B2B

 View Only
  • 1.  Survey design - Importance levels among a LONG list of attributes

    Posted 10-17-2018 02:00 PM

    Hi everyone,

    I need to set up a survey where we are asking retail furniture shoppers about which product and experience attributes are most important to them, when they shop for sofas for example. We have 25+ attributes for each area, and I'm trying to figure out the best way to ask the questions so that they are not burdensome and they get us the results we need. I haven't worked with a list this long in a while, so I would love your thoughts on this - any new methods?

    Below is a partial example list (not cleaned up).

    We are looking for what's must-have, what's nice-to-have, and what's not needed, and we will be asking those who have recently shopped (past 3 months) for these items, or who plan to buy them soon (next 3 months).

    We're thinking of a design where we present each attribute and ask whether it's a must-have, nice-to-have or not needed, and then once we have a shorter list of only must-haves (recognizing that that will still be a long list), perhaps presenting pairings and asking which is more important in each pairing.

    I need to have a design direction this week, so anything you can suggest would be terrific!

    Thank you,

    MaryEllen

    Partial example list for sofas:

    • Price
    • Sale price/discount
    • Color
    • Sofa Style (mid-century, casual, modern, etc.)
    • Brand (Ashley, La-z-boy, Nigel Barker, etc.)
    • Material/upholstery (chenille, leather, linen, suede, velvet, etc.)
    • Free fabric samples (yes)
    • Sofa Design (standard, convertible, sofa bed etc.)
    • Delivery Date
    • Arm Type (flared, pillow top, etc.)
    • many more


    ------------------------------
    MaryEllen DeMarco CCXP
    Lead Consultant
    Avalyne Research & Consulting, Inc.
    Wayne PA
    (610) 506-1206
    ------------------------------


  • 2.  RE: Survey design - Importance levels among a LONG list of attributes

    Posted 10-22-2018 05:46 PM

    MaryEllen,

    I'm afraid I'm late to the game here, but the old-school MR practitioner in me wants to yell: Run a conjoint study and let the respondent decisions made in replicating the purchase process determine the relative importance of each product attribute.

    The risk with stated importance is that everything is important to me when no trade-offs are involved. 

    In lieu of a conjoint study, can you use a single force-rank importance question instead of rated importance for each attribute?

    Regards,
    Sarah



    ------------------------------
    Sarah Simon CCXP, CX Expert
    Senior Director, CX Consulting
    Confirmit
    "Ability may get you to the top, but it takes character to keep you there." ~John Wooden
    ------------------------------



  • 3.  RE: Survey design - Importance levels among a LONG list of attributes

    Posted 04-30-2019 10:08 AM
    100% agree with Sarah - you need a choice tradeoff framework, either conjoint or maxdiff.  Stated importance is not going to be helpful.  One other approach is pair-comparison--you ask them to choose between a series of pairs (but it's a lot more questions).

    ------------------------------
    Lori Laflin
    Customer Experience Champion
    Cargill
    Wayzata MN
    (952) 742-2684
    ------------------------------