Social BPM At IRM BPM London

I’m in London this week at the IRM BPM conference. Today, I’ll be taking my Making Social BPM Mean Business 3-hour seminar for its first real outing (although I’ve presented most of this material in other contexts, just not in this combined form); over the next several months, I’ll be presenting this at three other conferences, but it’s always changing because there’s always something happening in this area.

Tomorrow, I’ll be moderating a panel on business architecture, where I’ll have the chance to discuss what it is, what the value is, who should be involved, and the skills required with John Gøtze, Neil Ward-Dutton, Mike Rosen and Martin Frick.

If you’re here, track me down and say hi.

Increasing Revenue Through Multi-Channel Decisioning

John DeMarchis from PNC Financial Services gave the final morning keynote on how they’re using multi-channel decisioning to increase revenue through a personalized customer experience. They looked at some of the compelling user experiences in the consumer space that are setting expectations of how people want to interact with service providers – Amazon and Google, for example – but also had to consider the loss of confidence in the US financial system that many consumers have experienced through the financial crisis. This led them to develop a customer interaction management system that uses Pega decision management to inform and direct customer interactions through all of their engagement channels. This provides centralized decisioning for consistency, real-time decisioning to allow next best action capabilities during the interaction, and adaptive learning to adjust the models based on market and customer behaviors.

Key to this is determining each customer’s current state so that they can be appropriately targeted for offers and notifications without annoying them, using the channel of their choice. Seriously, I wish my bank could do this.

The results: they are at the top of the customer experience rankings in specific geographic regions, and are ranked highly in terms of how they interact with their customers for marketing purposes. Interestingly, they started this with the simple idea to do cross-selling, but found that using the data and decisioning capabilities, they could do much, much more.

We finished up with some closing thoughts from Alan Trefler, since this is the last of the general keynotes; it’s all breakout sessions from here on in. I have (paper) notes from my meetings with executives yesterday, and a few more meetings today, which I’ll try to summarize here later today or tomorrow.

Decision Management And Next Best Action

Rob Walker, Pega’s VP of Decision Management (who joined Pega as part of the Chordiant acquisition), gave a keynote on day 2 of PegaWORLD about next best action. He started with a great visual of the analysis of a particular baseball pitcher’s stats, showing that there was some degree of predictability of what type of pitch (e.g., fastball, slider) will occur after an initial pitch of each type. In other words, if you knew those statistics, you would have a better chance of being prepared for a particular type of pitch once you were smoked by that initial fastball. Sticking to a baseball theme, he moved on to talk about the concepts behind Moneyball: using statistics to build a better team than you can with beliefs and biases. In other words, more data and a lot of analysis can give you a competitive advantage, an idea that holds value far beyond the realm of baseball.

He spoke about adding in all of that extra data and analytics to the customer view as “color” being added to the picture in order to make details more evident throughout the lifecycle. This can, for example, allow you to detect when a customer is likely to defect to a competitor based on their service history and behavior, or what new services that they are most likely to buy at a given time. Bring this next-best-action marketing to the Pega unified marketing portfolio, and you have this capability baked right into your processes and multi-channel customer communications. Instead of using an inside-out strategy of trying to determine the right market to target for a specific product, this allows for an outside-in strategy of maximizing the value of a specific customer through understanding their propensities, and doing it fast enough to make the next best offer in real time, while they are on your website or on the phone with your customer service or sales rep.

Relationship-Driven Customer Service At American Express

Jim Bush, EVP of world service at American Express, delivered a morning keynote here at PegaWORLD to talk about customer service, and how they’re transforming it to provide better value to their customers. 93% of those surveyed say that companies fail to exceed service expectations, which is a complete disconnect with the fact that companies that provide superior service will get 13% more customer dollars because of that. They looked at a new customer service paradigm to deal with the business realities: multiple integrated service touchpoints; experiences benchmarked across industries; consumer choice; increased regulatory scrutiny; and better-informed, more powerful consumers through social media and other means. The customers are back in the driver’s seat in most consumer-facing businesses.

AmEx responded to this by deciding to service customers, not transactions. That’s an important distinction: a specific incident (whether positive or negative) needs the context of the entire customer relationship in order to understand how to best address it. They now consider service to be not a business cost, but an investment in business growth, and focus on respecting and deepening the customer relationship. To do this, they reconstituted their service organization as World Service, with the goal to enable, engage and empower. No small feat, considering that their 20,000 customer care professionals handle hundreds of millions of customer interactions in 22 markets, 15 languages and eight engagement channels.

They have moved past the granular measurement of “how did we do on this transaction” to the net promoter measurement of “would you recommend us to a friend”, which changed how they think about the customer relationship. In fact, they have just trademarked the term “relationship care”.

I’m in the middle of reading United Breaks Guitars: The Power of One Voice in the Age of Social Media – a fantastic and funny read about one man’s journey through a customer service nightmare (and if you haven’t seen the “United Breaks Guitars” video, get on over to YouTube right now) – and one of the points that author Dave Carroll makes in the book is that managing customer service on a transaction basis tends to make companies ignore what they think are statistically insignificant events such as a specific bad transaction with a customer. That, in short, is exactly how not to treat a customer if you want to foster a relationship.

Thinking about the customer relationship rather than just servicing a specific transaction puts AmEx on the right track towards service innovation. They’re also looking at engaging customers through the channel of their choice, from paper to telephone to SMS. If a 162-year-old company can do this, every company has the potential to do the same, and yet many continue to put their head in the sand on turning their customer service around to actually serve the customer in the manner that the customer wants to be served. To serve the global citizen, AmEx combines relationship care with channel convergence, integrated technology, and a global scale with borderless solutions. Through that, they want to turn indifferent (although satisfied) customers into promoters and advocates for the brand. Not many of those customers will end up with the AmEx logo tattooed on their arm, but a single voice can go a long ways these days.

Oh yeah, and they use some Pega software.

AWD Monitoring Technical Deep-Dive

Great keynote at AWD Advance this morning by Captain Michael Abrashoff, author of It’s Our Ship, a book on leadership; I confess to tearing up a bit when he described how he supported and encouraged the young people who worked for him, and hope that I did a fraction as well when I ran a company.

Back to business, however, I’m in the technical session on AWD monitoring and business intelligence, following on from Kim Smyly’s introduction to the new monitoring yesterday, where Dirk Luttrell and Bob Kuesterteffan are giving us a peek under the covers for their new monitoring offering. They are implementing dimensional data modeling in their new offering – which, as I pointed out yesterday, is based on Oracle’s BI – in order to provide better business-based metrics and analytics. We got a brief tutorial on dimensional data models (star schemas in relational databases, or cubes in multidimensional databases), making me wish I was paying a bit more attention when my other half was talking about how he was implementing one of these in his data warehouse. In short, relational data models are organized around transactions, whereas dimensional data models are organized around business entities and information. Business entities are represented in fact tables, and dimensions are key to selecting, sorting, filtering and summarizing the data contained in fact tables.

The core AWD data is based on relational models, since it is a transactional system, but both the process and line-of-business data in AWD can be published to the dimensional (star) model for easier reporting and monitoring. If you’ve ever written a report or dashboard based directly on the process transactional data in a BPMS (which I have), you know that it’s not pretty: BI tools are optimized for dimensional data models, not relational transaction models. In the past AWD has allowed for reporting directly against relational models, but it was (is) not very flexible and could be prone to performance and scalability problems, requiring either extremely complex (and compute-intensive) queries, or denormalization and data duplication. Furthermore, it requires that report writers know and understand the underlying relational data model since they’re writing directly against that physical schema, which further locks in the core AWD product to that schema rather than being able to mask it behind a logical data schema.

In the new dimensional data model, they represent business entities directly: work items, queues, users and various other attributes of work including time dimensions. They also include a single line-of-business data dimension for all LOB fields (this seems like they are relational-izing their dimensional model, but I can understand the administrative and design complexity if they didn’t do it this way), so that fields such as account number can be used to cross-reference to other systems or for filtering, searching and sorting within the BI context.

They are creating the following fact tables:

  • Assigned work fact, with dimensions regarding when and to whom a work item was assigned and unassigned, and the current state regarding assignment and work selection. This is used, for example, to report on assigned work by user.
  • Completed work fact, which tracks work steps as they are completed, including duration, user experience level and other information about how the work was completed. This is used for reporting on work that was completed.
  • Locked work fact, tracking items when they are locked by users: who, when and how. As with assigned work fact, this is used for reporting on work locked by a particular user.
  • Login status fact, tracking when users log in and out, and whether they are currently logged in.
  • Queue fact, tracking work as it moves from queue to queue, and the status that each work item is in.
  • Suspended work fact, including when items are suspended and unsuspended, and who did it.
  • Work fact, which including historical information on work but includes a “current” flag to filter for just work that is in flight.

[This is probably way more detail about their dimensional data than you’re interested in, but I blog because I have no memory, and this is my only record of what I see here. That’s right, it’s really all about me.]

Given that the same underlying relational model will still be there in AWD, customers can continue to use the existing AWD BI (which would hit against those tables), but I’m guessing that a lot are going to want to move over in order to take advantage of the ease of use, performance and scalability of the new BI environment. They’re also planning on some future features such as scheduled report delivery; I’m not sure which of the new and upcoming features are based purely on the underlying Oracle technology, and how much that they’re building themselves, but if they’re smart, they’ll leverage as much of the Oracle BI package as possible. They also need to figure out how to integrate/publish to enterprise data warehouses, and work up full replacement functionality for the current BI product so that it can be retired.

Case Management In AWD 10

Judith Morley presented on their new case management capability; she started from some pretty basic principles explaining knowledge work, so likely a fairly novel capability for most of the audience.

She described case management as a new application or user interface, meaning that the AWD 10 BPM capabilities are there as part of it, but it has additional capabilities such as collaboration, content, ad hoc processes and deadlines. This circles back around the ongoing discussions in the industry about the relationship between BPM and ACM; certainly, process is a part of ACM (even structured process), but it’s more than that. They did research with their own BPO companies and some of their customers spanning retirement, mutual funds, insurance and healthcare industries, and came up with four design imperatives for a case management solution:

  • A humane way of working with files
  • Reorienting yourself to a case: making it easy to pick up where you left off after some time away from the case
  • Immediate responsibility versus ultimate responsibility: understanding ownership and responsibility for meeting milestones
  • A system that suggests rather than dictates: supporting the knowledge worker rather than enforcing a specific process

The primary workspace now for knowledge workers (as defined in their profile) is a dashboard listing their top 10 tasks – as defined by what they own and due date – and a task forecast for the next three weeks, then their top 10 cases and the case workload of all members of the worker’s team. There are two other tabs for cases and tasks; on each of those are interactive filtered views of the cases and tasks in progress. Both cases and tasks are types of AWD work items (with a predefined process model, even if just a single-step user task), with tasks being children of cases; opening a case or a task takes you to a view of that work item with the related data, content and activity. Tasks can be added to a case by the worker, using a template, and content can be added at the case or task level. Messages get passed around between cases and their tasks to allow for processes to be started, paused and rendezvoused appropriately. Cases can be created from templates as well, where a case template contains one or more tasks of any degree of complexity. Both task and case templates are, in fact, templates: if they are changed, work that is already instantiated is not impacted. Furthermore, cases can be organized into folders as a collection mechanism, although folders are not routes as cases and tasks are.

This is not yet a released product: it’s scheduled for the end of 2012 or the beginning of 2013, and they are currently researching different representations that they might create of manager and team views, as well as reporting on knowledge work. This latter issue is one that I’ve been talking about a bit lately, and proposed it as one of the “unanswered questions” in my presentation on the nature of work at last year’s academic BPM conference.

Document Capture to Process

The second half of the morning at ISIS Papyrus was a pair of sessions on document output and document input, or as the session titles put it, “unified print, web and mobile output management” and “scan to extract to process”. The first was a great deal more information on yesterday’s session on correspondence generation, and you can check out their website for more information on the features and functions.

The second session, on document capture, started with business drivers for capture and scanning, particularly the link from capture to process. A lot of organizations are doing capture to archive, but a smaller percentage are using recognition technologies (classification and extraction) or triggering processes from the documents. Interestingly, a much larger percentage are using document classification (i.e., what type of document is this) but not data extraction (i.e. what customer-specific information is on the document); there’s a big ROI just in document classification without additional recognition since that can be used to automatically route documents to the right department for handling.

ISIS recognition architectureThey handle a variety of input channels and devices: scanners (including check scanners), MFPs (“Mit Netzwerkanbindung”, according to the slides Winking smile ), fax, email, camera phones, social media and file import. I had a demo yesterday of their recognition and extraction capabilities, and it can be applied to both structured forms and freeform documents. A typical capture path includes steps for rearranging, splitting and merging (often for fax documents, over which there is less control on the input), then document classification, automated extraction and validation, then manual verification and correction. In this way, they provide the same functionality as products such as Kofax or IBM Datacap.

Document classification can be based on layout (fixed format), keyword (text or barcode) or text-based (phrases), and their capture design environment allows for learning by example to create any of these. 2D barcodes (which may have been generated by their outbound correspondence module) can be identified and used to match an inbound document to an existing case.

There are a number of administration and monitoring capabilities to track how many documents/batches have been captured and where they are in the processing queues.

They have a number of customers that are capture only – this, in addition to correspondence generation, are two of their key use cases – but provide greater added value if the capture process leads directly into distribution and case creation/management.

ISIS Papyrus Adaptive Case Management

ISIS Papyrus defines (and implements) ACM as the full range (dare I say, a spectrum?) from straight through processes through dynamic processes to completely unstructured process driven by ad hoc content arrival such as email or social media. This, I believe, is at the heart of discussion/argument about BPM versus ACM: this definition has traditional structured BPM as a subset of ACM at the structured end, with ACM covering a much broader range of structure as well as being inherently content (document) centric, and including a number of additional capabilities such as goal orientation and business rules.

At the structured end, this can be fully automated service orchestration, driven by events or by (document) state: this can be modeled as a flow diagram, but the individual tasks are adorned with additional information about state and events that impact that task, such as a task firing when a document reaches a specific state.

At the unstructured end, this provides a collaborative case-oriented environment for a knowledge worker to manage a response to some sort of inbound content, including integrated correspondence management based on the core technology that we saw in detail yesterday.

In any type of application along the spectrum, but especially at the unstructured end, you can have access to social media channels (both inbound and outbound) as well as data from other systems and related documents. External events can impact the case, and business rules created in natural language can be applied to constrain or act upon the case. Tasks within cases are linked to goals, as defined in the business architecture; goals can be linked, and sub-goals defined for more structured dependencies. The system learns as more cases are processed relative to their goals, allowing for the next best action to be suggested to the user on a case based on the history of similar cases; this could, of course, be applied in an automated fashion rather than as a user suggestion, but this level of machine intelligence tends to make some organizations uncomfortable.

Although processes can be represented as flowchart-style models, they can also be represented as Gantt charts (or PERT charts, for that matter) to be able to visualize the critical path through the process and provide some predictions around due dates for various milestones. I’ve seen this representation from other vendors, most noticeably BP Logix, for whom I’m writing a white paper on predictive analytics and the importance of adding a time dimension/representation to processes.

There is a task-specific user interface for the details portion which must be customized for each task type, although the framework includes standard information such as a history of the case activity and resources. The task interfaces are developed using widgets, and a single UI definition can be deployed on any platform (including mobile) without customizing specifically for that platform. A mobile deployment environment is becoming critical in application development, as we saw last week at IBM Impact with their focus on using their Worklight acquisition for mobile development and deployment.

