Making Knowledge Work for You: Interview with David Kay, KM Guru

When I analyzed all of my member inquiries for 2009, 31%–nearly a third–related to knowledge management, search, and web self-service. You’ve all heard me complain about how shockingly low self-service success rates are, with the industry average dipping down to 40% at the beginning of 2010. Clearly this is an area that companies haven’t figured out, even though spending on knowledge technology has been strong for the last decade and a half.

Wouldn’t it be nice to spend a day with someone who has all the answers? Here’s your chance. At our Fall TSW conference in Las Vegas, David Kay will present a full day workshop, “Making Knowledge Work for You: Best Practices in Support KM,” Monday, October 18 from 8am-2pm. David is founder of DB Kay & Assoc, and co author of Collective Wisdom: Transforming Support with Knowledge, a must read for all support knowledge workers. David is my go-to guy for all things knowledge, and I wanted to take this opportunity to interview him about his workshop, and why KM is such a challenge for today’s support organizations.

John Ragsdale: What a pleasure to speak with you David! I’m excited about your upcoming KM Workshop, “Making Knowledge Work for You: Best Practices in Support KM.” You have spent time with many SSPA and TSIA members over the years, helping them create new knowledge processes and implement new tools. What would you say are the 3 most common problems you find regarding KM initiatives within tech support?

David Kay: I’m excited about the workshop at the Las Vegas TSW conference!  We always have great conversations. The three problems I see the most have changed over the past several years.  It used to be that KM efforts lacked executive sponsorship, but–as your numbers suggest–most support executives understand the benefits of knowledge, so that’s less of a problem.  What we’re seeing instead is resistance among operational managers, measurement challenges, and a paralyzing fear of being wrong.

I feel for the operational managers I work with.  Every day, some executive stops by and says, “hey, do this one other new thing,” and then walks away before the poor manager can ask, “what do I get to STOP doing?”  With rising pressure to serve more customers, with more complex issues, with constrained resources, it’s not surprising to me that they see knowledge management as just one more thing they don’t have time to do.

The reality, of course, is that knowledge management will really streamline the job of support.  Done right. it makes the job not only more efficient, but more fun:  who wants to answer the same question over and over again?  But line managers will need some convincing, and expecting that line managers will support KM just because they’re told to is a mistake.

Knowledge measures are different from other support measures.  Typical measures are straightforward:  all things being equal, we should close more cases per person, resolve cases more quickly, and get higher CSAT scores.  But when it comes to knowledge, numbers don’t tell the whole story.  Is authoring more content good?  Yes…but only if it’s needed, and if it’s findable, usable, accurate, and timely–otherwise more content is actually bad.  Knowledge measures require a mindset shift.

Finally, there’s something about writing something down in a knowledgebase that makes people just a little…crazy.  Say someone to a customer, OK…write it in an email…fine…but write it in the knowledgebase?  EMERGENCY!  EMERGENCY!  Someone’s going to take their entire network down!  Let’s get 13 subject experts to review it first, with a side trip to Marketing and Legal.  Never mind that, by the time the content goes through its review process, we’re going to be shipping the next major product release.

I don’t mean to be cavalier about quality, and much of the work that Jennifer and I do with clients involves building quality and continuous improvement into the KM process.  But even if your knowledgebase is 100% perfect today, customers will still misinterpret it, and it’ll be obsolete tomorrow morning.  Perfection isn’t an option, any more than it is in product development or in case resolution.  The goal is the most value for customers, and efficiency for ourselves.  And I find this requires taking a deep cleansing breath and giving up the illusion that perfection is an option, while figuring out how to get things as right as possible, as quickly as possible.

Ragsdale: I’ve been fielding KM questions since I started as an analyst in 2000, and a decade later, with all the new tools, social media, and Gen Y, the questions haven’t changed that much. Why is KM so hard for companies? What part of “people, process, technology” causes the most grief?

Kay: It’s the intersection of those–really, this issue is, why are we doing support in the first place?  TSIA and The Complexity Avalanche are really helping here, by broadening our view of support.  We’re not just in the business of closing cases.  Our real job is creating value and loyalty by helping people use our products and services.  Closing cases is just one means to this end–and a pretty reactive means, at that.

So KM success requires that people get what you, JB, and many TSW keynoters are talking about.  If support organizations believe we’re in the business of creating customer value, then knowledge is the key asset for doing that, whether it’s through closing cases, self-service, product improvements, better PS engagements, you name it.  If you believe we’re all about FCR, backlogs, and case SLAs, it’s going to be pretty hard to view knowledge as anything other than second priority–and it just won’t happen.  Not really.

Ragsdale: I don’t want to open a big can of worms here, but as you already said, “Knowledge measures are different from other support measures.”  What metrics should be tracked, how do you calculate them, what are realistic goals, etc.? Things like self service success, hit rate (percent of issues resolved by existing knowledge), satisfaction with self-service content, etc., are all pretty ‘grey’ metrics, meaning how you track them differs from company to company–making benchmarking hard. What are the key KM metrics you think companies should track?

