Research in a mobile world

A lot of our clients immediately jump to the concept of "reach" when they first begin to think about mobile research and wireless surveys. They worry that only a small percentage of their panel might be willing to take a survey on their phone, and they think that hard-to-reach groups (like teens) are the perfect audience for mobile research. 

Sure, using consumer engagement techniques that make your interactions more personal and portable should make them inherently more relevant, thereby increasing response and reducing churn.

But that's not what's really exciting about cellular surveys. My favourite conversation with clients is the one when the light goes on the realization sets in that mobile research is about so much more than putting traditional online surveys on a small screen.

What's really exciting is the new reality that as researchers we can take advantage of the billions of dollars that device manufacturers pour into R&D. Built-in functions such as taking photos and capturing GPS coordinates are just the beginning - even so, these basic capabilities provide us with some pretty astounding options for gaining insights from targeted mobile groups and communities!

So how many people do I need in my mobile panel? I by no means decry the important science of sampling, but check out this article that explores how big the ideal online research community should be, and hopefully you too will start to become a believer in quality over quantity.

We talk a lot about trying to get deeper insights by using new mobile research tools that are personal, portable and relevant. But maybe we're missing the boat. There's a lot of money to be made in being a professional respondent, right? 

"It's hard to keep up with all of those online polls on my busy schedule..."

Check this out for a fun view of our industry that brings the concept of "professional respondent" to a whole new level:

The rules of research are about to change in Spring 2010. The MRS (UK's professional association for researchers) will begin enforcement of a new ban on use of client goods or services as incentives in a research project.

So many corporate communities, especially branded panels, rely heavily on this very technique for survey conducting.

Who can deny the reason behind it? I remember answering a poll in a teen magazine (was that really decades ago?) where I purposefully skewed my answers to highlight my desire to see more of the content that they were providing in that very issue. I thought it would increase my chances of winning the cash prize! But in fact I couldn't have cared less about Menudo.

With DIY research making a stronger push than ever, how does this affect the position of the professional researcher who is already challenged when trying to coach clients about the "right" way of doing things?

Maybe the promise of mobile research where true engagement through 2-way communication occurs will emerge one technique that can gain answers without bribes. But we'll never escape it will we? Time is money, especially when it comes to hard-to-reach populations.

Maybe it's no big deal. Cash always been king when it comes to the incentive that works for everyone. But where does that leave financial research? :)

In my last post, I wondered about device market share and it's affect on our ability to reach the right people for survey conducting.

I focused on the manufacturer / operating system, but another element also has a significant affect on reach. I guess you could call it Mode. We normally talk about mode in terms of whether we're collecting data over the web, over the phone, face-to-face, etc. Mobile is it's own mode, but there are choices within that mode.  

Specifically, the primary mobile mode possibilities for a mobile survey system are text message, mobile web, and applications.

Check out this interview/podcast from a recent interview at a CASRO event. Heidi Dickert has some good points about how a different type of technology is required when looking at mobile surveys. The message? You can't just replicate the same old approach when it comes to mobile research.





Text messages give you ubiquitous reach, but the commonly held experience among those I speak with is that SMS surveys are good only for very, very quick polls. The burden on the respondent is high, and you can only squeeze in 3-4 questions before participation drops off - and if you haven't seen it yourself, believe me, it doesn't just stumble off the curb, it careens off the grand canyon.

Mobile web feels like it should be familiar since CAWI has taken over as the primary mode of choice in so many countries. However, the problem is that the mobile web isn't great yet for the vast majority of users. You still need a lot of patience. Think about a news page that requires 10 seconds to load, but then you spend 5 minutes reading. That's much more palatable when compared to refreshing screen after screen after screen of a mobile web survey. To reinforce this, check out Mobile Marketer's recent article about slow load time is the #1 performance issue on the mobile web.

Off course we believe in Applications (particularly Java where possible, native where needed). With iPhone App store just reaching it's 2 Billionth download in September, and Facebook's install base leaping from 2 million to 10.5 million in two months with the release of targeted Java versions, the path seems pretty clear to us...

Hardware markets have never been so attractive. 


Check out the hot deal on the ASUS Eee PC with Windows at $285.

