When new things get launched in the world of ICT, one of the questions that typically gets asked is if large or small companies will be fast or slow to adopt the particular technology or technique. This month we will be celebrating the one-year anniversary of the "official" HTML5 specification on October 28th and now is as good a time as ever to check trends in the adoption rate.
We're lucky with HTML5, since tracking how companies have accomplished or delayed the integration of this technology into their own corporate websites requires little effort - once you know the company's URL it is easy to visit their site and see for yourself how they have coded their web resources.
In fact any company with a website is advertising (most of them 24 hours a day) not just if they have adopted HTML5, but precisely which elements from the HTML5 specification they have implemented and how many times they have done that. The transparency and accessibility of this information, coupled with its always being up to date (unlike quarterly earnings, for which you typically have to wait weeks and sometimes over a month) means this is some of the most accurate and reliable data about themselves these companies make available to the public at large.
This weekend I wrote a bot to visit the websites of a sample of 80% of the companies of the Russell 2000, and to scan their home pages for HTML5 and pre-HTML5 elements. My bot went further to collect more information on Bloomberg Business and other financial websites about these companies. And now with my mass of data, I have correlated HTML5 usage with these non-technical indicators, and hope to shed some light on the question if we can predict which companies have made the step and which haven't.
But first, the question I'm sure is on your minds: In total, almost one year after the official HTML5 spec was released, how many are still not using ANY HTML5 elements on their homepages? The answer is 782 out of the 1615 sites analyzed, or 48%, have yet to include a single element from the HTML5 specification.
There are seven non-HTML indicators I tracked and attempted to correlate with HTML5 use. They are:
And the correlations, from least to most significant, are:
Income is technically unrelated. With a correlation of 0.00 there is no link at all between a company's improved income and their probability of using more HTML5 in their corporate website. Those companies using some HTML5 have an average income of $22.4M; those without have an average income of $19.9M.
A negative correlation here, so bad news for the HTML5 club: the less the top-compensated officers make, the more HTML5 elements there are on their homepages. With a correlation of -0.01 there is a weak link on the continuum, but one that gets stronger if you compare companies that have any HTML5 at all with those that don't: Top-compensated officers at companies with no HTML5 make an average of $556k; those at companies with some HTML5 make only $480k, a decrease of 13%.
Negative correlation again, so bad news for HTML5 fans: the less companies sell, the more HTML5 on their homepages, with a correlation of -0.02 between the two variables. But simplify the data to the HTML5 haves and have-nots and the picture reverses: companies with no HTML5 make average sales of $873M; those at companies with some HTML5 make $990M, a gain of 13%.
Negative correlation again, so bad news continues for HTML5 fans: the more the market values a company, the less HTML5 you are likely to find on their homepages, with a correlation of -0.03 between the two. The absolute picture is even closer, with companies with no HTML5 running an average valuation of $1.05B; those at companies with some HTML5 being valued at $1.01, a difference of only 4%.
Negative correlation continues! The more employees work in a company, the less HTML5 you are likely to find on their homepages, with a correlation of -0.04 between the two sets. The absolute picture is one of the most pronounced: It takes an average of 3572 employees to run a company that has no HTML5; those companies with some HTML5 get by with only 3134 employees, a difference of almost 14%.
Finally, we get a positive correlation. Companies that were founded more recently have more HTML5 on their homepages, with a correlation of +0.07 between the two. In absolute terms, companies with no HTML5 were founded on average in 1967; those companies with some HTML5 were founded on average in 1973.
As you might expect, correlation swings back to negative. As a company's C-level ages, the less the probability they will have HTML5 on their homepages, with a -0.09 correlation, the strongest measured in this study. In absolutes, companies with no HTML5 are run by people aged 54; those companies with some HTML5 are run by people aged 53.
Contrary to popular preconceptions, one year on from its official recommendation, small companies (in dollars) do not seem to be significantly better at implementing HTML5 on their websites. What is more telling is the age, not just of the company itself, but especially of its management; with younger companies and executives more often succeeding in making the step to HTML5.
Manpower also appears as a significant factor, with firms with a 14% smaller headcount successfully integrating the new web elements compared with companies that have more staff.