The following are notes that I took live while Fabien Rolland, Director of Marketing Research, Aeroplan Michel Girard, Director, Analytics at Aeroplan 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.
Outline:
- look at Big Data analytics in practice
- Compare volume of data and value of data
- Evaluate the pros and cons of Little Data versus Big Data
- Examine the impact, opportunities and challenges of big on research practices
Want to see how/if Big and Little data can be merged within the same organization
Aeroplan — is the focus of AIMIA company, 5 million members partners with more than 150 brands globally.
What is Big Data — The Three Vs?
- Volume -Tera/Peta bytes, Transactions, Tables
- Velocity — collecting and reporting near/real time
- Variety – structured, unstructured, multiple sources
Close Neighbours but Different
- Customer centric model
- next level: demographics, model scores, transactions, touch points
- then: attitudes, satisfaction, share of wallet, lifestyle
- outer circle: syndicated data, neighbourhood data and socio-demographic segmentation
Shape a relevant, personalized experience via member-centric communication
- organization has moved from building models and segmentations schemes to targeted marketing, and then finally 1:1 marketing
Vision for Change: The Analytics Roadmap
- reduce pressure on analytics team with the creation of a self-serve access to data (velocity)
- centralize member touch points data (volume variety)
- transaction from “reactive” to proactive agents for change (velocity)
- doing this represented a change in mindset
Traditional MR is under Pressure
- Promise that Big Data will improve decision making is creating expectations among users, but this may be overblown.
- Research is seen as being a very long process to insight, we cannot afford to get to the answers in a long period of time — decisions need to be made faster and faster.
Other challenges:
- Expensive
- Slow
- Exclusive
- One off
- Self reported
- Not reliable
- Not accurate
- So what?
- Not what I was looking for
Little data is getting:
Better, faster, cheaper, bigger
Elements of research AIMIA using today:
- Online panels
- Sample sizes
- Online forums
- Text analysis
- Standard approaches
- Reporting
- Web intercepts
- Mobile QR
- Event triggered
- Location based
The biggest evolution in Aeroplan is in the technology side. Have tools to analyze faster, large amounts of data, cross tab and create reports in PowerPoint in a number of minutes.
AIMIA Proprietary Panel (100k members)
- Can find out how promoters differ from detractors
- Each day interact with members through many channels
- Can not just collect data, but also act upon them — like connecting complaints to call centres
Bigger isn’t always better
Little data can answer the following questions that big data can’t, such as:
- Which consumers should we target?
- Which products will be successful?
- What price promotions will be successful?
Problem: big data can be incomplete
- Members don’t always use card at a partner
- Doesn’t analyse competitive behaviour
- Big sometimes can be too big if they are desegragated or too difficult to mine
We can also understand the ‘why’ behind the ‘what’
- Who
- Where
- When
- What
- How much
But, big data cannot uncover consumer opinions and motivations, small data can do this.
- Can connect intentions with behaviour
Result: Little and Big data can complement each other
2009-2010 — moving big data to research
2012-2013 — integration of research and big data
2013-2014 — building data assets