Getting to grips with mobile phone market research is certainly a learning curve, any new methodology is. When using a pre recruited panel who have agreed to conduct surveys over their mobile phone is much easier than surveying individual’s cold.
At Fly Research one of our methodologies is to use the internet browsers on mobile phones, we send an SMS message to an individuals mobile phone with a URL embedded into the message, once the URL is clicked on, it launches the internet browser on the phone and the individual is then able to take part in the survey.
From our own experiences, we know that 15 questions is the limit, we can ask different types of questions from, yes/no to multiple select, open ended, ranking, and scale questions. When asking more than 15 questions per survey, we’ve noticed that the difference between the response rate and the completion rate increases from 10%, to 30% or even 50%.
Fly Research has just recently carried out a study for a mobile phone operator using the clients own customer database. The client was looking for 300 completes’ and we were given a sample of 15000 customers.
Having decided to send out the questionnaire to a sample of 1000 customers at a time, the next major challenge was to decide on the message that the sample should receive inviting them to take part in the survey, the client wanted us to mention the costs involved in taking part in the survey on the mobile phone (in our experience, this is normally 1p-2p per survey on the condition that there is no big logos inserted) , how to get the survey started i.e. clicking on the URL and to state that the survey was from the client. The entire message that we sent out consisted of 320 characters (2 SMS messages).
The survey was sent out at 1pm on a Tuesday and we received 20 completes within the first 2 hours (you have to remember that this sample is cold and has never taken part in a mobile survey before). Within the next 24 hours no further surveys were completed. The next day, the survey was sent out to a further 1000 with very similar results.
We then decided to shorten the SMS message that was sent out to each recipient and we sent the message out at 6pm instead of 1pm. This time we were determined to keep the SMS message down to 1 SMS message i.e. 160 characters and we also removed the cost element from the message.
This time the response rate increased dramatically with 100 survey completed within the first 2 hours. One thing to mention is that with mobile surveys, most questionnaires are completed within 2 hours and then only a handful l is ever completed after that.
We continued to follow this procedure and the 300 sample was captured within the next day and a half. It’s also important to point out that no incentives were being rewarded for taking part in this survey.
At the end of the day it took us a sample of 7000 to get 300 completes. As stated before, this is the first time that these people have taken part in any mobile survey, they were not incentivised and there was a small charge for completing the survey on their mobiles.
Lessons to learn from conducting mobile surveys:
· Keep the SMS Invite message to within 160 characters
· Aim to send out the survey between 5pm-6pm
· Try not to mention the cost (this is not ideal) but for people on PAYG phones it essential, those on monthly contract normally have data bundles and the costs are included in this monthly plan
· Offer an incentive
· Keep the questionnaire to with 15 questions.
· Keep the survey question types fairly simple i.e. don’t ask complicated matrix questions.
· Try not to include too many images within a survey
· Most responses will be captured within the first 2 hours of the SMS invite being sent out