Don’t Ask! Gaining insights without questions

Technology has made rapid advances this century allowing us to capture more data than ever before. This conference looked at how organisations collect non-survey data such as biometrics, passive mobile tracking and observation.
Clearly these types of data can potentially substitute more accurately for self-reported survey responses. However this could be intrusive and raises ethical issues which were discussed at the conference.
The conference also looked at how this data can be analysed on its own and with data from surveys and other sources.

Event Schedule

Have you ever stopped to wonder why personalised is still best done by a human? I mean, with all that data on who you are, and yet, a person can just “get you” in a matter of moments with only a fraction of knowing your claimed intent or behaviours. We call this the emotional conundrum. In our talk, we share the last 5 years of exploring the intersection of people and machines and how body language might just unlock intuitive artificial intelligence… or in human speak, empathy.

Location: Hall 1, Building A , Golden Street , Southafrica

A discussion on automated collection of biometrics at scale.

Location: Hall 1, Building A , Golden Street , Southafrica

Everyone in Europe has been forced to focus on data protection by the introduction of the GDPR later this month. This regulation contains requirements that will impact the way we collect and manage passive data, in particular whether we have appropriate consent for processing. There is still a lack of clarity in some areas about what is allowed and what is not. In addition, we are still waiting for a definitive text for the new e-Privacy Directive. I will do my best to explain the background to these laws, what we know, what we expect will happen in what is still an evolving regulatory environment and how you can get up to date help and advice.

Location: Hall 1, Building A , Golden Street , Southafrica

During a recent European Horizon 2020 research project (http://www.bison-project.eu/) MyForce got the opportunity to discover the amazing world of speech technology and find out how this fits into the big AI puzzle. This covers a lot of different aspects such as legal compliance, recognize people based on their voice, improve quality during ongoing conversations…
And this also offers a tremendous amount of possibilities when executing market research. During this presentation we will first of all explain how speech technology works. This covers elements as voice biometry and speech analytics. And this leads to new possibilities, insights, results, ideas… that can be used in the market research world. Think about obvious things: transcription of open ended questions, improved quality monitoring… but also conversational surveys using voice bots!

Location: Hall 1, Building A , Golden Street , Southafrica

The PASS panel survey is a major data source for labour market and poverty research in Germany with annual interviews since 2007. In January 2018, the supplemental IAB-SMART study has been started, in which selected respondents were asked to install a study app on their smartphones.

The IAB-SMART app combines short questionnaires that can be triggered by geographic location with passive data collection on a variety of measures (e.g. geographic location, app use). The triggering of questions allows us to enrich yearly retrospective data with data collected immediately after a certain event (e.g. placement officer visit). Passive data collection allows innovative measures, e.g. for social capital that complement traditional survey measures. Furthermore, the additional smartphone measures create the potential to address new research questions related to the labour market and technology use (digital stress, home office performance). Finally, the study provides new insights in the day structure and coping behavior of unemployed persons and thus replicate aspects of the classic Marienthal case study with modern means.

In this presentation we will provide an overview of the study and share our experiences in conducting an app project. We will focus on data protection issues, implementation of the fieldwork, participation in the study and of participation in short surveys.

Location: Hall 1, Building A , Golden Street , Southafrica

Mobile data has become a necessary element of transport analysis in recent years.  Its value and limitations have increasingly become understood, but perhaps never before in as difficult an environment as the whole of Greater London. Project EDMOND (Estimating Demand from Mobile Network Data) has been a major exercise, involving many new techniques to deal with the challenges of data development for such a complex city.

The Strategic Analysis department at Transport for London (TfL) is responsible for the transport models that are used to forecast future traffic and public transport demand, congestion and crowding in London. It also considers demand for cycling, walking, taxi and other travel. In Project EDMOND, TfL worked with Jacobs, AECOM and Telefonica to develop new demand data that will be used to help maintain and update TfL’s transport planning models and to provide broader policy insight using mobile network data from O2 fused with various other datasets relating to transport in London.  The mobile network data used was anonymised and aggregated to protect the privacy of the users.

The presentation will consider insights that were gained from the data and which led to techniques to improve its value.  In particular, this will emphasise the determined focus on understanding the biases and suitability of every data source used in the project, including a project requirement to develop a metadata statement for each source.  Two specific surveys were also undertaken to support bias appreciation and correction. These were:

  • Additional questions on mobile phone ownership and usage added to the rolling London Travel Demand Survey, a household interview programme with an 8000 annual sample.
  • An opt-in survey, which compared respondents’true travel diary records for individual journeys against the travel attributes inferred from the data processed, with permission, from their mobile phones.

The presentation will explain the unique perspective this analysis has brought to understanding the strengths and weaknesses of processed mobile phone data and the adequacy of the processing assumptions and methods applied. There will also be a brief discussion of the role of data fusion in the project to improve the results by use of other data sources, including separate surveys and models.

Location: Hall 1, Building A , Golden Street , Southafrica

respondi has been involved in passive metering for more than 2 years. In this talk we would like to share our experiences.

Since 2016, we have been continuously collecting web navigation data from approx. 7000 persons in France, Germany and the UK. To date, around 1 billion data points have been collected. It is already possible to draw lessons, on several dimensions, from the various experiences we have had.

  1. What do clients expect?

Which demands do we typically receive?  How do we answer them efficiently (time-, as well as costwise)? Which kind of demands can we meet, and which are not possible?

  1. Which usages per se?

Here, we will present some uses cases which have been run, based on passive data only. In particular, we will present some results from various research about online purchases (conversion rate, abandon rate, payment method).

  1. How do you combine with traditional declarative surveys?

The principle of this data combination will be described through this simple and interesting question: are heavy internet users happier? To answer this, we crossed the web navigation with the answers to a survey about their well-being feeling (inspired from the UN’s world happiness report).

We will then run through more sophisticated data science output (of course no technical knowledge required) about customer journey analysis (about the retail industry in the UK).

  1. How does dataviz fit in?

One of the main problems of web navigation data is the complicated data (huge amount, unstructured…). Dataviz is part of the solution as it helps us to understand, immediately and in a visual way, complicated topic, huge dataset, etc.

We will present our dataviz solution in saas access which  visualizes online behaviour of a precisely defined target (more details here: https://www.respondi.com/EN/beaterypublic )

  1. Final learnings about challenges and development of MR?

To wrap it up, we will share with the audience our best practices, the difficulties we still must address and improvements we have in mind.

Location: Hall 1, Building A , Golden Street , Southafrica