SUNZ 2011 - Vince Moder - Loyalty - One dataset to rule them allPresentation Transcript
One Dataset to Rule Them All Vince Morder Loyalty New Zealand SUNZ 2011 Conference 24 February 2011 Te Papa, Wellingtom
Introduction Need for data to give a consistent, complete picture of the customer Data can be too fragmented across your organisation Too much data preparation for analysts Results cannot be easily translated to other products. Need for better integration of analysis into the business
Even the best analyses are useless unless they are used.
Analyses can have limited lifespan if not adjusted for the needs of your organisation and the changing habits of your customers.
Inferring a person from outside in Demographics, Occupation, Location, Family Structure Beliefs, Attitudes, Behaviours Knowledge, Enculturation Emotions, core motor skills Genes, Consciousness
Understanding customers using data and statistics Demographics, Address, Occupation Transaction, Account data Hard, Explicit Profiles, Models, Segments Data Surveys, Panel Groups Soft, Tacit Knowledge Information
Data, Information, Knowledge Data
Bits of unorganized and unprocessed facts
Data is a prerequisite to information.
Information can be considered as an aggregation of data
Information has usually got some meaning and purpose.
What resides in the minds of people in your organisation.
Used to transform data into information.
Knowledge is derived from information in the same way information is derived from data.
Knowledge Management (KM) Cannot define knowledge KM is different from Information Management The function of KM is to create a shared knowledge context.
Varies from org to org
Requires a cultural change.
KM is what you put into place to deliver value from your knowledge.
About Loyalty NZ
About Loyalty NZ Renowned for its Marketing excellence Recent awards: Asia Pacific and Japan HP Digital Print Awards – Direct Mail Award 2010 2008 TVNZ NZ Marketing Award - Consumer Services Gold AXIS Craft – TV/Cinema/New Media – Animation/Design & Motion Graphics Silver 2008 AXIS Craft- TV/Cineam/New Media- Visual Effects Bronze 2008
Fly Buys – “Dream a Little” Statistics of the Fly Buys Programme 14 years of shopping history 70 Partners (Participants) 1.2 million active households (70% penetration) 2.2 million active cards 1000’s of rewards Business Model Many ways for consumers to collect points, as the coalition of participants covers the full range of retail products. Strong retention. Participants pay LNZL for each point collected. The carrot is the reward. “Dream a Little” is ‘the one thing’. Cycle of usage and redemption. Leading innovator in the industry. Recognised from the start that the real value is from the data.
The Customer Insights Team LNZL Customer Insights exists to deliver our Participants with insights about their customers (and potential customers) to enable them to gain maximum benefit from their involvement with Fly Buys. CUSTOMER INSIGHTS TEAM DATA WAREHOUSE Fly Buys Member and Transactional Data ParticipantSKU Data External Data We do this through leveraging the power in the Fly Buys database by applying advanced analytics tools and techniques to turn data into actionable insight.
The Base Data Fly Buys holds
The Loyalty New Zealand Customer Insights team Loyalty New Zealand’s Customer Insights team is driven to provide compelling outcomes for Fly Buys Participants leveraging the very best data. This is represented in the vision: “Providing unrivalled levels of Customer Insights to drive outstanding outcomes” To enable this to occur, Loyalty New Zealand has invested significantly over the past two years to provide market-leading infrastructure and expertise. A team of 12 specialists in Wellington are focussed on extracting the right information and insights to support desired activities/requirements.
The Pyramid of Delivery
Every participant gets a monthly summary report showing their the volume of spend and points accumulated.
Monthly Dashboard Outlet Reports Spend volumes, # customers, and points issued by month for last 60 months.
Demographic Dashboard Report Distribution of income, age, segment, commitment of an outlets customers .
How to Best Organise the CI Team
Handling so much data
5 TB database
Hundreds of Millions of transactions a year.