MCAPI used to only apply to PDA Survey Software, but with the introduction and proliferation of Netbooks, and powerful mobile survey packages like the ones offered by my company,  now you can easily deploy complex surveys to small and large screens alike, depending on your project need.

The two main considerations in picking a mobile research methodology are capability (what kind of work am I enabled to do?) and coverage (who can I reach with that capability?). Those of us in research care so much about market share of mobile devices because it profoundly affects both elements.

We need to reach people. But it's not just any people we need - we need to reach the right people. A colleague of mine recently held a focus group and noted - "hey - there are an awful lot of artists and musicians in this group..." Did his recruiter get the mix of participants right? Maybe, but probably not (you'd think so too, if I told you the category ).

To a marketer, the more people you can reach, the more mindshare you can gain with your ads to promote revenue growth. Marketers talk about how to reach a target market with their message because they need to know they are spending their money wisely.  However, the reason researchers care about the number of people who see our "message" (say, a survey invitation), is because of our sampling frame. We need the right people to provide us with feedback in order to help solve the business problem at hand.

In short, researchers have to worry about it even more because the science of our analytics depends on it. Without reach, my colleague wouldn't even have had artists and musicians to talk to. 

So, it's very interesting to see what Gartner is predicting for the future of Mobile OS market share. They say Android will overtake Research In Motion’s BlackBerry OS, the iPhone OS and Windows Mobile to capture 14.5% of the smartphone market within three years.

"Symbian’s market share will fall from around half of the global market to just 39%, according to Gartner. Apple will maintain its third place and Windows Mobile will stay in fourth. However, RIM’s BlackBerry OS will fall from second place to fifth,"

If you're thinking about partnering with a survey system supplier to provide cellphone survey capabilities, then make sure they have an expertise and plan in place to meet all of the different flavours of these platforms as they explode in the next two years, otherwise you may never overcome the problem of reach.

If you're a real cheapskate like me, and you've been biding your time before making a new hardware investment, then look no longer. The next release of inexpensive smartphone technology for MCAPI interviews and diary studies has arrived.


Check out the unlocked Palm Treo 500 for $119.  Incredible.


Just make sure you're willing to live with a QWERTZ keyboard - (in Germany, they switch the Z with the Y.. go figure).

Some shameless promotion here, but with Techneos' new SaaS pay-per-complete options, there's never been a more affordable time to equip your field force or diary respondents with a world-class mobile interview system.

For many outside of our industry, Market Research has always been a fuzzy concept, and let's face it, a weird choice of career.

But no longer! A major Hollywood movie (Last Chance Harvey) features a market researcher as a main character.



But what is she doing with pen and paper? What an antiquated approach! We would expect her to be using a computer assisted personal interviewing system or some kind of PDA survey software, of course. Where are these movie script writers doing their research?

Legal Note: This post in no way confirms that my wife convinced me to watch a movie which could be described as a Romance Drama.

Can anyone deny that the conventional understanding of what constitutes "real" research is being challenged? This Research Live article nails it:

"The rise of the Internet and social media has caused the received wisdom to be questioned on issues including sampling, anonymity, the relationship between researchers and participants, and whether there aren’t better ways to understand what people think, feel and do than just asking them."

With online research being at the forefront of so many researchers' mentality, do we move to an approach where we simply monitor and observe? Of course there's the problem of access as it relates to your sampling frame. But at least thanks to Anderson Analytics', we know 'Who's using Social Media?':




Personally, I think a moderate, measured approach is what's needed. If you read one article on this, check out the Zinc research blog entry on Social Media and so-called "legitimate research":

"...the traditional model for marketing research needs to be overhauled, and social media will be one of the impetuses for changing the role of research in the marketing toolkit. That said we should not throw out the existing research model too quickly"

And finally, if you want to read a sound, measured debate on both sides, then I'd suggest checking out this article: Is Social Media Measurement Meaningless?

As it gets harder for businesses (and researchers!) to reach people, the buzz about the affect of social networking applications is starting to get louder in our ears...

Apparently some people are wondering if the rise of direct forms of customer interaction like voice of the customer programs will make the use of Mystery Shoppers for mobile data collection a thing of the past.

