Table of contents for Teradata Third Party Influencers 2010
- Teradata – Key Messages
- Teradata Database 13.10
- Teradata Customer Management Portfolio
- Teradata – National Australia Bank and next best action
- Teradata – Coca Cola and customer intelligence
- Teradata in the cloud
- Video from the Teradata event
- Teradata – Q&A with Stephen Brobst
- Teradata Business Analytics Innovation Center
- Teradata Customer Warner Bros Home Entertainment
- Teradata Active Enterprise Intelligence
National Australia Bank group (a 15 year Teradata customer) presented on their use of Teradata for multi-channel marketing. NAB has 10.9M customers worldwide and about 38,000 staff. Brock Lynch, from the marketing group within the retail bank, gave an overview of their analytical CRM approach. NAB sees its analytical CRM assets as:
- Complete customer data stored in Teradata.
- Campaign Management: Product across mutli-channel access points, multi-step campaigns, inbound and outbound.
- Analytical skill set – specialist data miners
- Tools for analytics including SAS, KXEN, Salford
They build out from a core of integrated data, to analysis (data mining and predictive analytics), to campaigns (around households, relationships) and finally to multi-step efforts across channels both inbound/outbound and call-centers, ATMs, branches etc. NAB works trough a funnel that drives down through mandatory exclusions (bad debt, poor credit) to consent/privacy preferences, to customer contact preferences, to channel management limiters (limited number of relationship managers, for instance) and finally a control group. The end result is a timely, appropriate and quality conversation.
They focus on what I would call next best action not just next best offer as sometimes they drive to branding or relationship actions not just offers.
The particular project was around improving their inbound system. First, targeted tellers – had to be quick and low-impact so that the teller could see the offer being proposed but only act on it if there was time and if the conversation allowed it. Had to be low impact so did not disrupt the main business of the teller. In contrast, when they integrated the offers into the call center system (Siebel) it was higher profile. They identify customers as they call in and then provide this information to the call center person to whom you are routed (they plan to use the analytical CRM decision to route you too). The information is presented more aggressively so that it is noticed by the call center people. Like the teller, the call center person captures responses. Finally they integrated the offers into the internet banking home page. The ATM network is next, but it has only just been updated. Plans include coupons and perhaps driving folks to the branch as well as “service” offers.
They picked inbound because they have over 600M inbound interactions (Internet banking, ATM, tellers, call center). In addition, the outbound channels were being throttled by capacity and they believed that people would be more responsive if targeted while doing banking things. They took advantage of this new inbound channel to broaden from offers to things like collecting corrected cell numbers or addresses, have service conversation and talk about branding and community activity. Next best action indeed.
Mixing inbound and outbound did cause some changes. They held workshops on triggers and events and found they had to make it ok to have more conversations. They also had to understand the value of fixing data quality problems and relax their recency rules (because inbound calls are so much more numerous). It was important to keep a “service first” mindset through simple topics, scripting and keeping it fresh. Customer trigger and event-based conversations went well (“we’ve received returned mail so ask about their address”) as did those based on promoting alternative ways to self-serve. Scripting generic life stage conversational starters to help relationship bankers “connect” with customers of a different age. Finally kept track of negative responses and found this highly predictive longer term.
Challenges included making sure the conversation was appropriate to the channel (wasn’t simple to fix data in the teller system, for instance, but was in the call center) and ensuring that there was not more insight than the conversation could support (“How do you know I am pre-approved for a loan? Why?”). High volume (everyone) campaigns performed poorly as did partner programs.
Critical Success Factors:
- Tactical projects focused on channels to drive strategic initiative
- Included program in targets and rewards
- Identified champions in the call center to direct/mentor
- Training and communications
- Volume driven world so inbound volume matters
Benefits included more customer-centric conversations, quick response marketing (24-48 hours to get new offers out), more personalized and targeted offers without costs, and more measurable returns.
Very cool case.