The sheer volume of targeting campaigns (5 per week)
The sheer volume of analytical requests
…. And still keep developing and improving our services
How to Best Organise the Work
Redesigning how we do our work New DW and SAS Historically Raw Data (Loyalty Host) Raw Data (Loyalty Host) BIW: Transformed, Normalised, and Summarised Data Bespoke code to extract data, transform. Summarised Tables Select templates, change code parameters for specifics of job Bespoke code to select data based on specifics of job, enhance with fields of interest SAS templates to pull data, run analysis generate automatic profiles and create models. Intermediate Tables Code: Analysis, profiles, develop a model. Output tables Output tables Reports/Presentations Reports/Presentations
BAU Targeting, Bespoke Models and Analysis Sequence of DevelopmentCI has been preparing to do less work Productisation Web Portal Jan 2010 Dec 2010 Productisation Productisation Stored SAS Procedures Reference Tables, Formats, Macros Marketing Comms Process X-Camp Optimisation Marketing Comms Tables Hopper Data Warehouse (BIW and SAS)
Ensuring Data Consistency Across All Analyses and Customer Interactions All our analysts are focused on the customer and what drives them. The team wants to ensure we are continually building and enhancing our single view of the customer. This view needs to be readily available for all analyses, reports, models, campaigns, etc… Same data can feed into our communication management framework Capture data about interaction Use relevant customer data to drive the message/offer Analysis data combined with customer interactions maximises our understanding of what drives the customer and ensures relevance of communications
Final Prepare – A Single Customer ViewOne dataset to rule them All Data Warehouse Analysis Output SKU Transactions Profiles Rewards Campaigns Cardholders Final Prepare Models Transactions Census Mapping Real time data Account Reporting Scores
Segmentation LNZL had been doing only RFM segmentations on a participant basis. Simply, yet effective. We wanted a more mass customised segmentation (like Mosaic), but we did not want to use traditional demographic data. The key objective was to build a lifestyle based segmentation that is equally applicable for all Fly Buys participants, rather than focused on any particular participant or type of participant. Using our Customer Lifestyle Surveys undertaken by Loyalty NZ over the past two years with 50,000 respondents in each survey, CI team developed knowledgeCUBE segmentation.
Enables CI team and our participants to move beyond the standard geo/income dominated segmentations – provide an understanding into what makes the people tick.
This was a risky approach because it could have meant that we have segments that do not correlate with behaviours that we measure. However, it has worked spectacularly well.
The knowledgeCUBE Segments
Example:Ranking a Target Group by the Segments Across all participants, we can show how their base ranks according to the segments. Advantages include instant ranking for any data profiling request for any participant (Example below shows ranking for customers who redeemed through our Premium Rewards catalogue) Segments can facilitate knowledge in your organisation. Results across all activities can be stored at the segment level .
Improving the Marketing Campaign Process
Marketing Campaign Process Design The campaign process is completely standardised and integrated with core systems yet process can still handle a wide variety of situations and levels of complexity. Bespoke code has been minimised. SAS EGuide (Analyst) End Processing(Standard) Comms Tables (Data Warehouse) LeadInitialisation Final Prepare (standard) Hopper Campaign Code (Bespoke) Model Development Mailfile
The Gatekeeper The Gatekeeper becomes the common final funnel to all campaign files done by various analysts. Campaign Files Hopper Mailfile Mailfile Mailfile Mailfile Cross Campaign Optimisation
Selection Profiles Campaign files always need to be checked for quality, so we have improved our processes involving quality checking and signoffs as well as improved standard selection profiling reports :
Post Campaign Analysis Basic Sales Response v Non Response By Selection Variables Top performing outlets ROI Calculations
Industry Recognition for Loyalty NZ Earlier in September LNZL won the international Direct Mail (DM) Award for the industry leading Fly Buys Point Summary mailing. Publicis Singapore and Jon McKenzie, Digital Creative Director Leo Burnett, commented ‘that the new look loyalty statement showed that with great design thinking and an underpinning data strategy, this communication represented – best in class. It was the stand-out entry in what is a hotly contested category’ The underpinning data strategy is in fact driven by Gatekeeper in being able to allocate 175 different messages for 750,000 customers. This is over 600 trillion variations! The CI team will continue to evolve the Gatekeeper to handle more sophisticated simultaneous optimisation criteria. A great example product that does this type of optimisation is the SAS Marketing Optimisation.
What CI has become at Loyalty NZ The CI team now runs 10 campaigns in a week for our participants.
Half of these have had models or segments applied to them.
We are on track to do over 500 campaigns by 31-March-2011.
The team offers over 20 analytical products, from simple reports to profiles, to maps and even SKU-based models. Continuing to broaden the scope of our thinking to think about the customer from a single view. Contact strategy and strategic segments are being refined for 2011… Knowledge management framework for realising synergies across analyses. Layering our data and insights onto our common frameworks in order to continually understand what drives our customers. And this is just the beginning...
Take Aways Never stop thinking about what your data can do for your marketing and your business Make synergies in your Analyst team by making One Dataset to Rule Them All. Establish knowledge management practices to give life to the One dataset.