The following are notes that I took live while Melanie Courtright (ResearchNow) and Prince De, (Conversition) presented at MRIA Netgain 7.0. The notes were originally posted on my Tumblr account (paullongsblog.tumblr.com), but I have moved them here and revised them slightly to remove typos.
Background: Research Now used the 2012 Olympics as a data lab to learn how multiple modes of digital data collection can be combined to tell a more complete story about consumers.
Modes used:
- Web based surveys
- Social media listening
- Web behavioural data
- Mobile surveys and behavioural data
The ambition:
The key was to weave these unique data streams together — during the discovery process key learning emerged.
- General attitudes toward the games including opening and closing ceremonies
- With ongoing nature of social media commentary and longitudinal behavioural data, the team segmented out this point in time
As data analysis began a natural collaboration occurred between different modes, found information on one that could not be found on others.
- Web survey – develop hypothesis in advance
- Mobile — survey standardization and invitation timing critical
- Mobile app – app usage trends are more valuable than point-in-time snapshots
- Mobile search data – aggregating terms provide insights into topical subject
- Social media – data exists for all points in time, harder to answer specific question
- Web behaviour — high volume of data means you must formulate specific business questions
The Trials
Online and Mobile — compared favourably on consistency measures (both global samples)
Mobile skewed slightly younger, when weight on age and gender no differences on responses
Other general findings:
- When formulating an analysis plan using multiple data sources, first focus on business questions you want answered
- Focus on what modes you can use
The Olympics Story
The Behavioural Timeline
- Social media tells us the Olympics ran from July 27 to August 13, 2012 — accurate, from looking at the spikes in discussion about the Olympics. This is accurate
- Sentiment goes up for major victories/gold medal, drops for major losses.
- For U.S. social media data closely aligns with site visitation of www.nbcolympics.com and www.teamusa.org
- Opening Ceremony – the highest interest in watching the Opening Ceremony came from the host country (UK)
- Examination of mobile behavioural data reinforced this, as top visitations were to UK sites
- The Canadian market was active in voicing their opinions in the social sphere
The Games
- Data on sports of interest were extrmely consistent, when asked survey participants stated their favour events (track and field, gymnastics, diving, swimming) was consistent with social media.
- People didn’t generally mention which apps they were using, but behavioural data yielded this
- Star athletes (like Michael Phelps, Usain Bolt) had highest social media mentions.
- Locally, in Canada athletes like Simon Whitfield were mentioned quite often.
Sponsors
- In surveys, awareness low for all sponsors except for Nike (who wasn’t a sponsor). The same finding was found in social media
- Nike yellow shoes used in Track and Field drove the belief that they were a sponsor.
The Agony of Defeat
- Due to time zone difference, NBC chose to tape delay coverage
- NBC had very poor sentiment due to tape delay
Summary findings
- Each stream had helpful data, but using all together created synergies:
- Survey data Imobile and online) delivered answers to pre-deterimined specific questions
- Behavioural data enabled real-time monitroing of events and information gather activities
- Conversation content brough the clarity through the voice of the consumer