Where can I find case studies on productivity gains from dual monitors for developers?

February 2, 2006 1:50 PM

I am a Java programmer working for a reasonably large (read: not-so-quick-to-get-with-the-times) organization, and I'm trying to put together a case for having a dual-monitor setup for all our developers.

I know for a fact it will be a huge productivity boost (as I've used dual monitors in the past), but I have to prove this to my manager via case studies, scientific studies, etc. that already exist on the matter.

Unfortunately, all searches on google for the topic just bring up vendor sites trying to sell their overly expensive dual monitor setups and obviously some very biased opinions on productivity gains.

I've only found one rather weak argument for "anywhere between 5-50%" on a Microsoft site but it is by no means sufficient.

Any relevant links would be greatly appreciated!

Thanks!

posted by PWA_BadBoy to computers & internet (6 comments total)

A huge discussion (at slashdot) about a study carried out by NEC-Mitsubishi, ATI, and U. of Utah.
posted by SpookyFish at 1:59 PM on February 2, 2006

Joel Spolsky discusses how important it is for programmers to have the best tools possible. He gives only short mention to dual monitors specifically, but the article may buttress your overall argument.
posted by ldenneau at 2:07 PM on February 2, 2006

When I was a developer at Microsoft in '99, just about all the developers had dual CRTs. The more hardcore developers had three monitors. This was in addition to the extra testing and email machines we usually had.

Now I'm a freelance software developer, and pay for all of my own equipment. My primary monitors are a pair of Dell 2405 FPWs that are shared between my PowerMac and my biggest x86 notebook computer. I'm considering buying two more 2405 FPWs and a stand to place them directly above the two I already have.

More information on the screen == more productivity for developers. When you work with a highly object oriented language like Java, you really need to have a few source code, object browsing, and help windows open at any time. When you are debugging code, it is very helpful to be able to run the application AND the debugger at the same time, without one obscuring the other. When you are designing software, you often want a copy of Word/PowerPoint/Visio (sometimes all three) running, a few dozen web browsers, and maybe your IDE's help browser running, all at the same time, and all mutually visible.

Next, monitors are very cheap. My 2405's cost $750 on sale, but you are more likely to pay $850 or so now. Suppose they were $1000 each. I would rather have a $1000 development machine and a pair of 2405 FPWs than a more powerful $2000 machine and a single 2405 FPW. More screen space makes me far more productive than more disk/ram/video (which I have already maxed anyway) would. In addition, I will probably keep and use my 2405 FPWs for a long time. My last big monitor was an Apple DVI 22" cinema display from 1999. It is still a great monitor, and is in use on a secondary machine. 7 years of useful life is a lot more than you can expect from a computer.
posted by b1tr0t at 2:58 PM on February 2, 2006

[yeah, I know you wanted studies, but you may be able to construct a logical argument from some of the above]
posted by b1tr0t at 2:59 PM on February 2, 2006

Your organisation probably pays 100K plus after management overhead, office space, health benefits and so on for each programmer.

The cost of more monitors for developers in minimal. The company gets an asset they can resell. The cost is arguably the depreciation on the monitors, and even getting good 24" flat panels as b1tr0t described is only about 750. Even if their depreciation is set at 20% per year you are talking about $150 per year per developer. If this makes your developers even 1% more effective it is a tenfold return on investment using the studies provided by slashdot and your web page reference. If you were to buy high res CRTs for $2-300 it would be even less.

Stiffing people on development tools is a classic piece of a false saving.
posted by sien at 3:38 PM on February 2, 2006

One thing I did was take screenshots of our dev enviornment (jbuilder) and figure out how much space there was by default for the text entry window after the class browsers, message window, etc were gone. It was suprisingly small, like a portable DVD player. That may have helped a little.
posted by true at 5:52 PM on February 2, 2006