Wednesday, 21 May 2014

Innovation in our own backyard...

...or how a market researcher ends up attending a start-up competition

Pretty bold words. Whilst you might be hard pressed to find statistics to back up this claim, it is undeniable that there is a thriving community and support network in the North East which allows those with good ideas to flourish. This community was out in force yesterday at Campus North to watch 10 start-ups pitch their ideas at the Thinking Digital Conference Start-up Competition. You can see my summary of the pitches at the bottom of this post.

So why was I there? Market Research has often been accused of being stuck in its ways, slow to innovate, and too ready to bury its head in the sand. Like the metaphorical oil tanker it is assumed it will take too long to turn, in which time all of the start-ups and tech companies in their speedboats will have zipped past and beaten us to our destination.

As you might have guessed, I don't believe this. I think research companies, start-ups, and tech companies can co-exist and actually their combined forces can sometimes present great opportunities. I'm not the only one (read this great article by Allen Fromen on why Big Data will never fully replace MR here).  My role means I have start-ups as clients, suppliers, and as partners. I think if Market Research companies avoid the tech and start-up community for fear of being marginalised, they are limiting the opportunities where they can benefit, but also where they can give back to the community. So I thought I would pop along and see what was being pitched. Here are my thoughts on their ideas (in order of appearance):

Bountiny – The Amazon for Fashion
In a nutshell – Buying fashion online is a pain*. Rather than having to open multiple tabs and browse around various different sites, what if you could find everything in one place through an aggregator and go through a single simple online checkout? Add to that social sharing and searching of looks and you have Bountiny's vision.
+ I could certainly sympathise with the idea of too many tabs even if I don't buy fashion items. The idea of the crowd-curated looks if done right could be a winning feature, as it gives more flexibility to explore looks rather than forcing the consumer to think in terms of rigid categories such as tops, blouses, dresses, shoes etc.
- I think the implementation may be difficult and cumbersome. For it to be successful there needs to be broad support from retailers as if there are a couple of key ones missing people will still have to go to their sites directly thus undermining the one platform system. I think some retailers may be wary of the system, and reluctant to relinquish so much control. After all, it's only a couple of months since M&S took control of their online store from Amazon.

*This is what we were told, as you might have guessed my personal experience of this is limited.

Rormix – The Pandora for unsigned music videos
In a nutshell – Rormix promises to be a discovery platform for music videos for unsigned artists. The aim is to help users discover new artists, help new artists get exposure, and inserting “billboards” into videos opens a potential new revenue stream.
+ The ability to discover new bands based on commercial likes will appeal to consumers. They have an existing agreement with networks of gyms which could give the product a launch platform.
- The model seemed a little unclear and confusing and given the bureaucracy of licensing issues and the blurred line between signed and unsigned artists, it could be difficult to get the momentum this needs. Artists may be unwilling to relinquish control and so the platform may struggle reaching critical mass.  Assessments of attempts to insert advertising into virtual worlds have been somewhat underwhelming and so brands may not buy into this, and artists may not like the idea.

icanmake The shapeways for educational content/iTunes for 3D printed educational content
In a nutshell – They plan to create a marketplace for educational content for 3D printers. They have been busy laser scanning museum objects and aim to sell downloads and kits for people to use with their own 3D printers (or those in schools) and produce a book with Wiley to get people engaged with 3D printing.
+ Very admirable cause, plus their presentation included a picture of a ZX81 so bonus points there. The idea is a good one and could help broaden the adoption of 3D printers.
- It is a difficult market, hard to protect from copycats and there are a lot of question marks outstanding about the implementation. Changes to the 3D printing landscape outside their control over the next few years could skyrocket the brand or cause it real trouble.

Fonesense – The vouchercloud for ringtones
In a nutshell – Brands want advocates, consumers like offers. Let brands hijack your ringtone and everytime you get a call they get exposure and you get a deal.
+ The potential appeal for big brands was pretty clear. Also consumers like offers, give them the right incentive they might bite. The potential charity donation was an interesting spin on the idea.
- The idea opens a can of worms about data privacy and trust. Whilst it works for some big brands, it wasn’t clear how the model would work for new brands or smaller brands, and I’m not sure the appeal to customers will be enough to download and use the app. Moreover my phone lives it life on silent, and it wasn’t clear how the single solution would work for people who use their phones in very different ways. Given that mobiles often go off at the most inopportune times, there is potential for the branded ringtones to become associated with annoyance, and no brand would want that.

