However, I am concerned with the description I just read.
I see many classes of products where the IKEA effect could be done very easily, yet is not. Does it really apply across the board?
I wonder if anyone has done a very strict AB test against user satisfaction: for example, shipping boxes of hardware product that are fully assembled, or where the user must snap one thing together (that is easy and obvious).
Do they observe a difference in user satisfaction?
I am not certain I believe the effect is as described in this article, or as strong as described in this article, and wonder if anyone has done blinded AB tests.
The two part yogurts are just so the granola doesn't get soggy...
Edit: The greatest example of IKEA Effect in food is the story of how all-included cake mix (vs from scratch) didn't sell, until they removed the egg powder and required the addition of your own eggs. Which apparently is an old wives' tale http://www.snopes.com/business/genius/cakemix.asp
There's definitely more do it than that... apart from anything else, sometimes the other side is honey, or essentially fruit jam, which would be fine just sitting at the bottom or top of the yogurt (the way some other brands do it).
I remember a TV ad for a brand of yogurt with separate compartments, so you can add "a little or a lot" of the side item to the yogurt. It was a gimmick that everybody made fun of because not having all of it seemed implausible, but somebody must have done their research and figured that people like the involvement or feeling of customizing their food.
Similar to the transgressive image of Oreos being twisted apart in order to eat them properly. "You don't have to let The Man decide how you eat this."
Regarding your edit: what I'm looking for is a genuine AB test. Example: suppose there was one packet (mixed dry batter) or two, and for B the user had to mix them. The Ikea effect says B gets better customer satisfaction than A.
Someone downvoted all my comments in this thread. Apparently despite the fact that this is a product design guideline that in my opinion is easy (though not trivial) to test, either all the science in all the world can't test it, or we should accept the theory proposed in the article (that the Ikea effect is real, exists, and should he used in product design) without any testing.
Says you. :) I've seen versions that would do just fine with the separated part mixed in. (They're not just with Granola.)
At least we can agree that if there is an Ikea effect, then it is present in mixing your own Yogurt.
However I am not convinced about the presence of the Ikea effect, as there are many classes of product it could apply to but, I don't see such product packaging. don't.
If you go into an electronics store (a demographic that you would think responds stronger than average to the Ikea effect) most items are absolutely fully assembled. There is nothing to snap on, pull off, a sticker to apply, a thumb screw to screw in or unscrew, nothing.
So if the Ikea effect is real (across a variety of product types), I am sure someone has done a test like to see it demonstrated by someone doing a blind AB test using two shipped versions and comparing customer satisfaction.
I'm not going to blindly believe it because someone says it's real. It would be one thing if this explained an aspect of all the products of every category that I see everywhere - but it's just not the case. Companies aren't in the business of turning down free money, so if the Ikea effect were real then you'd see it in products everywhere. More likely, the theory is just false: that an AB test shows no increased customer satisfaction from token user-assembly, or even decreased satisfaction.
This is why I asked if anyone has, in fact, done such a strictly controlled AB test.
> I've seen versions that would do just fine with the separated part mixed in. (They're not just with Granola.)
My guess is that, if you're going to be offering a version with granola, you're going to be designing/fabricating packaging with a separate mix-in container anyway; and since you have to have that package, it's just easier/cheaper to re-use it for other mix-ins that don't technically need to be separate than it would be to come up with another package just for them.
I think you may be missing a bit of the nuance of the IKEA effect - it's not about superfluous assembly, but rather providing a context in which a little bit of user effort is massively amplified to produce results otherwise unobtainable. Take a look at the chart of the sweet spot for IKEA effect: high value with high contribution and minimal effort. The contribution the user puts into the final IKEA product is really huge - going from a flatpack of scraps to a complete piece of furniture - while the effort to do so is arguably pretty minimal.
fair criticism of my reading. Still, there are no AB tests at various points on that chart (different real-world examples that were actually AB tested) - it's just theory.
Surely you can agree with me that AB tests can't hurt. Anyone can say something plausible, after all! (Just see this comment thread - literally every commenter in the thread says something plausible that I could agree with, including you.) The only rigorous experiment the article mentions is about guessing the value of pre-assembled and unassembled Origami cranes.
This is obviously a problematic proxy for real-world products that manufacturers actually sell.
> If you go into an electronics store (a demographic that you would think responds stronger than average to the Ikea effect) most items are absolutely fully assembled. There is nothing to snap on, pull off, a sticker to apply, a thumb screw to screw in or unscrew, nothing.
I've had to connect the monitor stand to the monitor for many/most of my monitors, and pretty much everything at least needs to be plugged in eventually. Unless the batteries are permanently installed, those are often separate too. To say nothing about building your own Desktop...
Your monitor example is good and real, however due to the shape, it's not clear that it could easily be included fully assembled in the same-sized box.
What I point out the lack of, is assembly where it could already be assembled. So, while you do have to attach a monitor to its base, you don't have to put on a rear panel, even if it snaps on, or do anything else gratuitous. (Added by the manufacturer as an assembly step purely for the Ikea effect.)
In cases where something ships with batteries, it's usually not even included separately, but instead, already fully inserted.
Since products don't add a gratuitous user assembly step, this leads me to believe that the Ikea effect is not real. An AB test would probably clear up whether it is.
The "IKEA Effect" was not designed as such, it was stumbled on accidentally. The reason IKEA furniture is assembly-required is the same reason that your monitor stand comes detached from the monitor: because these companies operate on the principle that their products be flatpackable (minimize shipping dimensions).
Giving a theory a name doesn't mean it's a real effect. I would like to see a blind AB study that shows the effect is actually true and not someone's pet theory but which is actually false. I agree that the theory is nice and has explanatory power. Let's do a controlled study to see if it is real.
I wonder if you would like to see an AB test though? I haven't read any of the other comments in this thread yet, so stop me if you already answered that.
You also missed my point that IKEA effect is not necessarily a "gratuitous" assembly step, just any step(s) at all.
However, I am concerned with the description I just read. I see many classes of products where the IKEA effect could be done very easily, yet is not. Does it really apply across the board?
I wonder if anyone has done a very strict AB test against user satisfaction: for example, shipping boxes of hardware product that are fully assembled, or where the user must snap one thing together (that is easy and obvious).
Do they observe a difference in user satisfaction?
I am not certain I believe the effect is as described in this article, or as strong as described in this article, and wonder if anyone has done blinded AB tests.