4

From time to time, we feel the urge to encourage users to upgrade their browser, especially IE. And from time to time, we've rolled out features that attempt to send this message to users, usually "lightbox" dialogs, but sometimes as banners, begging the user to upgrade.

But nothing works. If the dialog has a "close" button, users immediately click it. If it has a "cancel" link, however small and unobtrusive, users find it and use it, even if the link says "continue at your own risk." And if it's a banner, they just ignore it, not even bothering to click a button to close the banner.

(Obviously, we could just block certain old browser versions completely, but that's an extreme choice that we wouldn't want to do to more than a few percent of our visitors.)

It's easy to brainstorm more/different ways to send this message (and there are a number of cute open-source projects to display an "upgrade browser" warning), but I'm wondering: has anyone ever rolled out a "browser upgrade" dialog that actually worked?

By that I mean, does there exist any documented case on any website where the data showed that rolling out a "please upgrade" message was followed by a sharp measurable drop in old-browser traffic?

Or even a reasonably large percentage of users actually click on a "Download" link in the dialog/banner to upgrade their browser?

EDIT: Folks around here can get prickly about proving causation, so I want to clarify: I'm not asking for data that "proves" that a banner caused users to upgrade. All I'm asking for is a documented example where somebody installed a "please upgrade" banner and observed that their old IE usage decreased shortly afterwards.

(Proving causation is hard, but if no one has ever even observed a correlation, that's a surprising and important discovery.)

3
  • 4
    Keep in mind that the people who are hitting your old-browser detector are likely in a situation where upgrading the browser is not an option. Dec 26, 2013 at 6:02
  • 1
    I would love a button in my browser to send websites the message: "be standards compliant", do you think that could work?
    – Pieter B
    Dec 27, 2013 at 12:23
  • Keep in mind this question asks if anyone has evidence of the effectiveness of the upgrade request. NOT why.
    – Itumac
    Jan 6, 2014 at 23:42

7 Answers 7

6

They all 'work'.

The confusion is assuming they are designed to get users to upgrade their browser. They are not designed to do that. As many others have said, nothing you do will likely get them to upgrade short of simply not supporting their current browser. As long as your site works in their browser, there is absolutely no reason for them to upgrade.

As such, the upgrade message's real purpose is to let you off the hook. It's a disclaimer. "Hey, stuff may not work. We warned you..."

In that sense, they work just great.

4

WHY should a user upgrade their browser to use your site? Sure, it sounds reasonable and makes your life easier... but simply saying please without giving them a concrete reason why is unlikely to have an effect.

  • Do you want them to upgrade to make your layout easier? Just stop going out of your way to "support" their old browser. If they insist on using IE 6, let them see the lousy HTML it renders.
  • Do you want to use a fancy feature that their old browser doesn't use? Go ahead and implement it, and pop-up a "please upgrade" link when they try to click on that document-editing HTML5 gadget.
  • Do you want to use a slick presentation tool that's hard to achieve in the old browser? Then when they click on a picture that would be in a lightbox, open a new window with your nag-banner and their image.
  • Do you want to take advantage of slick semantic-web design and a cool new GPU-accellerated menu? Go ahead and do that, but have your old menu remain in-place along with an appropraiately named "nag" banner.

Just telling people to upgrade will rarely work. Telling them that they'll get something SPECIFICALLY BETTER will.

3
  • "Telling them that they'll get something SPECIFICALLY BETTER will." This is not an answer to my question, because it doesn't cite any documented evidence that a browser upgrade warning, however phrased, has ever worked. Dec 26, 2013 at 16:23
  • You're asking for something that's almost impossible to glean, and largely irrelevant unless you're just trying to get out of doing IE-specific CSS hacks.
    – DougM
    Dec 26, 2013 at 18:22
  • 1
    I don't know why you'd say that it's impossible to glean. All I'm looking for is a case where somebody stood up and said "We put up this banner and our IE7 usage decreased 20% within the next week, with no measurable loss in overall traffic." Dec 26, 2013 at 23:32
2

It might help if you looked at the reasons why people continue using what many people consider to be an archaic browser:

  • Users who are unable to upgrade because they do not have permission to do so. These could be people at work or using a public computer, such as at the library or an Internet Cafe. These people will never upgrade, no matter how much you nag them.
  • Users who have the ability to upgrade, but are afraid to. Think of a non-tech savvy user, such as an elderly person. Nagging is unlikely to convince these people to upgrade in a timely fashion, though they might ask a tech savvy relative to help them if they remember. Then again, they may just go to your competitor instead.
  • Users who have a locked mobile phone. These users are unlikely to unlock their phone because they've been scared into voiding their warranty or whatever the popular excuse is. Android 2.3 is still one of the most popular Android OS versions (Google stats showed that 40% of Android users are running this version back in April), and its stock browser is one of the least capable browsers that's widely being used at this point in time. These users could switch to a mobile version of Chrome but...
  • Users who are running Windows XP (or, heaven help us, Windows 2000) or other old OS. These users cannot upgrade beyond IE8 (or IE6 in the case of Win2k). Microsoft has established a date when WinXP will no longer be supported, but older OS versions are quite popular in developing parts of the world because the hardware requirements are considerably lower than that of newer versions. Even if current versions of Chrome or Firefox work on these old OS, there will come a time when they won't.
  • Users who prefer a specific browser where development has effectively halted because they prefer that browser's features. This is where many loyal Opera users (like myself) currently sit. When Opera switched to Presto to Blink, they gutted just about every feature their fans had grown to love with the promise of adding it to future versions. It is worth noting that anyone using the last Presto version of Opera (12.x) is not notified that there are newer versions available.

