Why Exception Handling Should be the Rule

In the world of work, we encounter three primary tasks:

  • First, there are many processes that are, in fact, repeatable in the enterprise. Some examples: how we process orders, how we assemble products, how we deliver products to end customers.
  • Second, project work where the overall steps are repeatable but the ingredients are not. Examples: product development, managing marketing campaigns, executing a sale and the like.
  • Then there are those that aren’t exactly predictable: A question a prospect or customer may have before making a purchase decisions, a complex product that has customizable/subjective uses or accessories that work better with certain models. These come in both transactive/process as well as project flavors and almost always show up unannounced.

The Exception Misconception


We generally have a good handle on the first, when things go by the playbook. The second depends on our management ability to set direction and then hire, coach, lead and foster collaboration to execute well.

But as to the third, the sheer impracticality of channeling exceptions in any scalable way to get the right answers has plagued organizations for ever. Each exception requires a different set of experts or problem owners, some known but most unknown, and often spread across a global footprint at large organizations. Historically, it seemed far more practical and scalable to just stick to the book and assume the risk that once in a while, things will go horribly wrong.

In fact, if you think about it, when you used the phrase “Honey I had a crappy day at the office today” it was unequivocally a result of an exception that occurred and your inability to handle it in any painless way.

The thing is, completing business activities efficiently and to the customers’ (internal or external) satisfaction, is not just a volume game. Exceptions maybe just that – stuff that happens less frequently. But the mistake we often inadvertently make is equating lesser frequency to relatively lesser criticality in terms of loss or risk.

What’s worse, in the age of the social web, those very ‘once in a while’ instances cannot be shoved under the carpet. The social web has a way of exposing and then amplifying our ineffective handling of exceptions. And no, this isn’t just about failed social media campaigns or online service. Its when the social web gets wind of something seemingly dopey we did in the offline world as well. And whilst there’s no trending data suggesting that negative press on the social web instills long term financial damage, find me one executive who wants to own or be a victim of the PR disaster that resulted in lower revenue or operating margins this quarter.

Its about Flow, not Values

So why is it then, that we pay more attention to facilitating repeatable process, over exception handling. It’s because machines can emulate the intended flow of a repeatable process (you’re welcome ERP, SCM, CRM vendors). In comparison, exception handling looks like the wild west. Yet the risk profile of screwing up either one can be exactly the same.

Adrian C. Ott at Harvard Business Review has a nice article stating that rigid scorecards and metrics are the problem. She cites a Delta Airlines case, where (contrary to a deal worked out), front line staff insisted on charging Staff Sgts. Fred Hilliker and Robert O’Hair of the Military, $200 each for carry-on luggage when returning from the battlefield. This resulted in terrible PR, including this YouTube video:



Adrian takes on the issue of rigid top down management practices saying:

It would be wrong to place all the blame on workers for their failure to take discretionary steps. The blame lies with management that sets rigid rules and metrics that disable employee judgment and create so many approval hurdles for mundane decisions.

She suggests that the root cause is a values issue.

The blame for poor employee action should be placed on the managers who set rigid metrics, and fail to invest in employees. Yet customers need more judgment, not less, from the employees they come in contact with. When customers contact a call center, it’s because there is an exception within the existing process and they need judgment that only employees can provide. Corporations need to build guidelines and values — not absolute rules and measures. "Doing what’s right for the customer" is a value that can drive appropriate action. Judgment requires coaching, practice and training.

Fair point. No doubt that loosening up things will help. But I’m just not so sure that we’ll manage exceptions well with values and guidelines alone.

Customers certainly need more judgment from us. But the answer won’t come from values alone. That’s just dreamy stuff and an expensive change management bill if the pipes don’t exist to channel those values. And I’d like to believe that as individuals we, for the most part want do right by colleagues and customers. The reality is that we’ve spent decades blaming middle and senior managers who get sandwiched everyday between strategy and delivery. Our process centric world no doubt solves 90% of issues by volume, but provides neither line employees nor middle managers with the facility to reach across disparate organization (expertise finding), loop in the best people (wrapping around problem virtually), use data and content needed (in context), to address exceptions.

Consider the Connected Enterprise

It’s, in fact, the lack of internal wiring of people (as opposed to systems) across our organizations that’s the core problem. Even if you have a mechanism to spot exceptions early (which most of us don’t), folks on the front line at large organizations often don’t have a clue as to who to approach or collaborate with to handle exceptions, other than their direct chain of command.