Judi Hess, president of MSPA calls mystery shopping "a means to measure customer service, product knowledge, and sales ability." More important, she says, it offers subjective, targeted feedback that companies just can't get from less structured kinds of feedback, like social networking tools or surveys.

Ok, I can buy that. Read more here (you'll need to sign up for a free account).


Tom Anderson in particular is pushing us to think about the role of Social Networks and what he's calling "next gen marketing research". He recently interviewed representatives from Facebook and LinkedIn. It's just a start, but you can see that the social networks are starting to think about the convergence of these worlds, too.

Click here for the Podcast


All three of them will be a highlight of an upcoming ESOMAR panel session. This will be one you don't want to miss...



These are great articles to get us thinking about engagement, the future of survey conducting and the way we deliver insight from the market... But one of my favourite approaches comes from a seasoned researcher right here in Canada.

More on that next week...

I recently read that in this economy "opt out is the new opt in". 

Yeah right.

This was literature on a site (which I won't mention) that sells marketing lists, so I had to chalk it up to wishful thinking rather than benevolent distribution of informed market information.

For those of us that take this stuff seriously and know that blatant Spamming erodes respondent cooperation, we need to know that even when we think we're following the rules, local laws might derail our efforts. As as we move too add SMS survey invitations to our toolkit, the landscape is getting tricky to navigate.

Now, marketers who advertise products and services such as alcohol, tobacco and gambling cannot market to any phone number or address which is registered on a do-not-contact registry within Utah or Michigan. It's all about protecting families, and you can check it out here.....

"But I don't advertise tobacco! I'm in research!"

Maybe... But my point is that we need to be careful. Extend this to our world of mobile data collection for a moment, and is it really so hard to believe that similar legislation could be introduced to protect certain groups.... like children for example, a segment for whom we already need explicit parental consent before we can engage.

"One odd aspect of the laws enforcing these programs is that even if a consumer double-opts-in for an SMS campaign and verifies that he or she is 21 or older, marketers would still be in violation if that person previously submitted their mobile phone number to either state’s do-not-contact list."

Could you be unwittingly inviting minors to participate in your mobile marketing survey? Or are you treating SMS invitations with less ethical scrutiny than email? Might be time to re-scrub that list...

It seems as though reseach online has recently reinvented themselves. They were always a great source for information, but they seem to be exploding with news and original content lately.

A recent article titled "Survey Geek vs. The Cord-Cutters" offers an audio interview with self-confessed 'survey geek' and blogger, Reg Baker.  There are also great links to relevant articles on the cell-only population and the affects on research. You can listen to it here:



This interview is great in my opinion, particularly in that he begins by describing the "cell-only" problem in terms of the sampling frame. He also eloquently yet concisely descsribes the issue of bias due to undercoverage, and how the target group of the research and the topic may affect results. Not to mention cost and data implications of doing a survey on mobile phone. Reg also identifies the "wireless mostly" problem. I won't replicate every point here in text, but here's a researcher who knows his stuff.

Brilliantly, he surmises (and hopes) that researchers will adapt to the problem by matching the method to the problem - execute "fit-for-purpose methodology decision", to paraphrase slightly.

But one thing stuck out to me... there's no mention of three significant and growing ways to conduct a mobile device survey, cellphone survey, mobile marketing survey (mobile research by whatever name you choose): WAP, SMS and Survey Applications.

Check out the latest numbers from CASRO. You might find it's worth getting ahead of the curve by adding other mobile techniques to your methodological tool box.

Mobile is mandatory. 

You can't get away from it and the evidence is everywhere. Even the University of Missouri has made smartphones mandatory for journalism freshmen!

And everyone now agrees that "Mobile" is more than 'my cell phone'. Mobile is changing the way we live. Business and society is seeing the benefit, but we are also seeing the pitfalls.

Rapid growth of the use of the wireless Internet is causing wireless broadband services to slow down and interrupt from data overload, and according to one report, the result is that mobile broadband sales are slowing. Don't get me wrong - there is still impressive growth overall, it's just that the growth is slowing as consumer frustration sets in.

