Why Software is Like Wine


It’s no news to regular Neuromarketing readers that people’s experience with products is heavily influenced by their expectations. Expensive wine tastes better than cheap wine, wine from California tastes better than wine from North Dakota, “powerful” pills relieve pain better than others, and so on… even when in all of these experiments there were no differences between the products. Apparently, this applies to software, too. Microsoft’s latest operating system, Vista, has suffered from bad press from the outset. Early users encountered bugs, corporate IT execs demanded that they be allowed to keep using Vista’s predecessor, Windows XP, and Apple heaped scorn on Vista with its PC Guy/Mac Guy ads.

Even after initial bugs were cured, opinions about Vista have still been negative compared to past OS introductions. This was no doubt annoying to Microsoft execs, just as if a North Dakota winery was shipping superb wine but found that people still found it less tasty than inferior wine from California. So, in a move out of the neuromarketing playbook (does Steve Ballmer read this blog?), Microsoft conducted a study that asked people to test-drive and rate a new operating system, “Mojave.” The new OS was, of course, actually Vista.

Lo and behold, software users turned out to be as impressionable as wine drinkers, pill takers, and the rest of humanity. 94% of the users rated Mojave higher than they had rated Vista, and Mojave (post-demo) scored 8.5 on a scale of ten vs. Vista, which was scored at 4.4 (pre-demo). Read about the experiment here.

Unfortunately, this experiment wasn’t as controlled as I’d like – the ratings of Vista were scored before the users saw the demo and hence were really more of an opinion survey. Microsoft could have altered the approach by simply running the same demo for two groups, with one group seeing a “Vista” demo and the other “Mojave.” I’d guess that even under that circumstance, Mojave would still have outscored Vista by a wide margin. A test like that would have done even more to demonstrate that at least part of Vista’s reputation problem is, well, irrational…

  1. Dare says

    Isn’t this like more social proof? The other people didn’t like it at first, nobody liked it, it must be bad.

  2. Roger Dooley says

    Social input is certainly influential. Research has shown that subjects will answer a simple question incorrectly when everyone else in their group (actually confederates of the researcher) does so.

    In the case of a product like Vista, one’s expectations are likely to be shaped by a variety of sources. In addition to direct social input, one also has expert opinions from industry pundits, product reviews, advertising (like the Mac ads that assert that Vista is buggy and that many users would like to uninstall it).

    That’s rather like “California wine” – that the best US wines come from that state is more or less common wisdom supported by a plethora of reviews, articles, and personal comments.


  3. Jos? Luis Campanello says

    I’ve started reading this blog recently and i have to say that i’m not a marketer, but i find this very interesting.

    I’ve read recently a post on this same issue (Mojave experiment). The post link is:


    All in all, i have to say that i dislike windows vista in many aspects. I dislike having to wait 40 seconds for it to calculate the remaining time to copy a very small file from on location to another. I dislike that my dual core machine suddenly stops responding for no apparent reason, etc. It is now even slower that the service pack 1 is installed, etc, etc.

    In the end, i recognize the science and effect of MS trial, buy i have to agree with the conclusion of the linked article:

    Microsoft is trying to tell me that i’m an idiot that doesn’t know or understand how good vista is, despite it is worse than xp.

    I believe *THAT* is not good marketing. The fact that this “expectations” effect exists doesn’t mean it is wise to use it in this particular scenario.

Leave A Reply

Your email address will not be published.