Live blogging from the Net Gain 2015 conference in Toronto, Canada. Any errors or bad jokes are my own.
What’s Hot and What’s Not Hot: Ray Poynter, Director of Vision Critical University
- Ray’s book are for sale at the back of the conference. Find him and he will sign your book! (Yes, please!)
- What is still hot?
- Mobile is really big and that’s why Ray has written a book on it [buy it 🙂 ]
- Why is CATI so big – in this room, most people do NOT answer the landline in their home. Mobile used to cost more. Not sure if the person will be driving when you call a mobile phone. Hard to geographically target mobile phones like you could RDD.
- PEW research is top notch CATI probability surveys. It is the majority of what they do and they have just recently bumped the percentage of their calls that is mobile.
- Online surveys – 30% are attempted by people on mobile. Some people KNOW they are doing mobile and others don’t. May be 50% in just a couple of years. But only 15% of surveys are suitable for mobile devices. Most surveys are not optimized for mobile. Not thought about wording or question types. Not even checking the data to see if mobile vs laptop data are different.
In 950 Tesco stores, they do surveys on tablets with geolocation, datestamp, etc.- Heineken did a beer audit in Africa. Recruited interviewers, gave them a smartphone. Phone made SURE every location was geotagged. Photos of every location. Quality of data was far superior.
- Communities
- Companies doing so are beginning to disappear because communities are more mainstream. Everyone has their own community.
- DIY is enormous in society. DIY travel, DIY bank machines, Uber, AirBNB, ZappiStore.
- DIY has spawned automation. If every idiot can write a survey, they will. So let’s make it safer.
- SurveyMOnkey is the biggest survey platform in the world.
- To be hot, it must be scalable and it must work – NPS doesn’t do this. 🙂
- DIY isn’t great with efficacy. There won’t be many neuroscience for dummies books in the near future.
- What is HOT right now
- In the moment – Ask the breakfast survey the very second you finish your breakfast. Survey about the hotel registration before they open their hotel room door.
- Location Based Research – Put a geofence around a starbucks so you know who walk in or out. This also attracts aggressive marketers, not just researchers. So the message on your phone could be a survey or a sales pitch – ShopKick. Do they turn on the microphone on your phone? Do they turn on your camera? Do they tell you they have done so?
- Microsurveys – RIWI, google consumer surveys. Usually 1 to 3 questions. Google is up to 10 questions. Won’t tackle your problems that have a high dollar value associated with them.
- Automation – Automate reports as well as research process. What do we add to this? What do we add to the trends? What canNOT be automated?
- Always choose the simplest tool – don’t need to take a picture of every window and find software to count those pictures. [sounds stupid but really think about it]
- What’s bubbling new and exciting
- Text analytics – sentiment analysis is getting better for all except twitter. much better for emails and letters to companies, comments on youtube, inbound call centers, which letters are genuine sales leads or complaints or bomb threats, reaction marketing.
- Web Messaging – Whatsapp, WeChat. People are doing less talking to everyone and more talking to individuals. In comparison, whatsapp grew WAY more quickly than facebook and twitter. This is massively scalable. Panel companies will go this way. [They already are!]
- ResearchBots – Processing time and moderators takes a lot of time. New things don’t work all the time and that’s why it’s bleeding edge. Not very scalable at this point
- NOT so hot –
- Facial coding – good with an extremely experienced trained person sitting in the same room. Via webcam isn’t quite so good. Fully automated is very clever but delivers almost nothing. Software can identify specific pictures but a human must still go and interpret all those pictures. Great for assessing people’s reactions to packages. Not a general purpose tool. Doesn’t suit most research problems.
- Webcam Qual – You don’t want to take video from home because you still have to brush your hair and change out of your pajamas. Webcam on the bus means everyone behind you on the bus sees the images too.
- Social media research – We thought it would destroy MR but it’s really a niche. Most research teams have scaled back on this. Maybe using tweets only. not used so much for insights but more for reactions to advertising campaigns. Social does answer questions not asked. Social usually doesn’t answer your specific research questions. Vendors often say “I agree it has under-delivered but my company is doing it right!”
- Social media 2.0 – integration with marketing, integration with survey research, integration with tracking, interrogative.
- BT Case Study – Net Easy – how easy is it to work with BT was a better measure than NPS. They looked online for people talking about ease or difficulty and responded with solutions. Achieved a 3.5million reduction in costs by doing this. 600 000 people who would have called a telephone were able to DIY from the website.
- What about passive data, gamification, biometrics, wearables, quantified self, Internet of Things, single source, neuroscience. There is too much stuff to register the quality of everything. You can’t learn it all.
- Gamification doesn’t solve a lot of problems but it HAS made us rethink what we’re doing it.
- Behavioural economics is really efficacious but it is incredibly specific.
- Passive data from phone recording everything you press and everywhere you go. Won’t see big movements here. It will be mostly qualitative.
- Big data is beginning to move but predictiveness is limited right now.
- Wearables – sharable is great but these people are not yet representative. Mostly qualitative and very targeted.
- Geotracking – very tiny right now, works well in qualitative. Can draw maps of where individual people went. Mapping ebola is a different story – limitations of cell phone towers in other countries makes it impossible to map journeys in small locations.
- Internet of things – only exists in minds and publishers right now.
- Single source – Means tying together many data sources, it’s a power battle, a methodology battle. WHO is the single source? The telcom? A research company? Privacy battles of combining data.
- Top 2 Things to think about.
- Mobile – traditional, in the moment, multimedia, passive
- Integrative and participative – 360 panels, databases, communities, social, mobile, qual, collaborative all together
- “We will always need faxes” “We will always need horse and buggies” ….. We will NOT always need surveys. Ray thinks no more surveys in 20 years – classic 20 or 30 minute surveys. Suspects only 33% of spend will be on surveys by 2019.
- We need to redesign our ethics – most of our ethics were established 60 years ago mostly by men, all of them white, and most of them dead
The LoveStats Blog 
Can’t wait to prove how wrong Ray was about “Microsurveys”. I’ve never done a survey 1-3 questions and my last “microsurvey” was a 160 question modularized and randomized behemoth. Thanks for blogging Annie!