Wibbu – The personalised Duolingo
In a nutshell – Everyone is different, and people from different countries learn other languages in distinct ways because of the discrepancies between their native tongue and the dialect they are trying to learn. Rather than a generic offering like Duolingo, Wibbu offers content which is tailored to the individual the mistakes they are likely to make based on their mother tongue.
+ As someone who has learnt a few languages the more tailored approach is very appealing. Correcting the most common mistakes for leaners from your country is a good starting point, coupling this with personalised practice based on your mistakes gives a much better feedback system that should help you learn faster. They had a clear plan in terms of which markets to go after and why, and a unique selling point (content style) that is hard to replicate.
- The vast amount of content needed for a tailored approach becomes a huge burden, which is one of the reasons why currently Wibbu is focusing on Spanish speakers learning English. The difficulty of scaling the approach could limit the success of the idea. Convincing consumers to go for this approach when there are already so many options out there, could be difficult. – The soundcloud commenter for video
In a nutshell – Reframed allows commenting on videos in situ ensuring all comments are linked to the video timestamp, offering a social experience of video whenever you are watching it. The platform also provides you with a catalogue of your favourite moments in time from the videos you have watched.
+ The ability for third parties to utilise the system for their content increases platform use and could be very useful for content creators. The vast amounts of data the system could generate would be really useful to publishers if handled correctly. Also Reframed already have an agreement with Channel 4 News which will help them road test and refine the platform. If the favouriting of timestamps is implemented correctly this will give a compelling reason for people to use it.
- I don’t think the platform is going to replace generic video consumption via YouTube etc. and am uncertain if there are compelling enough reasons for the general audience to adopt the platform. I can however see how it would be useful for longer form videos and specific types of content (conference videos). I’m also not convinced that this solves the problem of “spoilers” and the experience far from being uniform people coming to video at different times will see very different comments and there is a danger of the videos being overwhelmed by spammy comments. Whilst I think it will be successful, I doubt it will become pervasive, but I would love to be proved wrong.

Audiowings – The Google glass for your ears
In a nutshell – Wearables are big, headphones are the most worn bit of tech, let's make headphones smart.
+ Love the name and branding. This was one of the most confident and polished presentations. I can see how something like this would have broad appeal if done right at the right price.
- Despite the confidence, I wasn't quite convinced. Perhaps rather unfortunately the headset was rather reminiscent of those awful cheap mp3 player headphones of yesteryear and so they didn't quite look the part. There were also question marks over how easy it would be to protect the idea. I also wasn’t convinced that the user experience would be a vast improvement over readily available alternatives. Finally it may be silly but smartheadphones isn't catchy and smartphones is already taken.

All 4 coaches - The Excel/Tableau for football coaches
In a nutshell – A personal assistant for football coaches, that allows them to instantly collect, store and analyse match data in a way that makes sense
+ Makes data instantly accessible, simplifies the collection and analysis process. The system is structured specifically for match data making it very useful for this one purpose and it looked very slick.
- Not sure as of yet how broad the market is and what the revenue potential is. The app would clearly be desirable but given the (relatively) niche market pricing could be an issue. Expanding to other sports will take significant investment given the specialist nature of the tool.

Formisimo – The Google Analytics for forms
In a nutshell – Their platform makes forms smart so that you can understand failure to convert, where the pain points are for users, and even potential intercept them in real-time
+ A simple solution that has a very broad potential audience. Sounded simple to implement so priced right the adoption could be rapid.
- The idea is hard to protect and open to competition from copycats. One of the judges asked the obvious “what’s to stop Google from doing this”. Whilst Formisimo are confident they can fend off the competition, it could be a fiercely competitive market in years to come.

Zipcube – The airbnb/ for meeting rooms and event spaces
In a nutshell – A one-stop shop for finding meeting rooms or event spaces, checking their availability and booking them instantly online.
+ Simple idea, simple interface, and a very flexible timesaver. It is probably the right time for an idea like this to take off. Also the potential for the system to be used by companies internally was an interesting idea and could be a good revenue stream if pitched right.
- Needs to have wide support and a good selection of venues before it can reach critical mass and be truly useful. There is the threat of competitors entering the space.

Judge's choice – Zipcube
Audience's choice –

Both of these will go on to pitch at the Thinking Digital conference today

Whilst the winners weren't surprising, both were in my top 4, honourable mentions I think should go to Wibbu and Formismo.

As someone who has previously learnt French and Spanish academically and is currently trying to learn German via Duolingo, I can see how Wibbu's highly tailored and individualised approach could leave them miles ahead of the competition and the tailored content means they are putting up barriers for competitors to challenges on their own terms (although this resource drain is also one of the big drawbacks).  Formismo's idea was simple, but I could see how it would benefit a lot of companies and how it might be readily adopted. Their biggest threat was copycats but they seem confident they can stand their ground.

Overall the calibre of ideas and pitches was very high and I'm looking forward to seeing who is crowned champion.

Update - Zipcube were announced as champions at the TDC conference

Dan Siddle 
Research Innovation Specialist

Monday, 31 March 2014

Nothing to declare


In a world with increasingly sceptical consumers, understanding how to build consumer trust becomes essential. Many studies have attempted to understand how professionals give advice when they have a personal interest in giving biased advice, and how it is interpreted and received by customers. It is common in certain professions, such as medicine or finance, for it to be a requirement for any such Conflicts of Interest (COIs) to be declared. Unfortunately prior work has shown that such disclosure doesn’t always have the desired effects. For example consumers often fail to sufficiently discount conflicted advice, and advisors can feel justified in offering more biased advice once a conflict has been declared [1].

A new set of experiments by Sunita Sah and George Loewenstein, published in Psychological Science, suggests that there may be hope for a positive effect of COI disclosure [2]. Previous studies have focused on the effects when COIs are unavoidable, whereas the current study also included conditions where advisors could choose to avoid the Conflict of Interest.

The basic setup for the experiment involved participants acting either as advisees, who had to guess the number of filled dots on a grid but could only see a subset of the grid, or advisors who could see the whole grid and give the other participants advice to steer their estimates. In the first experiment, advisees were rewarded for accuracy in their estimate, whereas advisors were paid more if the estimate was higher than the true value, introducing a Conflict of Interest. In this scenario the classic findings were replicated with advisors who had to declare a Conflict of Interest giving higher (and more biased) recommendations.

In their second experiment, advisors could choose between a reward structure that would create a Conflict of Interest (they received a $10 reward for an outcome that differed to the outcome which saw their advisee receive their $5 reward) or one which did not (the outcome that was beneficial for the advisor and advisee was the same and each would receive $5). 63% of advisors chose the higher COI reward when the reward structure wasn’t going to be disclosed to advisees. However only 33% chose the conflict when they knew the structure would be disclosed.

In their final experiment, a voluntary disclosure condition was added, meaning some advisors could now not only choose to accept or reject the COI but also whether they would disclose the presence or absence of COIs to advisees. Voluntary disclosure seemed to lead advisors to avoid the COI and disclose their freedom from conflicts, i.e. that they had “nothing to declare”.

It seems then that whilst the disclosure of unavoidable Conflicts of Interest may not help consumers reach optimal decisions, it could act as an incentive to advisors to avoid COIs and declaring the absence of conflicts could even act as a signal of quality. These findings give hope that in the longer term COI disclosure could have positive benefits by encouraging low-quality providers to improve quality or exit the market.

Dan Siddle 
Research Innovation Specialist

Thursday, 20 February 2014

Winter Olympics Sports Selector

If you want to see what your event is Click here EDIT 24/03/14  - Survey now offline

To find out more about the study email

Monday, 16 December 2013

Blink and you'll miss it

We like to think that we absorb most of what goes on around us, however our ability to attend to things can be remarkably limited. That's why when we watch the latest blockbuster most of us fail to notice when an apartment number changes [1], a cup switches colour [2], or even when a previously dented car is driven away scratch-free [3].

The fact that we can miss these things is not new, but it still surprises us when it happens. We are constantly bombarded with a barrage of sensory information. It is our fantastic ability to automatically filter out most of this information and to selectively attend to the small fragments of this stream that make our experience of the world so effortless. However it is what also makes us particularly susceptible to misdirection used by tricks such as those featured in Derren Brown's: The Great Art Robbery shown last Friday.

A widely researched phenomenon in the psychological literature is the Attentional Blink [4]. It occurs when individuals are asked to identify 'targets' during the rapid presentation of stimuli in series. Identifying a target makes individuals often fail to recognise a second target presented 180-450ms after the first one. A recently published paper in Psychological Science suggested that a similar process may underpin the failure to detect multiple targets in a self-paced search task [5].

So what does this mean?

Real-world searches quite often involve the need to (potentially) identify multiple targets. The authors give the examples of radiologists searching for tumours, and airport attendants scanning luggage. Failure to spot a target in either of these cases could have serious consequences. Understanding the presence of this effect could help put procedures in place to tackle the problem and increase identification rates.

In the consumer world packaging vies for the attention of shoppers on the shelves in stores.  Whilst companies have limited control on the ‘procedures’ consumers adopt, understanding the effect could help optimise packaging.  For example it is known that faces can escape the Attentional Blink [6].  Recognising the effect could also be important from a usability point of view.  As designers seek to optimise the flow for users, knowing what affects the fluency of navigation matters. Of course in both situations there are a variety of other factors at play. Nonetheless, it seems obvious that a better comprehension of how we attend to things should enable us to improve almost any engagement opportunity.

Dan Siddle
Research Innovation Specialist

Friday, 11 October 2013

The price is right

Pricing has always been a difficult thing to get right, make something too expensive and you may struggle to sell enough of it, make it too cheap and it can be hard to make a return. In recent years work on heuristics and biases have helped explain some of the quirks or seeming irrationality in the way prices influence the decision making process. The usual behavioural economic suspects have written extensively on various aspects of the subject.  That said, there are still big gaps in our knowledge and the technology-led diversification of payment structures means we need to expand and re-evaluate what we thought we knew.

Two things that have emerged over recent years are the growth of Pay what you want (PWYW) pricing systems and crowd-funding platforms. The former allows individuals to decide how much they are willing to pay for a good or service. Whilst not an entirely new concept, the increase in digital delivery of content such as video games and music and the sophistication of online payment systems have made this approach increasingly common. Nonetheless, the most well known case of this is probably still Radiohead’s release of In Rainbows back in 2007. Crowd-funding projects allows individuals to pledge money to things that they want to see become a reality. In 2012 Amanda Palmer took this approach to fund the production of her next album (and art book, and tour). Having asked for $100,000, she quickly accumulated almost $1.2 million in pledges and about as much controversy as Radiohead did in 2007.

When and why these mechanisms work (or don’t work) is a rich area for continued enquiry.  A study recently published in Jena Economic Research Papers suggested that anonymity could have an important effect on the success of PWYW schemes [1]. They analysed Magnatune sales data during two periods from 2005.  During this year Magnatune made a change for a brief time whereby they told customers on the payment screen that the artist would be given their name and told how much they had paid. By contrasting sales figures from these two periods, the researchers concluded that reducing the anonymity led to a non-significant increase in amount paid, but a significant drop in purchases made, leading to an overall drop in revenue of 15%. Of course the research uses data from 2005, so whether this behaviour of what probably could be considered “early adopters” would match the wider audience, it is hard to say.

Personally I’ve always been fascinated by the Humble Bundle statistics. Humble Bundle operates a PWYW scheme for collections of independent games. The first collection was released in 2010 and since then the series has successfully generated not only additional collections of games but expansion into eBooks and music.

Since the very first bundle they have published aggregate statistics on the payments made for the bundles [2]. They show the total contribution and average donation by operating system. Although the total contribution has generally been dominated by those using a Windows Operating System (due to volume of sales), the average price paid by those using Linux is consistently higher.

Does the fact that the average amount contributed by those on Linux exceed those on Mac and Windows Operating Systems indicate something about the sort of people who use Linux? Or is there some kind of grouping effect? Maybe the visibility of the stats affects the amounts those using different Operating Systems are willing to pay, or initial availability plays a role? It is hard to say, but certainly raises interesting questions about willingness to pay. Making something free can lead to over-consumption and make us undervalue superior paid alternatives (see for example "The cost of zero" in Predictably Irrational). However if anything Linux users are less used to paying for the software that they use. In fact the idea that people may become used to paying nothing for software is often an argument given against freeware and the open source movement.

When it comes to pricing we certainly don’t have all the answers and not all prices or situations are created equal. What works in terms of PWYW and crowd-funding schemes is still the subject of a great deal of debate – currently well over 50% of kickstarter projects have failed [3]. However as new types of transaction emerge and grow in popularity, it provides the data and opportunity to improve our understanding of the decision making process and I'm sure will generate plenty of interesting new findings.

Dan Siddle
Research Innovation Specialist

Monday, 23 September 2013

All ears

Music can have incredibly powerful effects on us. Not only can it touch us emotionally, but it can almost effortlessly evoke quite vivid memories and associations. Using melody can even be an effective tool for learning text [1] as any How I met your mother fans will be aware (Apple Orchard Banana Cat Dance 8663).

It is therefore little surprise that when BETC were looking to celebrate 30 years of Diet Coke, they used the Etta James track I Just Want To Make Love To You to evoke memories of the iconic nineties ad and give a nod to the brand's past. Getting the music right, can do wonders for the brand (and song).

In fact, advertisements have often launched songs and artists into the charts. A Little Less Conversation by Elvis Vs JXL rocketed to number one in the UK after featuring in a Nike advert. Would Mr. Oizo's Flat Beat have made it to number one if it wasn't for the association with Levi's? I doubt it. There are also plenty of bands who probably owe Apple a thank you for being selected to appear in their iPod ads (The Caesars, Jet, The Ting Tings).

Sometimes songs become so closely connected with advertisements that it seems odd to hear them when the ad is not playing. I still struggle to hear Adiemus' Cantilena without seeing a kid retrieve a pearl form the bottom of the ocean [2], or Leftfield's Phat Planet without seeing surfers rush into the sea in black and white [3].

I was therefore surprised when, the other day I heard M83's Outro playing from the kitchen, and wandered in to discover not Red Bull on the TV, but an advert for Persil:

It seemed odd for such a big brand to have used a song that was featured in an advertisement so recently, especially as both ads rely heavily on slow dramatic visuals to accompany the music. Part of the successful use of music in an ad has got to come down to a brand being able to “own” the music in the mind of the consumer. Although these aren't competing brands, the recency of the Red Bull ad seems likely to create noise for any encoding of the Persil advertisement. For completeness here's the Red Bull ad World of Red Bull:

Maybe it's coincidence, but this isn't the first time I've noticed re-use of music recently. In fact here’s another example with another song by M83. It features in the Renault advert for the Captur:

And was previously used in Gucci’s ad from mid 2012:

Even Samsung are at it with the new S4 ad:

The ad features the third movement from Vivaldi's Four Seasons: Summer. A clip that has previously been used for the HP Touchsmart, Homme Exceptionnel and as recently as October 2012 in a Next advertisement (although technically the music in the ad below is Vanessa Mae's Storm).

It seems crazy that brands who are fighting to put a unique stamp on a fairly homogeneous product type (such as laundry products), would choose to use a song that had been claimed by another brand. Surely in a world where it is increasingly easy to discover a wider variety of different sounds and new bands, the opportunities to "own" a piece of music are greater and the potential benefit of this for the brand huge.

What do you think? Are there any iconic songs from ads that are forever associated with a brand for you? Have you stumbled across any other songs that have been used in multiple ads?

Dan Siddle
Research Innovation Specialist

Monday, 9 September 2013

Going Viral

Image courtesy of KROMKRATHOG /

Creating and sharing content is easier today than at any point in the past. This makes it both an exciting and challenging time for content creators. Ensuring that your creation cuts through the clutter and spreads is no easy task. Indeed it isn't hard to think of campaigns that failed in their attempts to go viral, or videos put up just for fun that have become worldwide phenomenons. If only there was an easy way of creating, or at least identifying, a sure-fire hit.

Well researchers at the University of California think they might have unearthed one secret in the mystery of what we tend to share [1]. They found that neural activity in the temporal parietal junction (the bit in red on the diagram below) was more pronounced amongst participants that did the most compelling job of recommending the ideas that they liked.

In the study, 19 students were tasked with the role of acting as “interns” which involved being content curators reviewing ideas for TV pilots and then recommending which shows should be considered for further development. They were filmed describing the ideas, which were then shown to their bosses, the “producers”, who were other students recruited for the study.

The researchers suggested that the activity they observed in the ventral striatum (part of the reward system), could index the appeal of the idea itself or possibly represent the expected value from sharing the idea. They proposed that activity in parts of the mentalising system, such as the temporal parietal junction, related to the effectiveness with which an individual could communicate an idea. Both of these elements would be essential in the successful propagation of ideas.

Of course with the cost of content creation so low and the cost of using fMRI so high, it is hard to see advertisers rushing out to test all of their material to see if it lights up these sweet spots in the brain. It does however open up interesting new avenues for exploration and raises the question as to whether analogous measures that are easier and more cost-effective to monitor could be found.

It is still early stages. The activity gives us some idea of what we are likely to share, but doesn't shed any real light on why. There remains a great deal to be learnt about propagation of content as opposed to ideas, and the spread in more complex social environments. For example, the study created an environment analogous to that of a worker sharing ideas with their boss, so conclusions about how we share amongst our peers must be drawn with caution. Nonetheless, what goes viral and why is an interesting question that is being investigated through a number of disciplines and I'm sure it is a journey that will fascinate us for years to come.

Dan Siddle
Research Innovation Specialist