Importance of Feature Phones in Emerging Markets
By Stefan Kuegler, Director of Mobile Research
Consumer behavior in emerging economies challenges
First, Device Definitions:
While it is important to understand what a feature phone is, finding a reliable definition can be hard. With the world of the mobile phone being complex and different types of phones being introduced all the time, we need to have an easy definition to reference. The two definitions below have been compiled based on multiple sources.
A smart phone is “a mobile phone that is able to perform many of the functions of a computer, typically having a relatively large screen and an operating system. It will have built-in applications and Internet access turning the once single-minded cell phone into a mobile personal computer which contains features that, in the past, you would have found only on a personal digital assistant or a computer.”
Taken from Oxford Dictionary, Techopedia and PC Magazine Encyclopedia
A feature phone is “a mobile phone that incorporates features such as the ability to access the Internet and store and play music but lacks the advanced functionality of a smartphone. Feature phones are basically low-end mobile phones. The difference is that they have more computing abilities than “dumb phones”, though are less feature-packed than smartphones. Traditionally speaking, feature phones do not run on a smartphone OS such as Android, Symbian, BlackBerry or Windows Mobile and often run apps based on Java ME and BREW.”
Taken from Oxford Dictionary and PC Magazine Encyclopedia
There is a third category called ‘dumb’ or ‘basic’ phone, which only allows texting and phone calls. Feature phones can fall into this category if the user does not access the Internet with their phone. These ‘dumb’ phones are a category of phone that we cannot access via Web or App; therefore, they are excluded from any discussion.
From a research point of view, the issues we need to concern ourselves with are the screen size difference between the phones and what the phones can actually do based on the differences in the operating system. These differences will drive what can be done on each phone type and what a respondents will or will not be able to do.
The smaller screen size means that questions need to be shorter and answer lists need to be restricted both in number and length. The operating system also affects the type of questions that can be shown and completed with more complex questions taking too much time to load and not being displayed properly.
Second, the Facts:
The world is awash with thoughts about smart phones and how quickly consumers are taking them to heart and making them a part of everyday life. They are part of popular culture and now combine all the things we need (i.e., wallet, watch, calendar, PDA, entertainment on the go and, of course, a phone).
Overall, there are close to 4.5 billion world phone subscribers, meaning that nearly 65% of the world’s population has some sort of phone. Of this number, 40% of the global population are smart phones users while the remainder use the feature or ‘dumb’ phones. The table below shows that the gap between mobile penetration and smartphone penetration in emerging countries is still wide.
Source: EMarketers Worldwide Mobile Users Report 2013
Most mature markets have about 50-60% penetration of smart phones among all phone users. With this level of penetration, we can be confident that we will reach a representative population.
However, in the emerging markets we find feature phones are still widely used with key markets. India and Brazil both have low smart phone penetration rates of 20% and 27%, respectively. With such a low smart phone penetration rates, it is clear that feature phones are still the dominant phone type in these emerging markets.
Even in markets that have a large number of smart phones, such as India and China, there is still a relatively low penetration rate due to the large population of those markets and the cost of operating a smart phone.
Today we can predict that the amount of smart phone units being sold will increase year on year. With each record-breaking year it will still take time (a few years at least) for the penetration of the smart phones to reach a high proportion of the population in these emerging markets. Until that time it will be difficult to reach a nationally representative group of respondents via smart phone only samples.
Due to the transient nature of feature phones (i.e., hand-me-downs, secondary phones, etc.), it is difficult to get an actuate read on what the penetration rates are for feature phones. In most cases, an estimate based on mobile penetration removing smart phone numbers is used. There are few sites (if any) that accurately track the number of feature phones shipped since there are many more manufacturers of these phones than smart phones.
The need for Feature Phones
The graph above highlights the trends of device and Internet penetration rates in the U.S. and Brazil. If you compare these markets, we see that Brazil lags behind the U.S. in both Internet and smart phone penetration numbers. With the penetration that we see for the U.S., we can potentially gain the representative sample that we require. From analysis of various partners we have seen that the distribution of the sample is quite similar for mobile respondents compared to the overall population. There are small pockets of demographic groups that we might struggle with, but they can still be reached.
