 39% 52%
CACHING Provide caching information in HTTP responses. 39% 50%
TABLES_LAYOUT Do not use tables for layout. 35% 38%
POP_UPS Do not cause pop-ups or other windows to appear and do not change the current window without informing the user. 32% 35%
NON-TEXT_ALTERNATIVES Provide a text equivalent for every non-text element. 27% 33%
HTTP_RESPONSE (Not a real best practice) 27% 33%
STYLE_SHEETS_USE Use style sheets to control layout and presentation, unless the device is known not to support them. 26% 26%
TABLES_NESTED Do not use nested tables. 26% 25%
CHARACTER_ENCODING_SUPPORT Ensure that content is encoded using a character encoding that is known to be supported by the device. 21% 27%
EXTERNAL_RESOURCES Keep the number of externally linked resources to a minimum. 16% 23%
PAGE_TITLE Provide a short but descriptive page title. 15% 11%
IMAGES_RESIZING Resize images at the server, if they have an intrinsic size. 14% 18%
IMAGE_MAPS Do not use image maps unless you know the device supports them effectively. 10% 14%
OBJECTS_OR_SCRIPT Do not rely on embedded objects or script. 8% 11%
MINIMIZE Use terse, efficient markup. 7% 10%
NO_FRAMES Do not use frames. 6% 8%
GRAPHICS_FOR_SPACING Do not use graphics for spacing. 2% 3%
REDIRECTION Do not use markup to redirect pages automatically. Instead, configure the server to perform redirects by means of HTTP 3xx codes. 2% 3%
HTTPS (Not a real best practice) <1% <1%
DEFAULT_INPUT_MODE Specify a default text entry mode, language and/or input format, if the device is known to support it. <1% <1%

A few notes

mobileOK, tests, and sub-tests

Tests performed by the mobileOK checker are defined in the mobileOK Basic Tests 1.0 document. The term sub-test used below refers to an aspect of a given test that may FAIL.

How to read the percentages

The statistics only take into account the URIs for which the checker could run the tests. The checker typically cannot run the tests when:

A given URI may yield different results in time. When a URI is checked more than once during the period, only the most recent result appears in the statistics. This ensures that statistics are not biased by pages that include a mobileOK logo along with a link to the checker for instance.

URIs or domains?

Viewing statistics per URI isn't the most representative view one could think of. A better view would be "per website". Unfortunately, there's no automatic way to link pages of a website together. The statistics include a view per domain name, but domain names are often used by many different websites, and thus this view is probably even more misleading than the view per URI. The better view is somewhere between the view per URI and the view per domain name, closer to the view per URI...

Does it represent the state of the Web?

No! One must keep in mind that the URIs that were checked come from users that chose to use the mobileOK checker. This does introduces a bias in the sample oriented towards mobile-friendliness. Figures in terms of the number of Mobile Web Best Practices not followed would likely be far worse if the stats were computed from a representative set of URIs. However, these stats do provide a useful view of the most common problems content authors encounter when moving to mobile.

On missing Mobile Web Best Practices

Also keep in mind that the tests only cover a restricted set of the Mobile Web Best Practices. Besides, they sometimes cover only a portion of one given Best Practice. In particular, a few Mobile Web Best Practices that cannot be automatically checked, such as CENTRAL_MEANING, SCROLLING are not covered by the tests, and as such cannot appear in these statistics.

Alternatively, some generic tests are performed that do not match any of the Mobile Web Best Practices and are identified by their internal name within the mobileOK Basic Tests 1.0 document, e.g. HTTP_RESPONSE, META_HTTP_EQUIV.