Kay: What, you don’t believe in 30 page blog posts? The really quick version is that measures need to be tailored to your business goals, but almost all KM initiatives should pay attention to measures that show whether people are participating, for example, search rates, link rates (also called “participation”), certification profiles, create and edit rates.  These are measures that we don’t put goals on, but track and look for variance over time and by individual.  These can suggest when programs are thriving, when they’re hitting challenges, or when individuals need some coaching.

They also should look at self-service value.  Here, I like to look at contact rates (sometimes called CPx), estimated self-service success, and estimated deflection rates.  A number of companies are using the deflection rate calculations from our paper TSIA published here with good success.

Knowledge quality is important:  citations, sampled quality, and maturity, measured as the ratio of reused knowledge to created knowledge.  A mature knowledgebase has a relatively high reuse to create ratio.

Finally, knowledge should improve all the standard operational metrics organizations track today.  Productivity should go up, resolution times should go down, satisfaction should go up.  That is, until self-service really kicks in…in which case you’ll be closing fewer cases, but they’ll be harder, so per-case metrics will look like they’re going in the wrong direction.  That’s actually a good thing.

As for benchmarking, I agree.  We’re working hard to standardize some of these measures, especially on the self-service side.  Perhaps this is an area where we can use the bully pulpit of the TSIA Benchmark Survey to drive standardization into the practitioner and the vendor communities.

Obviously, this is a big topic.  People who are interested should definitely join us in the workshop, because we’ll be spending quality time on just these points.

Ragsdale: That’s a good  segue back to your workshop. Could you talk a little bit about how the workshop is structured, and what you will cover during the day?

Kay: Sure!  It’s a full day–we cover lots of material, but we do it in a very interactive way:  no death by PowerPoint.  Our big topics are:

  • Models for doing support knowledge, and the advantages (and challenges) of Knowledge-Centered Support
  • How to structure content for rapid capture and easy reuse
  • How to make content creation, improvement, and reuse part of the support staffer’s job
  • Publication processes
  • How to measure and encourage knowledge work
  • How to help people with the mindset shifts we’ve been talking about
  • How to make it real in your environment, with real technology, busy teams, and financial constraints

Ragsdale: Before we go I have to ask a technology question. From my perspective, I’m seeing less focus on knowledgebases, and more focus on intelligent (and mobile) search, so support techs and customers can easily find information from any source–including social media channels.  Having a single knowledgebase–which was the standard for so long–just seems unrealistic for large global enterprises. Do you see the new breed of search technology changing the way companies select and implement KB tools? Or is the KB still king in the companies you work with?

Kay: I’ve long been a proponent of what you’ve been calling the Intelligent Search approach to KM.  One-stop shopping is so important, and with the proliferation of social content, it’s only getting more important.

I see customers interested in this concept, but they’re also lured by CRM vendors, some of which are making a big push in knowledge, whose knowledgebase modules don’t support cross-source search.  They’re also very interested in going to the Cloud, but for technical reasons it’s hard to do Intelligent Search today well from the cloud.  So I see a real tension here.

Ultimately, the customer and user experience should be the driver here, and that means one search box no matter where the content lives.  I think the closed-box KB is an evolutionary dead end, and I trust that the smart vendors out there will figure out how to get outside the box, even through the cloud.

Ragsdale: Thank you David Kay for sitting down with me and sharing your collective wisdom with my readers! I look forward to seeing you in Vegas!

Kay: John, as always, it’s a pleasure, and I look forward to continuing the conversation at the TSW conference.  Jacks or higher to open, please.

If you have any comments or questions for myself or David, please add them here or shoot me an email. And as always, thanks for reading!

Explore posts in the same categories: knowledge management, Technology

Tags: , , ,

You can comment below, or link to this permanent URL from your own site.

11 Comments on “Making Knowledge Work for You: Interview with David Kay, KM Guru”

  1. Radha Says:

    Hi John, David- thanks for the post. I have some questions and hoping you can answer:
    (1) The role of content team in KCS

    (2) How many people do we need in content team

    (3) How do you go about hiring? – Hire documentation experts or rotate tech support team members or recruit from tech support

    • Dbkayanda Says:

      Radha –

      Good to hear from you! And thanks for the good questions.

      The role of the content team changes dramatically in KCS. Typically, these teams were *the* people creating content. Almost inevitably, this means they became a bottleneck.

      In KCS, the job of basic content production largely goes away, because engineers themselves are doing the content capture and maintenance / improvement in the workflow. The roles that are left are higher value-add.

      First, there’s the program manager. She’s not creating content, but she’s making sure that all the other processes are working so that the right content is created, maintained, and reused. She’s also responsible for the Content Standard, solution quality sampling, and the certification model, each of which are critical for knowledge quality.

      Then there are the knowledge domain experts. These are usually escalation engineers / analysts who take some of their time overseeing the health of a particular topical area of the knowledgebase. We find having this be not a full time job is actually good: working escalations keeps their technical knowledge fresh.

      Finally, especially in a larger environment, there can be a role for dedicated content professionals creating value-added content of various kinds: for example, a troubleshooting guide for a symptom that has multiple causes and resolutions, or a multimedia how-to guide. KCS calls this “B Loop content,” and I like to think of it as value-added content–stuff that can’t be created in the workflow, but is created based on demand and patterns of usage in existing content. But many successful KCS implementations don’t have this as a dedicated role.