The same cannot be said for some of the emerging markets. Reviews of similar partners and our own panel have shown that certain demographic groups are missing from the sample frame. Even though smart phone ownership increases, it does not necessarily extend to all social classes and demographics. From our analysis we know that higher educated, married, urban and high income earners are more likely to have a smart phone. The same type of respondent is also visible among the smart phone owners in India.
There is also evidence in the type of devices used – for the smart phone users we have Apple and Samsung as their main brands. For feature phone users we see more Nokia and Blackberry phone brands.
These people will give their opinions in the studies, but it might be a very narrow one.
When we include an additional sample provider that has access to some feature phone users, we can see that there is a shift in the income levels to address the balance. Although this does not redress the total shift, we can see even a small proportion of feature phone users provides a more balanced group of individuals for your study.
If we concentrate on just smart phones, we miss a large segment of the community. This might not be such a problem when the penetration of mobile phones is high such as in the U.S., UK, etc. but where the penetration is low it means that we have access to only a small proportion of the population and this can introduce bias into the study results.
Feature phones are more dispersed across the population in many emerging markets – used by all social classes within a market. This may be due to distribution of phones or even the network that will not allow smart phones to work at their optimal performance level (e.g., G2 or G3 networks for smart phones). In emerging markets, high speed networks (required for smart phones) are restricted to main cities, which would indicate that smart phones would also be restricted to main cities. Thus, feature phones also provide us with the means to reach and access people in more geographically dispersed regions of a country. This is especially relevant when moving work from F2F to online (via mobile).
That being said, most suppliers of mobile sample struggle to reach feature phone users. This is mainly a result of the methods currently being used to attract mobile users into surveys – App advertising, Web-page intercept, etc – which might not be ideal for feature phones. At Lightspeed, we have worked to create a platform (based on our PC surveys) that will work with feature phones – creating a ‘lite’ survey template and guidelines to help keep surveys working well on feature phones. We have tested these and carried out research to ensure that responses are similar to other methods and devices.
Survey Design for Feature Phones
In contrast to smart phones, feature phones have restrictions on their features. There are limited touch screens, limited large screen sizes, potentially low connectivity (mostly due to location) and also slow processing speed. Furthermore, the operating system might be very simple to allow other parts of the mobile systems to be able to be used. With all this in mind we have to be considerate of how we build surveys for these phones:
- 15 screens max (or 15 question if one question per screen) – this is a limit imposed by most suppliers of feature phone respondents.
- Be succinct with your wording on questions and answer options – there is little space so don’t waste it.
- For short questions you can put more than one on one screen but be careful how long the screen becomes – a maximum of three as long as the questions are extremely short.
- Code lists should not be more than eight answer options. To make it even more space saving for single you could use drop down lists.
- Open-ended question are tricky. If it is for a single brand for a spontaneous awareness then that might work. For longer like/dislike/main message style question then it might become problematic for responses and they would avoid. It might be good to include a skip function.
- Grids do not work well within a feature phone. There is a risk that not all options will be displayed properly or that the layout will be distorted. It is suggested to convert Grids into single questions with a drop down list. Please consider the number of ‘questions’ that you have as this can induce respondent fatigue. Unlike online surveys, back-to-back split grids should be avoided. The number of questions becomes wearing and data quality can be quickly reduced.
- In your testing, make sure it is easy for respondent to navigate through the survey. When testing the survey, make sure that you test on different feature phone devices – how the survey works on one may not be mirrored of a different device. The more devices the better the ability for more respondents to access the survey.
When considering feature phones within your survey design, there are two main ways you can introduce them:
- Build the survey for feature phones first, using the above guidelines. This survey will automatically work for smart phones and other devices.
- Chunk the survey either by design (surveys into smaller pieces and the feature phone respondents do many small surveys) or by audience (the feature respondents see a smaller tailored survey for them – a subset for other devices)
Either of these methods will work but each have their pros and cons (see table below). Lightspeed recommends using the first method and designing for feature phones: