Decoy Experiment

An Experiment
◦ An experiment is the ideal way to test the effect of a
psychological principle in the real world
◦ For an experiment, we need to have two
conditions (otherwise known as a randomized control trial or
an RCT). These two conditions are the following:
▪ Control Condition (where you do nothing at all – test the
existing scenario)
▪ Experimental Condition (where we apply the
psychological principle)
◦ We need to have two conditions, cause how else can we
measure if the application of the principle has worked or not.
◦ Always start the design of your experiment with the control
experiment.
• Decoy Experiment (otherwise known as the asymmetric
dominance effect)
◦ You design a decoy experiment with a control condition and
an experimental condition.
◦ In today’s decoy example (popcorn experiment)
▪ The Control condition had 2 options:
▪ Option A – Small Size : $3,50
▪ Option B – Large Size : $7,50
▪ The Experimental condition had 3 options:
▪ Option A – Small Size : $3,50
▪ Option B (minus) – Medium Size : $7,00 (DECOY)
▪ Option B – Large Size : $7,50
◦ The Decoy is meant to sway people’s choice to the more
superior option, as you can see in the example above, as for
$0,50 more you get a better value for money option. It’s a nobrainer. When your mind makes this comparison and decides
B is better than B(minus), then your brain is tricked to believe
that B is also better than A.
◦ In the Control scenario, the data shows that the majority of
people will choose the least riskier option (cheaper option),
as insights from cognitive psychology show that it costs a lot
of mental resources to decide between Small and Large (e.g.
do I really want that much? I want to diet? did I already eat
something before doing this experiment? should I buy bigger
and share with my friends?), so the cheaper option is the
safest option to choose.
◦ My own data, from the past 10 years running decoy
experiments for B2C and B2B in various scenarios and
industries, shows me that in:
▪ The Control condition on average 70 – 80 % will choose
the smaller option
▪ The Experimental condition on average 65% will now
choose the more expensive option
▪ Comparing the change in choice (change from small to
large between both conditions), the difference is on
average about 35% thanks to the decoy effect.
Note when conducting your own decoy experiments: Don’t worry if
you don’t achieve the results you expected to achieve. Learn from it and
think (reflect) about what you can do differently next time! This is what
marketeers, psychologists and business managers do all the time (It
took me once 1 year to get an experiment right for a client)! Also keep
your experiment as simple as I did in class! Don’t overcomplicate!
Finally, apply it to any business situation or customer interaction you see
fit and where you can collect data quickly.
Compromise Effect
• Similar to the above, only now the middle option is the option you
want your customers to choose.
• In today’s compromise example (popcorn experiment – poll)
◦ The Control condition had 2 options:
▪ Option A – Small Size : $3,50
▪ Option B – Large Size : $7,50
◦ The Experimental condition had 3 options:
▪ Option A – Small Size : $3,50
▪ Option B – Medium Size : $5,50 (COMPROMISE)
▪ Option C – Large Size : $7,50
◦ The difference between the decoy and the compromise
effect, are the options and/or price you are offering. In our
experiment today it was only the price. Make sure it is exactly
in the middle. We humans just have this thing for things in the
middle.
◦ Also look at how the options are framed for the compromise
effect: A, B, C for the compromise effect. Whereas in the
decoy experiment the options are A, B-, B.