They’ve created the critical round trip between strategy and execution by connecting strategic objectives (in a strategy map) to business architecture (in a capability map) to process goals (balanced scorecard and other KPIs): not only is this top-down, where strategy defines capabilities, which in turn are used to define KPIs, but also feeding back so that the actual performance during execution is compared back to the architecture and strategy.

ISIS Papyrus stresses that ACM is just a capability of their content processing platform, and I think that this is part of the confusion around the definition of ACM: ACM is about how we do work, so requires a combination of activities, content, rules, user interface, and integration with external systems. However, there are a lot of application development environments that provide some or all of that without being defined as ACM, and more traditional BPM products are redefining themselves as ACM by adding some of these capabilities even if it’s not a good fit with their underlying infrastructure.

The ACM market is still emerging and will continue to evolve. Having some good examples of ACM in action through the ACM Awards (for which I’m one of the judges) will help with market understanding, but I anticipate many more discussions on this topic along the way.

Mobile, Social And Integration With The Papyrus Platform

Day 1 of the ISIS Papyrus open house was more about their capture, document processing and correspondence generation, which is what many of their customers are using. Today, the focus is more on newer functionality, and we’re starting with Roberto Anzola, senior manager of R&D, discussing their mobile and social capabilities.

They provide a mobile app that acts as a portal to any application developed on their platform; Forrester has recently identified this functionality as “mobile backend as a service”, where the application is defined on the server, not on the device, and accesses data and user interface components from the server. This provides access to the same application on iOS, Android, PC desktop and in the browser through their UI widgets. The mobile server supports SOAP and REST calls, as well as OAuth for authentication on social networks. The app provided for the conference (which you can find on the iTunes app store by searching for ISIS Papyrus), is built on their mobile server technology. We saw a demo of an iPad-based vehicle claim app built on this platform, and how the menus and features on the app are driven by the case definition in the desktop environment. Because it’s driven by the Papyrus platform, the app has access to the same data and documents as a desktop application, although rendered in a mobile form factor.

He went through a number of different integration points that they have with other systems:

  • Access to LinkedIn contacts for inclusion in an application.
  • Detecting and responding to Twitter messages (as we saw yesterday)
  • Set and retrieve events in Google Calendar
  • Use Google Translate to translate text building blocks in the designer, or on the fly for dynamic case information
  • Use SharePoint as a document repository for documents linked to a case as well as exposing cases within SharePoint using Papyrus WebParts
  • Integrate with SAP (and many other systems) using web services
  • Direct file system integration, where case data objects can be exposed for navigation directly in Windows Explorer if that is a more natural interface for users rather than using Papyrus directly, although it wasn’t clear how access control to the files is managed
  • CMIS access to any standard document repository, including EMC Documentum, IBM FileNet, Alfresco and SharePoint

Although the title of the session was about social media, this went through a much broader set of integration capabilities, as well as the mobile platform.

Personalized Electronic Statements at China Trust Commercial Bank

We’re running 30 minutes late and it’s right before lunch, but everyone is sticking around to hear Ignatius Chang, EVP at China Trust Commercial Bank, talk about their initiative for statement e-delivery. Based in Taiwan, they are in an overcrowded financial services market, and technology has become a competitive differentiator for them in customer retention through superior service. They have about 5 million credit card holders, and send out over 2.5 million statements per month. These statements aren’t just a statement, however: they include targeted promotional information and offers for other products and services in the same document.

Their move to e-delivery was driven by some basic business goals: they wanted to reduce print and post costs by moving customers to e-statements, but maintain the personalized marketing and visual appearance of the printed document without generating a PDF. They wanted a common document design and handling process for both paper and electronic, although electronic required additional features such as secure encryption and support for mobile devices. Since the statements included marketing campaigns, they wanted to be able to track the campaign response rate, since this contributed to their return on investment (in addition to the cost reduction when moving to e-statements).

They are only partway along their journey with the Papyrus platform, so some of this described future plans rather than current reality – it will be interesting to see how they achieve their goals as they continue to implement in the future.