Thanks to newer fluid, social and collaborative approaches and a vast array of relatively simpler technology, finally solutions exist to reduce the likelihood of getting Delta-ized via a decisively built but loosely coupled collaborative fabric.

We’re seeing organizations from Financial  Services to Institutional Banking to Healthcare and Hi-Tech starting to connect silo’d employee, customer and partner ecosystems, so that questions, knowledge, insight and action can find the right minds and generate the best answers to exceptions.  And yes, all of that with an audit trail so we know who said what.

Having worked with large organizations on such problems for over a decade I can tell you that this was extremely difficult and expensive to get right in the past. Back then we twisted and stretched technology to do unnatural things. Today, innovations in enterprise social and collaboration technology, along with seasoned, purpose driven strategy, planning and leadership allow us to deal with potentially expensive or embarrassing exceptions almost as effectively as plain vanilla process. And compared to those prices you’ve paid for ERP, CRM and SCM technology with 3X implementation multiples to streamline repeatable process, the technology that helps take care of non repeatable processes (including exception handling) comes at a tiny tiny fraction of the cost.

Look, collaboration is not the answer to every problem. But by the same token, I’m more convinced everyday that the unwavering belief in the notion that there is 100% repeatability in most processes is grossly overestimated and even downright dangerous. Even those process considered most straight forward often beg for a discussion when a fork in the road presents itself. I’ve long said that what rigid process systems are missing is a giant Discuss button that sits right between Submit and Cancel buttons that govern what in reality is not a very black and white day in the office. Same applies for face to face customer interaction that’s otherwise governed by rigid protocol, as in the case of Delta Airlines..

And so, the question is, as executives, will you wait to act in the aftermath of a very prickly situation or will you preemptively ensure that on your ship and on your watch, exception handling is the rule.


7/31/11: One thing I should have added to this post: The conversation around exceptions can’t be complete without calling attention to the work of Sig Rinde who have done intensive thinking on the subject of what he called Barely Repeatable Process. Lots on this topic on Sig’s blog.

Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Post Details
  • Vijay Vijayasankar
    Jul 22, 2011

    Very nicely done, Sameer

    It needs innovation on process design which is easier said than done. Workflows are generally rigid, with a clear way to determine the recepient. However, such workflows do not cater to the fact that most exceptions just cannot be predicted upfront. So BPM needs a flexible mode of working, which allows ad-hoc change in direction. Most tools just cannot do it – and a complex set of collaboaration, analytics and just plain old “ability to delegate” needs to co-exist to make it work. 

  • Esteban Kolsky
    Jul 22, 2011

    very, very, very nicely said.  good job.  couldn’t have it said it better myself (except i would’ve said 80% are normal and 20% are exceptions, but I am a customer service guy — so that is the difference).

  • Sameer Patel
    Jul 22, 2011

    I agree – its changes in process thinking and few people know how hard that is as you do. But I also think its a change in management thinking that will enable loose coupling of people as required by process. The fact that we have anointed people from start to finish brings a lot of risk. Yes, its a great way to know whose throat to choke but if things go horribly wrong, firing 1-2-5 people still wont make up for the damage caused. Its starting to happen in some cases. And the good news is that its moved beyond early adopter organizations.

  • Philippe Cases
    Jul 22, 2011

    I am still waiting for a true object oriented technology that  couples really loosely business processes and enable this type of flexibility. Agent was supposed to be just that but didn’t fullfil on its premise. At this point, it is just management commitment to manage the uncertainty.

  • Esteban Kolsky
    Jul 22, 2011


    I don’t disagree with what you are saying, but you are caught up in the how-to, the technology part of doing this.  Handling exceptions is totally un-related to technology and process – except where the process says “kick it to the human to handle”.  at that point, empowered (by management) employees use common sense and expertise to return an answer to the process, which then continues.

    it is giving ourselves the ability to do this (kick it to the human) and continue process afterwards that we are lacking as a corporate culture.  

    I half-agree with Sameer that is about management – it is about change in culture across all levels: management and workers.

    I do believe there is an opp in collaborative enterprises and future-of-work organizations to tackle and leverage this, but most people and organizations make it too complicated to even acknowledge.

    kick it to the human: it is not just an idea, it’s the best idea. :-)

  • Sameer Patel
    Jul 22, 2011

    When I say management issue, culture and change is a subset of that. Did not mean to imply top down/mandate only.

    Sent from my iPad

  • Charles Isaacs
    Jul 22, 2011

    Great post, great comments all around. Yes, kick it to the right human, at the right time, with the right skill, and the right tools at their disposal to provide quick resolution to that exception.

  • Sameer Patel
    Jul 22, 2011

    Re:80/20 – yep, every process is different. I have a follow up that digs into exception handling. Thanks for the kind words and the discussion.:)

  • Sameer Patel
    Jul 22, 2011

    Yeah great comments. Thanks Charlie.

    Sent from my iPad

  • A. Prem Kumar
    Jul 22, 2011

    Superb piece Sameer (as if you needed validation of that). I have read & talked on this topic for ages now … but from a BPM perspective (one vendor segment you missed). Though a BPMS is supposed to cater to business processes other than those covered by a ERP/SCM/CRM, I have often seen them being used for the aforementioned processes too.

    And in a BPMS you model & automate the process/work flows after doing rounds of discussions & interviews with the various stakeholders involved in the execution & management of them processes … with no scope for handling emergent / exception scenarios. Oft have I given a sheepish response to the end users that the problem they are facing is due to an exception scenario, and it is not built into the system … and that if they need to do it more often, why, they can give the requirements for the 3rd release cycle from now since the next one is almost due for production & the development has begun on the one after the next … or something to that effect.

    I started hating the term ‘exception’ at one time … thankfully I am at peace with the term now. Zen …

    But, social has come to torment me from the other end now … there is so much confusion in this unstructured world that it seems to beat the purpose of it all. Chris Lynch excellently captures the problem with unstructured data/content in his post on rethinking the social architecture in the enterprise [http://thelynchblog.com/2011/07/21/rethinking-social-architecture-in-the-enterprise/], but I doubt if the approach he chose is going to help much. I would like to take tibbr for a spin before I comment on it though.

    As with content categorization, so with process definitions … and ACM seems a probable solution there … need to take a dip in it too.

    We need systems that help us navigate from the structured end to the unstructured end of the spectrum (both content & process/work flows) and many other stations in between.

  • Emanuele Quintarelli
    Jul 22, 2011

    Sameer, superb thought provoking piece as always so I won’t spend too much time on congratulations!

    My two cents:

    This area is strategically important for so many organizations much more today than in the past given the speed of change in markets, customer expectations, global competition, knowledge obsolescence, shifting type of work most employees are called to do (hint more knowledge intensive tasks than transaction based tasks). Most organizations cannot simply afford to miss this boat for customer service, product innovation, reputation, efficiency reasons. To get a handle on how much this cultural gap is crucial John Hagel and John Seely Brown some year ago stated that “While 95% of IT investment goes to support business process (to drive down costs), most employee time isn’t spent on process but on exceptions to process”. The entire shift from push programs to pull approaches (see the Power of Pull book) is probably among the fundamental changes we’ll see in the coming decades.

    The second point is that it is not enough to reactively address exceptions. Exceptions are instead a huge opportunity for organizations. When you stop to see them as something exceptional and you accept that they are simply a big piece of the daily work, you can systematically take advantage of that thanks to approaches like Adaptive Case Management. Exceptions are the key to build a more agile, reactive, emergent, flexible, people motivating company. That’s mostly not a technology problem but more a cultural, management, leadership and process thing. Some resources here are Mastering the Unpredictable and Social BPM books by Keith Swenson.

  • Sameer Patel
    Jul 22, 2011

    Hi Philippe. 
    We’re starting to see the technology now now that allows a good portion of this. Business Process Management has been slow on the update as has any stand alone framework, relatively speaking. But traditional CRM and ERP software is responding – both from established vendors such as SAP or SalesForce.com (and others that haven’t announced publicly), and from nimble upstarts that embed collaboration inside process or at least sync events. But IMO, that’s the easier part.

    The first order of business though is for executives to realize that preemptive exception handling is as important as repeatable process completion -and to ensure that its infused in how the organization works – both operationally as well as technologically. 

    Based on our work, such thinking is slowly becoming part of process optimization or re-engineering to do list.

    Thanks for the comment.