There's more than an inconvenience factor... now we have to worry about 'cellphone elbow'. Doctors are indeed seeing more instances of this.

As people get more engaged with their devices, I'm betting we'll see more opportunities for doing deep and personal research with respondents, as can be done when people are recruited for a targeted, self-completed digital ethnography. 

Regardless of the specific research techniques to be employed, we will undoubtedly begin to see advances in the sophistication of mobile market research methods arising from this cultural shift.

In the meantime, watch this space for more updates on what fabulous new ailments arise as a result of increased mobility of the population!

According to Marketing VOX, mobile marketing is set to explode, with a forecasted 26% increase this year.

The article predicts that local mobile advertising will be the next hot trend - particularly when it comes to mobile search. This info comes from BIA's The Kelsey Group.

This is no wonder. In Canada, Eight per cent of Canadian households have cellphones but no land line. The number jumps to jumps to 34.4 per cent when looking at households composed solely of 19- to 34-year-olds! In the US, it seems that across all age brackets, 1 in 5 households have cut the cable.

This is a tide that no one can fight. To not be making plans for mobile engagement is to miss out on the next evolution of engagement. 

It's important to note, though, that growth in engaging with mobile advertising is coming mostly from Smartphone users (which comprise only about 5% of the global mobile market today).

So while most marketers are focused on advertising, I wonder: who is really focusing on engagement? Market insight will not be delivered entirely through analyzing click through rates on mobile ads. We will need to engage people, and find ways to encourage them to provide feedback and ideas in a time and place that suits them - this is the promise of the mobile device survey and mobile market research.

A recent study is claiming that mobile web browsing is 30% slower than typical online browsing.

I'd like to learn more about this study... Most desktop users (in North America, anyway) now browse with high speed Internet. But many only access the Internet at work where they have T3 speeds (learn more about T1 vs. T3 here). On the other side of the spectrum, some still use dial-up.

On the mobile side, some users have 3G, and some only browse when their HTC TyTn finds a WiFi connection.

Smartphone users are increasingly allowing their devices to be absorbed in to their daily lives. In particular, a new study shows the mobile web is popular with commuters. I'm one of those who browses on the way to work - I regularly browse mobile cbc.ca and The Onion Mobile, among other small screen friendly sites.

But The Onion recently eliminated their nice clean text only WAP site, and now forces me to download images. I could set my browser not to download images, but other mobile sites are more sparing with their use of jpgs, and I prefer that because like all consumers, I'm impatient but still want a good experience. I can only get through about half the content that I could before in the same amount of time, and I wonder if the flashier site is worth it.

Perhaps they are just staying ahead of the curve, but it's frustrating. And when users get frustrated, they go elsewhere.

What I'm really wondering is whether the stat is much worse than 30% for most of the population, given how bad mobile browsing is on some phones.

It just made me think about the implications for mobile market research. We need to make sure we present our mobile survey participants with fast-to-load and easy-to-navigate surveys, otherwise they're not going to participate in another cellphone survey until the wireless web catches up with the regular web for all users in every city on every phone.

I recently wrote about the Wild West of mobile devices - that was within the context of self-completed surveys and survey software on respondent devices. But what about mobility in terms of more traditional CAPI / MCAPI for a mobile device survey?

Netbooks and their equivalents are some of the only hardware projected for sales growth this year, and when I look at some of these entrants, I can see why. Check out the CTL 2go. Windows and a touch screen for $499? It even has a handle and is semi-rugged (tested for 50cm drop)! It's almost as if they had interview research in mind when they built this thing.





The popular Asus eee will also soon be available in a touchscreen..










Some other cool devices like this super keyboard computer can be found at PC World's "10  Most Anticipated Ultraportables and Netbooks of 2009" (Thanks to Mike for the forward)

I want to expand on my recent post about viability for wireless surveys on respondent devices...

People carry their mobile phone with them just about everywhere they go, and they increasingly view their mobile device as more than just a phone. The landscape is ripe for survey conducting anytime, anywhere, so why hasn't mobile market research completely blown up yet? 

If you take the cost of wireless data out of the equation, then the answer largely lies with the vast array of devices on the market.

