Question

Topic: Research/Metrics

Sms Service User Profile

Posted by Anonymous on 8/1/2008 at 12:16 AM ET 125 Points
Hello KHE experts,

My dilemma is fairly straight forward. I need to understand the profile of my audience.

The business: They use a group messaging service. There are two ways to create a group, either online or by sending an SMS to a message servicing hotline. You begin the networking (or grouping) process by inviting your friends/acquaintances either through the web or through SMS by providing their cell nos. (You can have upto 250 people in your group)

Current User base: 8 million

Question: How do I go about contacting them and obtaining responses to relevant demographic & behavioral questions, without disclosing who I am (to avoid bias). I have their mobile numbers. Does it make sense to first call them and ask for an appointment for a F2F meeting, to conduct the profiling...or should I ask them for their e-mail id and ask them to participate in a survey? While the former is going to be more expensive than the latter. I'd love to know if there are any other suggestions out there. FYI, we need this profiling information to share with advertisers.

Thanks in advance!

S

To continue reading this question and the solution, sign up ... it's free!
SUBSCRIBE
Or, SIGN IN

RESPONSES

  • Posted by Mikee on 8/1/2008 at 10:32 AM Accepted
    I think you could send them an SMS asking them to complete a quick survey online. You need to really pitch it as you need information to give better service.

    I do not think you need to worry about whether they know it is you or not. I am asked to take surveys by vendors all the time. I do not really think you will get that much bias.

    This survey can then accomplish two things. Info for your advertisers and info for you to improve your service. Do you have any enhancement ideas you want to float to your customers? This would be the time to do it. I do caution you to keep the survey fairly simple with a lot of radio buttons or checkboxes, something that can be done quickly.

    You may want to enter people in a drawing. Seeing that your service is catering to cell users, perhaps an iPhone or blackberry with 2 years of service would be a great prize.

    Good Luck,
    Mike
  • Posted on 8/1/2008 at 7:47 PM Member
    I suggest you look into vendors that can do the survey via mobile (SMS). Since you have mobile numbers, you may face a less complicated path doing it that way than by getting email addresses and then sending them an email survey. First of all, all of your users may not necessarily have email addresses, so you might introduce bias if you do an email survey. Secondly, asking for information twice (first for the email, and then for them to fill out the survey) would likely generate a higher non-response rate.

    However, you may want to ask for email address in the survey so you can contact them that way in later surveys. Also, by observing other demographic and behavioral responses in terms of who provides you with an email address, you may gain additional insight into your customer base.
  • Posted by Jay Hamilton-Roth on 8/3/2008 at 10:00 AM Member
    Why not contact an existing service such as: https://www.engenus.com/?section=10&sub_section=59 or https://www.mobilestorm.com/sms-marketing/ or https://golivemobile.com/ (for example)? They might already have the data in-house or would be able to help you develop a campaign to acquire it.
  • Posted on 8/3/2008 at 10:48 PM Author
    Thank you all for your answers, thus far. As far as recruitment/administration goes, your inputs have been valuable.

    However, in terms of representation, how will the weighting of the data take place? Knowing the customer base (8m) is one thing, but reaching out to the 'right sample' is another. Any thoughts?

    BR,
    S
  • Posted by steven.alker on 8/8/2008 at 5:44 PM Accepted
    The “What Questions” and incentives to answer a survey have been covered well

    To answer your last question about sample size and weighting, all you need to do is to decide what level of confidence you with to achieve using your user community as 8 Million. Remember, surveys can never be totally representative so you will need to set an achievable level of confidence to give you some degree of certainty that your results are meaningful and can be acted on.

    There is a law of diminishing returns with sample sizes and the question you decide to ask. For example a surveying a sample of 1,000 users might give you a 95% confidence result. Sampling 10,000 might give you 96%

    Let us know what accuracy you want to work to and I’ll either show you how to do the equation or refer you to a widget which will do it for you.

    Weighting doesn’t come into it unless you wish to either compensate for or encourage heavy users. A randomly selected sample should do the trick. If you want to weight your results depending on usage, then you need to construct an arbitrary weighting factor relating to minutes used, text’s send and money spent. Then you can use it to modify the answers, assuming that you have not anonomised the results of the survey.

    I’d also go with my colleagues to say that having a survey come from you won’t make a blind jot of difference!


    Best wishes




    Steve Alker
    Xspirt
  • Posted on 8/11/2008 at 12:13 AM Author
    Thanks all for your inputs!
BACK TO TOP

Post a Comment

Related Discussions

Community Info

Top 25 Experts
(Research/Metrics)