  • Dave Duggal
    Jul 22, 2011

    Hi Sameer,

    Great post. Good perspective. A business with no structure is inchoate, formless jello. A business with rigid structures is in stasis, it’s stopped responding/adapting to its environment.

    I feel like we are trapped in a false binary between structured and unstructured work. First, are there any people other than us industry folks that make this distinction. No one goes to work and says “Let me do some structured work, then I’ll do some unstructured work…”. Work is a continuum, structure is a variable based on particulars, our work is more or less procedural based on what we are trying to do and our authority.

    Collaborative work is social, it’s only rigid processes that make it anti-social.

    Right now the vendor community is just doing brute force integration of social with ERP, but that’s fairly lite – sure, you can chat about an errant process, but the resolution still happens off line, not inflight. You can’t start from non-dynamic flowcharts and just plug in the latest social tool and say voila, we’re dynamic. Business needs a holistic approach to work.

    We’ve developed an approach (http://www.ideate.com) to injecting real-time dependencies into loosely-coupled emergent processes. It is a parallel development to Big Data technologies in that we are schema-less and solve for distributed joins, but we can do it in real-time for an individual interaction. We enable real-time feedback loops to bridge the divide between analytics (OLAP) and transactions (OLTP). It supports a self-organizing and regulating system that provides agility with governance.

    Thought it might be of interest to you.


  • Brian Magierski
    Jul 22, 2011

    Great post Sameer! 

    Esteban – totally agree with your comment here … “kick it to the human”. Of course this implies a collaborative organization that empowers their humans to make decisions and get things done of course, and equipped with the proper tools. 

    On the latter point, mobile technology and apps seem to be a great potential enabler here for both employee and customer empowerment. 

  • Sameer Patel
    Jul 22, 2011

    Thanks for the kind words and thoughtful comment, Emanuele.

    No arguments with your thinking. Though I’d like to expand on one point and that’s the Culture issue.

    Granted, its important but I think in the context of traditional approaches to leveraging collaboration where often it can look like a mission without a cause, exception handling (centered on good examples, similar to what you provide) is a far more tangible problem/opportunity set to start a
    practical discussion on where these new approaches to connecting people can
    make a real difference. And in my experience, the cultural challenges
    start to look far more palatable if the business problem is framed correctly to end users who want to get the job done right and are goaled to do so.  

    Thanks for providing references to the work of others. All super smart folks in my book.

  • Sameer Patel
    Jul 22, 2011

    With you on the mobile opportunity Brian. As we consider the larger swath of needed design and pipes to channel exceptions, Mobile is right up there. Thanks for the comment. 

  • Sameer Patel
    Jul 22, 2011

    Thanks Prem. I don’t think beating your self up for any inability in the past to make sense of this is worth it. The approach, mindset, technology was very different. And most orgs were high on the promised efficiencies of process automation, by going from paper to machine, or EDI to Web and on and on.

    Now we have new challenges and opportunities. The customer is implicitly demanding a new engagement model thanks to the social web. If there’s a willingness to re-think how we work, the approaches and tools to get there are available. 

    What an awesome time to be around to witness and to help improve the world of work. :)

  • Chris Lynch
    Jul 22, 2011

    Great stuff here. Glad to see this topic resurrected to reflect today’s current technology environment. With regards to exceptions, I think we’re going to have to get better at stream management. For instance, we can set up filters to monitor certain systems, people and updates, but I think we need a Priority Inbox way of bubbling up exceptions from those individual sources.

  • Sameer Patel
    Jul 22, 2011

    Thanks Chris. Stream management is but one interaction metaphor. We have to
    be open to the idea that the answers might come from a combination of huddle
    mechanisms. It could be an activity stream, it could be internal
    communities, it could even be finding people via rich profiles and picking
    up the phone or initiating a video chat. But I get your point re: smart prioritization.

  • Sameer Patel
    Jul 22, 2011

    Thanks for the comment, Dave. I would not go so far as saying the vendor community is using brute force. There’s very different approaches being applied today. I do agree that were at the early stages thought.  

  • Jeff Shuey
    Aug 9, 2011

    Great post. I have said for many years that a well designed system will have graceful exception handling.

  • Jeff Shuey
    Aug 9, 2011

    Great post. I have said for many years that a well designed system will have graceful exception handling.

  • Leave a Reply

    Your email address will not be published. Required fields are marked *

    You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>


    Get every new post on this blog delivered to your Inbox.

    Join other followers: