Collaboration,Enterprise 2.0,Social CRM,Workflow / BPM

Service Agility thru Adaptive Case Management

27 Oct , 2010  

Who hasn’t been there? After waiting 20 minutes on hold with Customer Service, then 15 mins for explaining your problem, the Contact Center Agent says she is really sorry, but can’t do anything about resolving  it because something first has to be done by some other department, or it the system won’t let her do what needs to be done as this isn’t in the process, even though it should be simple and straightforward to do.

The customer is frustrated because she need to call some other number and explain everything all over again, and the agent is becomes frustrated and may take it out on collegues, or worse – the next client in the caller cue! The example above is an issue of empowerment and system inflexibility, but also of process rigidity. The rigidity of the process has just killed the positive customer experience.

What happens when BPM breaks down?

Thanks to Taylorism we spend bucketloads of money to model our understanding of how things should work, hoping to recoup the time and energy in the long run thru process efficiencies. Although this may work well for Easily Repeatable Processes (20%-40% of the time), resolving Barely Repeatable Processes effciently and sustainably will be a key challenge to meet, especially as customers now have more choice and are more informed through their social network and are more sensitive to how their ‘experience suppliers’ cater to their needs.

BPM and Social BPM

Business Process Management (whcih is part of what I did in my previous role) solves a real business need by identifying, modeling and optimizing processes – cutting work into repeatable parts and adding decision trees to ultimately lead to a known outcome. It is characterised by heavy upfront investment in modeling which the company intends to recoup through efficiences in the long run. However, BPM is not a panacae as it can be a straight-jacket leading to customer and employee  frustration when exceptions occur (and inevitably they do), with little or no ability to adapt the model once it has been set. Especially not by those that go through the drill everyday such as your Contact Center Agents!

There is a good group of people and companies looking at at how to make improvements, gathered under the heading “Social BPM”. As Tom Allensen says: “Social BPM is basically just collaborative business process management utilizing a collective network environment – it’s about extending BPM access and decision-making to partners and select external parties without compromising the exclusivity of the core group.” Michael zur Muehlen adds: “Social is all about providing context, a rich environment of data points that a streamlined workflow would be lacking otherwise. The challenge is to make this context useful, both from a social networking perspective and from an unstructured data perspective.”

So Social BPM is about getting all those concerned involved (and why not the customer?) and about collaborating whilst taking the context of the task at hand into account.  Although this approach certainly has its merits to get people to collaborate effectively on process steps within a pre-existing model, it still does not add flexibility and agility to reaching the desired objective or outcome in case of non-modeled exceptions. If exceptions do take place, employees just revert back to email and the company could lose thread of what is going on, making it unable to capture insights for continous improvement and learn from them.

Adaptive Case Management

I’ve been looking at the potential of Adaptive Case Management (ACM) as a framework for guiding and structuring issue resolution (or even collaboration in general), relying on the insights and experience of knowledge workers (such as your agents), pulling together the right resources to make informed decisions at each step based on the context, and choosing and adapting next steps as necessary to reach the desired outcome.

ACM makes a distinction between routine and expert processes. Routine processes are prime candidates for modeling whereas expert processes are emergent and the domain of Knowledge Workers – they give the organisation the flexibility to adapt and respond to changes around it. ACM also aims to bring understanding, visibility and control to  unpredictable knowledge work and serves to make routine work processes more reliable. It doesn’t set out replace BPM or workflow, but adds to the tools that the organisation can leverage to be sustainable and resilient to changing expectations and conditions.

Adaptive Case Management sets goals that describe what must be done rather than how the process must be done. It provides guidelines and looks to Knowledge Worker experience on how to achieve the goals but does not dictate the work or the flow itself. Control is achieved through tracking deadlines and goals, and having a case owner responsible for the process results. The flow is organised around the context and content assets, and can be modified to bring in participants as needed based on the availability of information and expertise.

My focus is on the customer and customer experience and ACM applies to non-routine Customer Service, but is equally valid for for example Sales Processes (such replying to an RFP, customizing a product to a client’s need), Loan Management, Financial Audits, Innovation, Strategy Implementation : basically anything that requires collaboration and negotiation. As another more concrete example, who knows beforehand how a patient will be treated when entering an Emergency Room? The flow of events that lead to the patient leaving the hospital in good health is highly unpredictable but is guided by fragments of processes (diagnosis, CAT-scan, medication, operation etc.) that are assembled as the situation evolves and more information becomes available.

This is a quick (certainly not exhausive) introduction to the concept of ACM, which in my opinion can be a valid framework for guiding emergent Knowledge Work – which Enterprise 2.0 is ideally suited to facilitate! In my opinion, E20 has been putting a lot of emphasis on devising the tools to improve information flows within an organisation to the detriment of actually helping the employees do their task at hand better. Where it can add a lot of value is by organising the content assets, identifying the people with the right expertise at the right time, and facilitating information flows and sharing, as well as tracking and consolidating lessons learned by capturing fragments as templates and distributing these. Social CRM – through Social Analytics – can help by providing a better insight into the context which can then be used by Knowledge Workers as information inputs on which to base their decisions, choose the next steps and who could provide the right expertise, which can include partners, suppliers and of course customers!

In my coming post I will discuss Adaptive Case Management in more detail as I think this has great potential to for linking Enterprise 2.0, social CRM, business and social analytics on the road to Social Business and the Collaborative Enterprise. Together they provide framework for work in an agile organisation.

I’ll be at the Enterprise 2.0 Conference in Santa Clara (8-11/11 2010) to participate in a customer panel on social CRM. I’ll be happy to meet and exchange ideas on the above subject when I’m there. Drop me a note and we can arrange to meet!

18 Responses

  1. Mark,
    very interesting post !

    The ideas on Social BPM and Adaptive Case Management have great potential for moving the focus beyond technology and automation towards smarter processes, where I believe the true value is.

    Keep up the good work !

  2. Mark Tamis says:

    Thanks Stale 🙂

    Not only smarter but also more agile, and giving Knowledge Workers a way to express their experience and insight. This in turn will also give companies a better way to tap into the full potential of their employees.


  3. […] mer om Social BPM og Adaptive Case Management på bloggen til Mark Tamis (engelsk). Del med […]

  4. […] BPM and ACM – Mark Tamis ACM makes a distinction between routine and expert processes. Routine processes […]

  5. vikas nehru says:


    interesting article.

    what are you working on? I am the VP of Marketing at KANA and would like to have a chat to explore possibilities of working together.

  6. Mark Tamis says:

    I’ll be in Santa Clara next week for the Enterprise 2.0 Conference next week, let’s try to connect then!

  7. Jennifer says:

    This is an area where we (Collective Intellect) have been moving – integrating our technology (social media analytics)/expertise into a workflow. You include this quote in your post:

    “Social is all about providing context, a rich environment of data points that a streamlined workflow would be lacking otherwise. The challenge is to make this context useful, both from a social networking perspective and from an unstructured data perspective.”

    This is Social Business Process exactly. Social fills out and gives depth and understanding to a workflow, without these very human touch points a company’s direction can make it irrelevant at worst or, at the least, aggravating to its customers.

    Thanks for the very interesting post and for providing additional clarity around this fairly new thinking.

  8. Ray Brown says:

    Hi Mark Interesting article. I believe that we’re seeing a coming together of the customer and the employee to create “actionable insights.” As you say the key is to create structure and process so that businesses can both hear and then act on these insights. Inside-out/top-down becoming more outside-in/bottom-up?

  9. […] is just to rigid to effectively handle ad-hoc, unpredictable interactions. This is where I believe Adaptive Case Management has a role to play. It empowers the Knowledge Worker, gather step inputs as relevant (history, […]

  10. Larry Irons says:

    I think the way people recognize cases is as important as the way they are managed. In other words, in my perspective the important point is how people (customers, employees, or other stakeholders) proactively recognize the emergence of an exception and act on it. Just my first blush response.

  11. Bayesian inference math can adapt paths to guide to these goals. I learned this math from Professor Daphe Koller at Stanford. Intelligence data bases now meander for bad guys without models looking for bad guys. Website links may help you.

  12. Oops edit:

    Bayesian inference math can adapt paths to guide to goals. I learned this math from Professor Daphe Koller at Stanford. Intelligence data bases now meander for bad guys without models. Website links may help you.

  13. Mark Tamis says:

    Hi Clive,

    I found the link through Kelly Craft (HT) . Interesting, but I wouldn’t want to go as far as have the system make the decisions, to me there needs to be enough room for the Knowledge Worker to bring her insights into the outcome and the decision on next actions so that there is enough variation to avoid making cowpaths (room for serendipity)

    Have you looked at the Rete Algorithm for reachng desired outcomes? Carole-Ann Matignon has done a good job of describing it here (@cmatignon)

    Thanks for your input!

  14. Carole-Ann is describing Bayesian Inference.

    Theoretically model free Bayesian architecture can learn to solve goals -> Evolving intelligent ‘workflow’. But assumes massive data sets.

    In the ACM context, I agree with your point, don’t remove the customer service knowledge worker, because there’s not ofen not enough data.

    But the knowledge worker can help speed up the solving the goal, while you capture tacit knowledge into the form of probability relationships (that can assist junior customer service evolve a workflow).

    Proof case, check your spam box, full of nsfw emails. Yet you can override the simple Bayesian filter, an teach it so that a certain email now goes correctly to your inbox.

    ACM can work the same way.

    Click my name for link to code MSR video and .net sample code.

  15. Update Mark,
    Via Cloudera, Mahout uses Naive Bayes, should help usher in predictive analytics for social CRM, you just need buckets of data, the cost is now affordable. Here’s a video pic recap

  16. […] more agility is required within the context and experience expectation of the customer. Such flexibility and agility can be achieved by empowering Knowledge Workers to decide on and carry out tasks in line with the […]

  17. […] L’implication d’un RSE me paraît naturelle dans de tels contextes ! Le potentiel de l’ACM semble promoteur pour définir le cadre de l’entreprise collaborative et agile où processus, informations, […]