To design with the attitude of "I'm just going to give you a horrible experience because you're using an old browser" is not going to convince these people to upgrade no matter how much you wish it to. Users of old browsers are used to a below average experience and are likely content just having a baseline experience (can they see all of the content you're offering? can they purchase your products?).

3
  • This isn't an answer to my question. These are all plausible explanations for why a "please upgrade" message might not work, but I'm looking for an example of when a "please upgrade" message actually did work. Surely it must have worked at some point, for somebody, right? If so, who? When? What exactly did they say? Dec 26, 2013 at 23:38
  • 1
    Why do you think it must have worked at some point? Just a crazy thought here, but maybe the reason you can't find any success stories is because there are no success stories.
    – cimmanon
    Dec 27, 2013 at 0:53
  • Don't you think you see less and less of those buttons is: because it doesn't work?
    – Pieter B
    Dec 27, 2013 at 12:25
1

I can't find any documented cases. But, that's not surprising (to me). This is not the kind of thing that people would bother to measure + analyse + write about. (Or at least, that would be my expectation ...)

But I don't think this should influence what you do. Presumably you are trying to encourage people to upgrade so you can stop supporting old browsers. You are never going to succeed in convincing everyone to upgrade. At some point, you simply have to force the issue by ceasing to support old browsers, and leave the "non-upgraders" behind.

The fact that some people can't upgrade (e.g. 'cos their organization won't let them) is not your concern.


The trouble with any strategy designed to encourage people to upgrade is that it could have a different effect. It could cause them to stop visiting your site at all. But you have to balance that against the new users you can attract because your website is better because you don't have to support old browsers any more.

3
  • I disagree that it's not something anyone would bother to measure. Certainly we have measured it, and we can't be the first company in the world to have this idea! But it's not something we've written about, because, you know, it didn't work. If, in fact, it were the case that "browser upgrade" warnings never work, that seems like a pretty important finding, and certainly counter intuitive. (See, for example, DougM's answer, suggesting that you just have to phrase your messaging correctly, which IMO is the received wisdom.) Dec 26, 2013 at 16:29
  • 1
    I didn't say "not something anyone would bother to measure". I said "measure + analyse + write about"! Doing some measurements is easy. The hard part is doing an analysis and forming conclusions that will stand up to rational scrutiny. How do you show that it was your "upgrade now" that caused the drop? How do you show that the users actually upgraded?
    – Stephen C
    Dec 26, 2013 at 18:15
  • I clarified my question. Proving causation can be difficult, but that's not what I'm asking for. All I'm asking for is a documented example where somebody installed a "please upgrade" banner and observed that their old IE usage decreased shortly afterwards. (Proving causation is hard, but if no one has ever even observed a correlation, that's a surprising and important discovery.) Dec 26, 2013 at 23:39
1

It seems to have worked when YouTube launched their banner in 2009.

YouTube IE6 Banner launches July 2009, sharply decreasing the popularity of IE6 and IE7 in favor of IE8

Why did it work for YouTube but doesn't seem to work for individual sites? YouTube is (and was) a major Internet destination. In addition to pushing the banner on users, it made tech media press, encouraging other sites to follow suit with their own banners. Quoting from the article:

Between YouTube, Google Docs, and several other Google properties posting IE6 banners, Google had given permission to every other site on the web to add their own. IE6 banners suddenly started appearing everywhere. Within one month, our YouTube IE6 user base was cut in half and over 10% of global IE6 traffic had dropped off while all other browsers increased in corresponding amounts. The results were better than our web development team had ever intended.

2
  • hahaha... you already knew the answer didn't you?
    – Ren
    May 1, 2019 at 19:37
  • Nope! I posted my question in 2013; this article was just published in 2019. I'm more than a bit surprised that I've never seen a graph like this before today. May 1, 2019 at 23:01
0

Maybe there isn't much data due to the difficulty of ascertaining it.

The usage of antiquated browsers, let's say IE 5-8, is dropping every day and no reliable statistics are available (most sites that care about it and release stats are design/developer oriented and unlikely to be representative).

Besides, as a US-based Internet marketer and web developer, my target market is likely people with iPhones >4 (or equivalent) and modern-ish browsers. While I am a fan of accessibility, accessibility for users of IE 5.5 is unlikely to result in any actual revenue.

0

Google was able to convince some users to install and use Chrome Frame, so I would say it's safe to say that an upgrade message has worked at least once.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.