How to discover what do your customers really want?

Last week, I claimed that people fly Porter primarily because of convenience and not chiqueness. Also, I ended up recommending that you should force your customers to trade-off attributes to find what they really want. Some of you asked me to go into more detail on this topic. Let’s see if smart guys at Bain & Company can help you…

Let’s start with the logic: If you ask a customer whether something is important or not (without a proper context), most probably the answer you will get is going to be  “YES.” For instance, is convenience important for an airline company? The answer is “Hell, yeah.” Is sitting on a leather seat important. “Of course.” How helpful is this exercise? Even if you ask your customers to rate these attributes on a scale, probably nothing will jump out, because for customers everything is important. If your goal is to figure out what do your customers really want, then Bain & Co says that you should use the research technique called “MaxDiff.” And I agree. Without going into too much technical details, let me explain what you should do.

First, come up with a laundry list of product or brand attributes-typically that represent potential benefits. (For olive oil it could be colour, viscosity, taste, packaging, shape of the bottle, name etc…)

Second, show participants sets of four attributes at a time, asking them to select which attribute of each set they prefer most and least. (Let’s say first round attributes are: name, taste, viscosity and shape of the bottle. I would pick taste as the most important and shape of the bottle as the least important, provided that I previously used this brand)

Finally, continue showing different combinations of four attributes to identify the most and least important ones. The beauty of this technique is by forcing your customers to trade-off attributes, you understand exactly which attribute is the most important, which one is the second most important etc.  (If you want to learn more, you can watch this presentation by Bain. If you are a researcher looking for more, then read this article.)

Going back to Porter Airlines, I speculate that no matter what, convenience would come first, followed by price, followed by on-time flights and (maybe) then chiqueness. But, then again it is just me. If anyone from Porter is reading this and decides to conduct a “Maxdiff” I will be more than happy to learn the results. Even it turns out that I was wrong!

Today’s actionable tip: Use MaxDiff to figure out what do your customers really want.


7 Comments Add yours

  1. Matt Lesle says:

    In web we call this MVT, or Multi Variant Testing. It allows us to test many different combinations of designs and functionalities to understand what works best in terms of conversions, or any other pre-determined KPI. We basically crowd source the optimisation of the site. It’s cheap, and very efficient.

  2. Soydanbay says:

    Thank you Matt. I am glad to hear that what I have recommended actually works! Cheers.

  3. In principle I like the max-diff approach and understand why you have recommended it. There is something awkward about the traditional approach where every attribute comes out equally important.

    Of course, there are still reasons to be cautious:
    1> The underlying assumption is that people can tell us what is important to them in their choices. All the max-diff approach is doing is forcing people to move considerations further apart (by artificially creating a separation between the rank 1 and the rank 4 in each group of 4). If you are sceptical about the assumption… you should still be sceptical.

    2>I am skeptical that the approach always brings out the right order. The implication is that a forced choice is more thoughtful but I am not sure that this will always be the case.

    3> The differences between low and high ranked preferences may be much smaller in the real-world than is implied by the results – especially in situations where there are a couple of table stakes and a lot of other differentiators.

    The max-diff approach is clearly advantageous in sorting out product attributes but in the brand sense I am actually more interested in what drives brand ratings (a somewhat different question). In a sense, the “food served hot and on time,” which comes out on top in Bain’s analysis may be the most preferred attribute but taking this strategy only matters if you can differentiate yourself on this and if brand ratings are actually related to how well a restaurant is evaluated overall.

  4. Soydanbay says:

    Thank you Richard,

    It is great to hear from one of Canada’s leading market researchers that Maxdiff works. Regarding your comments, I would like to believe that consumers should be able to tell what is most important to them in their choices (otherwise we would be completely irrational. By the way there is a great book on this: Predictably Irrational by Dan Ariel) .

    Going back to the Porter Airlines example, I know (based on my personal experience) convenience and price lead the decision, closely followed by on-time departure. I want to believe that I should be able to make a similar ranking for laundry detergents, mobile phones or shoes.

    As per your other comment, I definitely agree that this method would exaggerate the difference among attributes. However, I think it would get the ranking of top 3-5 attributes. right. That allows us to practice “so-called” Blue Ocean Strategy”, meaning focusing on key drivers, delivering exceptionally well and reducing or eliminating all non-value added offerings.

    I will write a post on this particular subject!

    Thanks again.

  5. Soydanbay says:

    Here is a free online MaxDiff tool:

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s