The True Impact of Flash Bashing


Recently we wrote about the new feature in Google’s Chrome browser that will automatically pause Flash advertisements to save battery life. Since then, Facebook’s security chief called for the end-of-life for Flash, and Firefox temporarily blocked Flash completely, citing security concerns. We thought we’d more closely examine how blocking or removing Flash will impact the advertising ecosystem as it stands.

Q: What if Firefox permanently dropped Flash support?

A: Flash was blocked by default in Firefox for just about 24 hours – but the new standard for tolerance with the technology is pretty clear. If Mozilla were to decide to drop Flash support entirely, over 11% of Flash advertisements would be blocked. That’s over 7 billion ads in the first half of 2015 on Sizmek’s platform alone.

Q: How many ads will the new Chrome feature pause?

A: To answer this question, we have to unpack a few more hidden inside it. If we assume the scope of the feature is limited to Flash advertisements (excluding in-stream video) on laptop computers that are operating on battery power, then we need to know:
• The number of Flash ads delivered to Chrome browsers
• How frequently those browsers are on laptops
• How frequently those laptops are on battery power

The first statistic is easy to come by – Sizmek data shows that Chrome has a 36% share of Flash advertisements (which is essentially the same as its overall market share). Laptop vs Desktop can be a bit more difficult to determine, but research shows that 84% of US households own laptops, vs only 70% with desktop devices.
Further market research indicates that more time is spent browsing on laptops than desktops, meaning that advertising inventory on the whole will skew toward laptop use. Most difficult to sort out is how often laptop users are on battery power. The web is full of studies, opinions, blog posts, and pseudoscientific rituals for optimizing battery life – but no studies about how frequently users browse while plugged in.

We made the conservative estimation that 1/3rd of the 35% of Flash-based ad traffic in Chrome is served to a battery-powered laptop. The estimation is particularly safe given the skew of younger users in non-default browsers (like Chrome) and the higher concentration of laptops among younger demographics.
With this filter in place, we find that 27.8 billion impressions would have been paused in Q1 of 2015. Analysts predict that the new Chrome feature will be released and fully adopted by September, which means that 12.1 billion ads will be paused between its release and the end of the year.

You can sail clear of these stormy waters by adopting HTML5 as your technology choice for digital creatives. HTML5 ads will not be impacted by the feature, and thus will behave as the advertiser and publisher expect, regardless of the user’s browser or battery status.

Q: But what about desktop environments that don’t support HTML5? How much inventory is put at risk by the format change?

A: We compared the rate of default – that is, when a rich media advertisement isn’t compatible with the site or browser where it is served, and so fails over (or “defaults”) to a static image – of Flash creatives and HTML5 creatives in Q1. In desktop/laptop environments, HTML5 creatives default less than 1% more frequently than Flash creatives; at rate of 7.2% for HTML5 vs 6.6% for Flash. If we treated the Flash creatives paused by Chrome as defaults, than the Flash default rate for Q1 climbs to 15.7%, more than double HTML5.

And this does not take into consideration how HTML5 solidifies a campaign’s mobile offerings – there are so few mobile environments that support Flash creatives, Flash default rate on smartphones and tablets is nearly 100% (98.6% in Q1), compared to only 8.3% for HTML5. We take a much deeper look at mobile inventory in our Sizmek Mobile Index, available here.

Q: So what’s next?

A: It’s clear that browser makers, major publishers, and other leaders in the web development community are comfortable with a hard line against Flash, even if loads of web content (and the ads that support that content) still rely on the technology. And even with generous allowances in our estimation, it’s clear that these moves could and will have a substantial impact. Many open questions remain:

• How will these paused creatives be measured?

• How will they be billed?

• Will they be counted against campaign frequency caps and delivery limits?

Nobody takes the time to rewrite their tools and upgrade to HTML5,” says Facebook security chief Alex Stamos, “because they expect Flash4Eva.” Fighting against this ingrained expectation will not be easy, and finding creative designers (or training designers to become) fluent in HTML5 is certainly not a simple endeavor. But agencies and web publishers may find that increasing adoption of the more stable format is much easier than answering all the questions now raised by their Flash based campaigns.


Please enter your comment!
Please enter your name here