Compare the infrastructure question to telephone surveys. Every landline phone transmits voice - it's the main purpose of your home phone. You have decisions to make regarding your CATI software, RDD and sampling strategy. And getting people to pick up and stay on the line and talk to you is a problem, but when you place that call, the phone on the other side will work no matter what brand of phone is in use.

Now move to the Internet... On the web, you have respondents choosing to interact with you through Internet Explorer, Firefox, Safari, Opera, and now Chrome. Your email invitation might get filtered out as SPAM, but there's a relatively small number of permutations that your survey software has to deal with when it serves up questions and answers on a computer screen.

Executing on a cellular survey provides a completely unique challenge. The number of devices, operating systems and even capability within a company's product line are staggering. Add to this that some users might only be able or willing to respond to you by SMS, not over the web, and it becomes clear that choice in the mobile marketplace is a problem.

Corporations don't develop with the notion that compatibility with their competitor is good. In the ever-changing world of mobile devices, they just want to get the next handset out, and get it out fast. Things don't always work as you'd expect...



I wish I could link to the source, but hearsay will have to do for now. At a recent wireless summit, a prominent panelist mentioned that when Transformers the Movie came out, they wanted the accompanying wireless app to work on every mobile device. It took 20,000 versions.

So which methodology do you choose if you want to reach respondents?

Do you go with a strategy that takes advantage of the iPhone, which sold 3.8 million phones in the first quarter this year, or do you choose an application that has been customized to excel on a wide variety of devices, or do you go with SMS - something everyone can use, but limits your research options?  It's a tough choice, because your sampling strategy needs to be considered in tandem with your method of data collection.

Will the market eventually show convergence, or will mobility follow the laws of entropy? Only time will tell...  

CASRO's popular Annual Technology conference is coming up. With budgets being slashed, you might not be able to get on a plane at the end of May, but if you're interested in mobile survey software and mobile market research, then try to catch AJ Johnson's talk "Avoiding the ‘Hang ups’ - Understanding Mobile Phone Self-completion Surveys".

Learn more about the conference here.

A recent post on this blog linked to some interesting stats:

• Worldwide mobile phone penetration continues to climb at a break-neck pace, with 4.1 billion mobile subscribers at last count (that's a global penetration rate of 61.1 percent).
• 1.27 billion fixed line subscribers (18.9 percent global penetration)

In areas like North America where traditional telephone research is a hefty percentage of the data collection that gets done, researchers might have a heart attack for the implications on their sampling plans! It's tempting to think that maybe moving to cellphone survey are just around the corner to relieve our response rate woes! 

Being at a company that specializes in wireless surveys, we're in a very exciting time. But I also have to have a dose of realism - Not all phones are created equal. 

In this story about Smartphone viruses, I learned that Smartphones currently make up about five per cent of the total mobile market, and the most popular smartphone operating system, Symbian, has 64.3 per cent of the smartphone market (3.2 per cent of the entire mobile market).

The line between smartphones and feature phones is blurring with the entry of WebOs, Android and the like. But have you ever browsed the mobile web on anything other than a Palm, Windows Mobile, iPhone or Blackberry? It's miserable.  My Samsung SPH-a920 is awesome - I'd recommend it to anyone, as long as you don't brwose the mobile web.

Setting aside connectivity problems for a moment, and just thinking about the hardware people own, wide acceptance of a mobile phone survey via the mobile web is probably further away than we would like to admit. Perhaps survey applications, text msg surveys and IVR will be our best options for doing a cellular survey in the short to mid term.

I recently wrote a post about a client who asked me about Palm, Market share of mobile devices, and what device to choose for their PDA Survey Software. Hopefully that helped spur thoughts on the state of the market today, but what's in store for the devices researchers will choose for tomorrow's Mobile Research?

In September 2008, professor Randall Stross from San Jose State University did an interview with the CBC where he noted that devices are getting ever more capable.

"There's no question that Google sees the biggest opportunity for growth is mobile access to the internet. Phones are becoming ever more capable of performing the same work that we use our desktops for. There's a whole new category of phones now, beyond the smartphone, that is designed from the ground up to interact with web services"