      So, how many people do you need on a content team, per se? Zero, at least to start. But there are these other higher value roles to staff. Often the program manager is a successful leader in support who has a particular passion for knowledge, and as I mentioned the knowledge domain experts are subject experts who have some time carved out for B-loop work.

      Let me know if this makes sense…

  2. Greeting John and David:

    An area I was surprised that you will not address in your upcoming workshop is how to get broad stakeholder involvement in creating and refining KB content to reflect and support real-world usage.

    Without active and perpetual engagement from relevant internal and external stakeholders to evolve KB content I do not believe it can realistically provide sustainable value to the community.

    The problem with most KB technologies is that they do not include the necessary metrics and supporting infrastructure required to facilitate, motivate, recognize and reward sustainable stakeholder involvement to evolve existing KB content.

    Community sites integrated with KBs can certainly be a source for new KB content, but once the content gets in the KB the vital collaborative evolution of the content will die and much of the content will quickly stagnate.

    More so than general purpose content found with unified search or open community generated content, editorially controlled KB content can be optimized for support environments where consumers and support staff expect and need to quickly get to accurate, succinct, anticipatory, complete and well written content that the consumer can count on and that the brand will stand behind. Regardless of the fine print, brands who delegate support to online communities are still held accountable for the customer experience and their satisfaction.

    Unfortunately, in the eyes of many, KB’s have been rightfully relegated to old-school technology that is void of the wisdom of the crowd.

    What say you?

    Chuck Van Court
    CEO and Founder

    • Dbkayanda Says:

      Chuck –

      You cover lots of interesting points.

      (1) You’re absolutely right that broad stakeholder engagement is a critical part of the program, and it’s an important part of the workshop, too. I probably paid it short shrift when I said “How to help people with the mindset shifts we’ve been talking about.”

      (2) The idea of community-based knowledge is a very important one – as a matter of fact, Robert Rose of Symantec refers to this KCS stuff as “knowledge communities.” Extending that to the broader customer community is very exciting to me, but outside of open source and relatively modest implementations at the-company-formerly-known-as-Sun, Cisco, Novell, and a few others, the industry seems hesitant to go there. I hope we get over our hesitation soon: even the New York Times lets people comment on stories!


  3. Hi David:

    Thanks for the reply.

    How important a part do you believe a reputation engine plays a part in measuring contributions, identifying specific expertise and motivating broad internal and external stakeholder participation in creating and evolving Knowledge Base content?

    Thanks, Chuck

    • Dbkayanda Says:

      Chuck –

      Our experience has been that reputation is an extremely important driver of engagement and quality, internal and external.

      Today, this reputation isn’t the product of a single engine, but a combination of reports, analytics, and the ineffable skein of human relationships that happen online and in the workplace.

      I encourage vendors to explore ways of systematizing reputation, and see lots of promising research. Reputation matters, whether or not it’s captured in a number, a vector, or any other computable artifact.


  4. David:

    Glad to see that we are on the same page on this.

    A reputation engine is built into our knowledge base and we actually have a patent application outstanding for it.

    The impetus for our reputation engine was that our customers wanted to have the ability to measure and motivate specific contributions to their knowledge base from internal and external stakeholders. They also wanted to define specific expertise rather than just expertise across all their KB content.

    You would be surprised how competitive people will get to earn Contribution Points!


    • Dbkayanda Says:

      Chuck –

      I’m delighted to hear that you’re working on this.

      But beware rolling performance or reputation into a single number — especially one based on activities. I would NOT be surprised to see how competitive people will be to earn Contribution Points, and any single index will be a target for gaming.

      Even Jeff Harling at Avaya, in his excellent talk at TSW last year, was careful to caveat their “Quarterback Rating” (his analogy) as a great tool for deciding whose details needed looking at more closely, not the number to incent.

      I’m continuing to try to swear off walls of words in these things — if this doesn’t make sense, grab me at TSW and we can talk it through.

      As for patents on reputation engines, don’t get me started 🙂 (I say this as a guy who holds 5 issued KM patents, and wishes he didn’t.)


      • Hi David:

        I too hate SW patents and wish they did not exists, but while they do we must begrudgingly play the game. My god, RightNow has a patent for incrementing a relevance calculation used in ordering search results each time a person views an item. I guess we are just lucky that the person who developed drop-down menus did not get the patent.

        Gaming can certainly be an issue, but we have several components to control and identify when it occurs.

        The ongoing challenge is engineering a reputation that is easily understood so that it successfully motivates behavior while also protecting against people inappropriately gaining reputation.

        No solution will be perfect but I believe that some day all KBs will have some sort of reputation engine built into it to motivate behavior, identify skills and get broad community participation to perpetually evolve KB content.

        KB can serve a very important purpose, but only when they

  5. oops… where’s that edit button when you need it?

    when they …engage stakeholder participation.

    Best regards, Chuck

  6. Carli Says:

    Ecemnoios are in dire straits, but I can count on this!

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: