This talk will focus on experiences Ericsson ConsumerLab have had with using virtual reality (VR) as a means to conduct actual focus groups. Rather than bringing participants to a central location and exposing them to various VR tests, this instead means designing a focus group room inside of VR and then inviting participants who already have their own VR headsets to virtually attend focus groups sessions from wherever they happen to be physically located.

Initially, we thought that recreating focus groups in VR would be relatively straightforward, but there was a wide range of challenges, many of which were related to participant behaviour and the fact that there are no established social norms for this type of social setting in VR. Virtual body language also turned out to be very expressive, yet different from physical body language.

Recruitment of respondents was complicated as traditional agencies could not recruit the user base we were looking for. There were also technical platform issues we needed to work with, not so much related to the VR environment used as related to the users’ own equipment.

Finally, we will discuss the pro’s and cons with focus groups in VR and how they might be a very interesting complement to traditional face-to-face sessions going forward. Despite a low initial user base that limits the type of topics suited for VR focus group work, we believe cost savings and efficiency gains will make VR an important qualitative tool going forward.

Quantitative researchers work with lots of numbers, whilst qualitative researchers work invariably with lots of words, written and spoken. But what about pictures? Instagram – the social media platform based around picture sharing, has over 1 billion active users, uploading 100+ million photos every day, which receives 4.2 billion likes daily. As researchers we can’t ignore such an enormous phenomenon, the explosion of the visual - so we set about exploring it via a radical experiment to find out what would happen when we conduct a piece of mobile research replacing words entirely with pictures.

For our first experiment we used pictures for the questions, the participant responses and the analysis – cutting out all words. The responses we received back were full of personal storytelling, but they lacked one important element necessary for consumer understanding: context. This lead us to conduct a modified follow-up experiment, replacing regular pictures with “augmented pictures” – pictures with levels of embellishment, as widely created and shared on Social Media.

The findings were extremely insightful, and gave us some food for thought for the way we do research – a reminder that the research world and the real world are not two planets existing in silos. Consumers shouldn’t change their behaviour to market research requirements – it’s us researchers who need to adapt our methodologies to reflect consumers’ worlds. Whether that’s using pictures, augmented pictures or otherwise – in order to understand consumers fully, we must do what we can to get close to them, adapting our approaches in synch with changing communication habits.

It is no secret that the market research industry is under pressure to deliver sound and strategic insights within shrinking budgets and time frames. Ten years from now efficiency is predicted to be the number one deciding factor when commissioning research. This presents a serious challenge for qualitative researchers whose traditional methods are not compatible with this need for speed. Whilst demand for qualitative research may be growing, unless we find a way to adapt our processes there is the risk of becoming irrelevant. Confronted with this scenario, we have set out to discover if automation and AI in qualitative research – especially in analysis - is even feasible, and if so, what are the benefits and drawbacks for research professionals? What is the impact on the time, cost and quality of insights? Moreover, how do clients evaluate these trade-offs? To answer these questions we collaborated with Danone, devising a head-to-head competition between human analysis, machine analysis and a combination of the two. This resulted in three research reports, evaluated by Danone in relation to their business needs. From this experiment we learnt that the outputs from machine and AI analysis do not offer ‘magic bullet’ insights. Standalone, automation delivers little value and is limited in its sophistication. Yet, if viewed as a facilitator, automated tools can certainly be used to our advantage during the ‘human’ qualitative analysis process. Thus, by using these, a full report was produced in half the time versus a full report created by human analysis alone. Moreover, our client evaluation confirmed that the time and cost benefit incurred did not compromise the quality of insights. In fact, this report was selected as the preferred option prior to revealing the methodology behind each one.