Dan Dickinson (@digitalbankguy) Managing Director, Online & Mobile Banking (Canada) BMO Bank of Montreal
BMO Financial Group Established in 1817, BMO Financial Group serves more than 12 million personal, commercial, corporate and institutional customers We provide a broad range of retail banking, wealth management and investment banking products through our operating groups: Personal and Commercial Banking BMO Bank of Montreal (CAN) BMO Harris (US) Private Client Group BMO Capital Markets $542 billion in assets, 46000 employees, 1500 branches 2 Our mission: to be the bank that defines great customer experience
What I ll be covering today Why we built a leads engine Why we made the call to go multichannel How it manifests in digital channels vs. assisted channels The results What we might have done differently What will come next 3
Why we built a leads engine 44% of direct mail is never opened * 84% of 25 to 35 year olds have clicked out of a website because of an irrelevant or intrusive ad * The cost per lead in outbound marketing is more then the cost for inbound marketing * For BMO Bank of Montreal: Our traditional direct marketing channels were delivering low response rates Customer data latency was making our targeted leads irrelevant to the customer We did not have a consistent 360 degree view of the customer across channels In 2010 we began building a Leads Management Engine, or LME 4 * Source: Inbound Marketing vs Outbound Marketing Infographics, Voltier Digital, Oct 2011
Objectives of the Leads Management Engine (LME) project Provide a new and innovative capability to deliver near real-time leads across multiple channels based on the customer profile and purchase likelihood Align the presentation of the leads to support the customer conversation / interaction Streamline the process of reviewing and discussing leads with a customer (in person or on the phone) from introduction to fulfillment (or appointment) Make leads accessibility and fulfillment truly multichannel Release 1 focused on inbound leads delivered to the branch and contact centre Release 2 focused on ATMs Release 3 focused on Online Banking Mobile banking release still pending 5
Before the technology: the process for creating a great customer experience 6
Defining great customer experience: know the customer 7
Defining great customer experience: identify potential customer actions 8
Defining great customer experience: create appropriate message 9
Defining great customer experience: deliver message 10
Defining great customer experience: listen 11
Defining great customer experience: action 12
Defining great customer experience: test and learn 13
How the LME works 14
Data about the customer and their holdings feeds the propensity model Customer Demographic information, segment Holdings Products, balances, reward schemes Session context Location, time, transactions performed, employee role (if branch/ccc) External services Credit scores (if on file) 15
LME architecture 16
We use IBM Interact for dynamic segmentation 1 Channel specific segmentation strategies 2 3 4 5 Product lifecycle approach, communicating with the customer at relevant touch points 6 Business rules allow for robust refinement 7 Customer Events that positively impact customer experience and inturn customer satisfaction 8 Global control group enables robust reporting at a channel, segment and cell level 17
How a lead appears to branch and contact centre staff New functionality will display the conversation cue text when the employee hovers over the row with the lead 18
How staff takes action on the lead. Pro tip: RSP is Canadianese for 401k The list of leads will be prioritized with the best cues at the top A list of potential outcomes lets staff continue with the flow of the conversation 19
Extending leads to self-serve channels Since the vast majority of our customer interactions happen at self-serve channels we moved as quickly as possibly to phase two: ATMs Note: leads were displayed only at our full-service ATMs, not our basic cash dispensers We learned some important lessons during the ATM deployment: The offer has to be very simple: there s no employee to explain the offer to the customer The action also has to be very simple: asking for anything more than a yes or no gets tricky It s vital to optimize the speed of the round trip to get the lead >2 seconds doesn t cut it Moved call earlier in the ATM flow, and added cycles to handle increased volumes In May of this year we completed phase three: connecting Online Banking to the LME 1.8M active customers 700k logins on an average day, >1M logins on a busy day Leads delivered via in-line ads and Offers section of message centre 20
The project team The core LME business owners worked on speed, responsiveness, and robustness of the LME with their IT partner team My online channel team worked on the placement and responsiveness of ads inside Online Banking with their IT partner team, while my online sales team ensured quality and applicability of offers Corporate Marketing (ebusiness) provided creative assets for the ads and offers, many of which are shared across channels (OLB, bmo.com, ATMs, etc.) Program Delivery team provided project management support Note: core product systems and online banking are homegrown. LME is IBM Interact. All IT resources are internal. 21
How leads manifest in Online Banking Customer views their offers and clicks to read more about one 22
Performance to date, and what we might have done differently Product sales have been good, but haven t blown the roof off yet Monthly product sales originating from the LME roughly the same as 5-6 branches Primarily lending products (credit cards, loans, LOCs) and simple investments (GICs, taxfree savings accounts) Good uptake on customer entrenchment offers as well Roughly 3000 service-related offers accepted per month Card balance insurance, NFC payment stickers, card limit increases, etc. We ve learned where the throttles are The offer pipeline is managed by marketing and product, who still operate in campaign mode working on moving away from that paradigm toward real-time offer algorithms Graphic asset creation for offers takes too long, especially if we have multiple offers if we offer a product at five different rates to five different customer profiles we have to create and load five graphic assets 23
What s next? Mobile banking will be connected to the LME in mid-2014 as of February our online and mobile channels will run on a single code base, so cost and project time to extend to mobile will be minimal Currently working on content management for ads and offer messages personalizes offer with customer s name, etc. reduces asset creation overhead Continued effort to move from campaign mode to sense-and-respond e.g., providing next-best-product for all segments Additional sources of data to feed decision engine Clickstream data from web session, holdings from wealth / private banking, customerentered data from PFM suite (budgets, savings goals, etc.) 24
* Dan Dickinson (@digitalbankguy) Managing Director, Online & Mobile Banking (Canada) BMO Bank of Montreal * See what I did there?