Walmart’s Content Management Journey #KMWorld

Speakers: Amber Simpson (Senior Manager, Learning & Development, Walmart) and Todd Fahlberg (Senior KM Consultant, Enterprise Knowledge)

Session Description:

[These are my notes from the KMWorld Connect 2021 Conference. I’m publishing them as soon as possible after the end of a session so they may contain the occasional typographical or grammatical error. Please excuse those. To the extent I’ve made any editorial comments, I’ve shown those in brackets.]


  • Initial Challenges
    • They deeded efficiencies to create, maintain and sunset content
    • BUT there was a lack of accessibility and awareness of existing content, they saw a lot of unnecessary time and money spent recreating, manually managing, duplicating, searching for content. PLUS there was an overliance on the “powers” of Excel. They were using Excel to do more that it should.
    • How to create associate experience and learning data as a service
      • enabling the delivery of learning to Walmart associates before they know they need it
      • providing recommended content and personalized eperience
      • needing a single centralized location for indexing learning content
    • How to align their taxonomies so that machine learning could ingest and parse the language(s) of Walmart.
  • Lessons Learned
    • “If we’re going to do anything, we first must be able to show value and how we’re giving back to the organization.”
    • “Have to stop referring to KM as a separate entity, should instead align with Ways of Working.”
    • “Keep things Sesame Street Simple”
  • Their Initial Content Management Strategy Roadmap had several work streams:
    • LCD MVP & taxonomy implementation
    • Findability (Solr implementation)
    • Knowledge Graph implementation
    • Analytics reporting platform
    • Business process improvement
    • Content governance
  • Their Simplified Implementation Roadmap:
    • Product discovery and analysis
      • define and validate product vision
      • conduct environment analysis
      • update architectural design
      • update content data model
    • Agile adoption
      • consolidate requirements and backlog to Jira
      • Modify agile processes to fit Walmart culture
    • UI/UX Improvements
      • conduct user interview
      • update site map and consolidate application screens
      • design and develop new search interface
    • Findability (Solr implementation)
    • Taxonomy (PoolParty implementation)
  • Digital Library — Their Learning Content Database (LCD)
    • Learning content database
      • Their LCD became the one source of truth. This enabled them to deliver high-quality learning content in the most efficient and reliable way, which resulted in time and cost savings.
      • Their LCD connects skill badges to learning content so they can track associate development, provide targeted learning to upskill employees, and identify gaps in content.
    • PoolParty
      • The LCD is their metadata hub
        • User-contributed metadata (manually added)
        • Derived metadata (harvested metadata)
        • Provenance & lineage metadata (metadata added by users and technology) — this is GOLD — it shows relationships, change history, and dependencies
      • PoolParty automated tagging, taxonomy/ontology management
        • the controlled vocabulary is stored in the LCD and applied automatically by PoolParty
    • Solr
      • indexes content from Walmart’s repositories
      • provides customizable options
  • Future Vision
    • They are transitioning from a relational database to a graph database
      • Their goal is to provide a semantic search experience. Solr, powered by Walmart’s Knowledge Graph and taxonomies, provides US Learning and its internal and external facing systems a Google-like search experience.
  • Recap of Lessons Learned
    • Keep it simple. What is understood is supported
    • Meet people where they are. Explain WIIFM — what’s in it for me — for them
    • Training, training, training.
    • Don’t try to boil the ocean
      • Clean data is a must but waiting for perfection will delay your start
      • Find a place where you can make an impact. Identify the problem and jump in.
    • Identify a trust partner/advisor who is willing to teach, mentor, and provide guidance
    • Stakeholder, business, and tech alignment
    • Know your organization’s culture and understand your customer
    • Show yourself GRACE!
  • Agile Content Management — They use agile methodology to ensure the millions of content items in their Learning Content Dabase are regularly reviewed and updated.

Content Catalysts

Knowledge managers  sometimes divide the world into two camps:  content creators (the folks on the front lines of an organization, as well as a small handful of knowledge managers with subject matter expertise) and content managers (the bulk of the knowledge managers and librarians). In this scheme, most knowledge managers are working well behind the front lines and feel best suited to the task of organizing content rather than creating it. And, the content they seek to organize is explicit knowledge.  In reading Nick Milton’s post, KM and content management – the turf war,  I was struck by the fact that the disputed turf at the heart of this war (i.e., explicit knowledge) is relatively small.  If you look at his Venn diagram, you’ll see he identifies three types of turf:  non-knowledge information, explicit knowledge and tacit knowledge.  The first usually is the domain of librarians, and the last the domain of knowledge managers.  Both areas provide ample challenges and rewards.  Yet the turf battles continue with respect to managing explicit knowledge.   When you consider whether many (or any) of us should be fighting over KM 1.0 efforts to create and organize document collections, the turf battle seems even more pointless.

Therefore, in an effort to shift the conversation, I’d propose to expand the roles available for information professionals.   What if we were to add a new category:  Content Catalysts?  Rick Ladd pointed to this function in his comment on my previous post, Librarians vs Knowledge Managers:

It seems to me we KM professionals have been saying for years that an organization’s most useful knowledge lies between the ears of our people; up to 80% (obviously an approximation) of the total available. What I’m seeing is the use of social media to discover, connect, build relationships . . . in other words, greasing the skids of close to real-time knowledge transfer . . . is transforming how we deal with information and knowledge.

I’m of the opinion most value – at this time – lies in developing those “social” capabilities in an organization. Not to say managing the explicit knowledge assets isn’t important (precedent and all that comes with it isn’t going to go away, whether it’s judicial or the laws of physics); merely that connecting people to people and facilitating their ability to make sense of their collective information/knowledge, etc. is likely to have a bigger payoff than organizing our explicit assets.

This seems like a more productive route for knowledge management.  As information explodes around us, we’re finding that we’re able to corral less and less of it.  Our best hope is to have search engines that find what is necessary in the moment of need while we spend our time as Content Catalysts ensuring that information flows rapidly and is shared easily within our organizations.  (This is the promise of Enterprise 2.0.) It seems to me that knowledge managers should be able to add a great deal of value to their organizations as Content Catalysts, without the distraction of engaging in outdated battles over questionable turf.  What do you think?

[Photo Credit: Thomas Hawk]


How We Manage Junk

A person stuck in KM1.0 spends much of their time maintaining their various content repositories.  Unfortunately, it can be hard to keep up with those maintenance tasks.  And if you fall behind, it becomes increasingly difficult to retrieve your content — even if that content is pure gold.  If you have dross mixed in with the gold and you’re not a diligent housekeeper, then you have a real mess on your hands.

If you spend too much time with database-loving people, it becomes far too easy to think of those repositories in clinical terms, as if they were pristine, properly structured and well-maintained.  However, that doesn’t reflect reality and it doesn’t reflect how most of us interact with information.  So I thought I might provide three alternative views of your content organizational scheme:
The Junk Yard:

[Photo Credit:  sigma, Creative Commons license]

The Junk Shop:

















[Photo Credit:  hortulus, Creative Commons license]



Vintage Chic:






[Photo Credit:  Tammy Manet, Creative Commons license]

Which is your reality?  Do you need to do anything about it?


A Place for Every Thing

There is an old adage: “A place for every thing and every thing in its place.” And yet, if you’ve ever shared space with another human being, you know how hard it can be to (i) identify that one place and (ii) get everyone to put each thing in its “proper” place. (As I write, I’m staring at a bottle of dish washing liquid that always ends up on the “wrong” side of the kitchen sink, despite my best efforts!)

So why is it we think we can do better in our law firm knowledge management programs? The reality is that people often define the “proper place” for content differently. You only have to look at the variations in social bookmarking to see this. So, for example, instead of creating a rigid top-down taxonomy that imposes a regime of a single place for each thing (and then devoting the necessary resources on enforcement), why not spend your energy creating systems that allow users to organize the content as they see fit? After all, the point is to enable their easy use of the content — it really isn’t about ensuring that they find and use that content only in particular places.

At the end of the day, the purpose of the adage of one thing/one place is to eliminate options so that that you always know where to find your keys, your wallet, your cellphone, etc.  With the advanced search tools available today, we don’t need to worry about this quite the same way when it comes to electronic content. So instead of enforcing a single way of doing things, meet your users where they are. I guarantee they’ll be happier  — and then so will you.


Context Matters

Mark McGuinness would like you to test your perception. Take a moment to read his post, Are You Trapped in Black-and-White Thinking, and then tell me which square is darker. He uses this test to illustrate his concern about our tendency to think concretely in black and white terms — no ambiguities, no shades of gray. In his view, this rigid approach cuts off creativity at the knees.

Spending a little time thinking about the restrictive lenses we use in daily life and how they affect our ability to think creatively is definitely a useful exercise. However, I’d also like to draw your attention to another aspect of this black/white test: Context Matters. When we saw the square surrounded by dark squares, we assumed it was lighter than it was. Equally, when we saw the square surrounded by light squares, we assumed it was darker than it was. In each case, however, the squares in question were exactly the same color. It was the immediate context of those squares that led us to perceive them differently. But, when we look at both of those squares in the context of the entire board, we get a much clearer sense of their true color.

Without getting carried away by this optical illusion, it is helpful to think about the value of context in knowledge management where there isn’t always an objectively right answer. There is a lesson here for folks who think super-search is the ultimate answer to knowledge management challenges or who believe that extracting explicit knowledge and warehousing it in a KM repository is the best solution. With each of these approaches you run the risk of losing valuable context that can help the user make sound judgments about the content in question. In the language of law firm knowledge management, model document X may be the perfect precedent in situation Y and a complete disaster in situation Z. The only way you are going to know for sure is by looking at document X in context.

As you think about your approach to content, think about whether you’re doing a good job of providing the context necessary to allow users to make wise decisions about the content they choose to use. Context matters.

[Thanks to ProBlogger, Darren Rowse, for pointing out Mark McGuinness’ post.]


The Right Stuff

A sidebar e-mail conversation with some thoughtful readers of my earlier post, Is Your Knowledge Management Strategic, raised the following interesting question: How do you find out if you have the necessary content and processes without doing a full-blown knowledge audit, yet how do you avoid the dangers of the knowledge audit?

Dangers, you ask? In the wrong hands, a knowledge audit is a bit like that old definition of a sailboat (i.e., “a hole in the ocean into which you pour buckets of money”). Knowledge audits have a way of absorbing and squandering resources. And they can go on and on.

Perhaps the right approach is more triage than audit. Once you know your business strategy, and have determined with the business folks what knowledge (both internal and external) is necessary to implement that strategy, then do a quick and focused check to see if you have the required content in logical places. Contrast this with a full-blown knowledge audit, which tends to be more like a comprehensive inventory of all your content and processes — whether or not they are pertinent to your business strategy.

The full-blown organizational knowledge audit is probably best undertaken just before you switch jobs or retire so that you at least will know the scope of the legacy you leave to your organization. At any other time, does it really matter how much content you have in your knowledge management system if you don’t have the critical bits of content that are required for operational success?

PS: If you’re still tempted to tackle a full-blown comprehensive knowledge audit, start with the interesting resources provided via the ever-helpful Knowledge Flow.


Have You Contributed Enough to Retire?

My last post talked about the dangers of letting baby boomers slip out the door without first ensuring that they had left in their firm’s KM system “knowledge nuggets” containing their accumulated experience and learning.  That post was intended to be a warning to knowledge managers.  But perhaps we should launch a parallel appeal to baby boomers along the following lines:

The American Association of Retired Persons (AARP) provides a handy online tool to help you calculate whether you have contributed enough to your retirement savings accounts to retire comfortably.  There is another non-cash account that you also need to contribute to before retirement.  Before you head out of your office door for the last time, please answer these questions:
Have you contributed enough to the firm’s knowledge management system?
And, how do you know what is enough?
  –  Do your contributions to the KM system represent the best of your work product and learning over the time you’ve been affiliated with this firm?  
    –  Are they a suitable legacy of your work?
If individual lawyers paid as much attention to their contributions to the knowledge management system as they did to their contributions to their retirement accounts, we would have much more content in our KM systems.  And, if those contributions were made with a view to ensuring a suitable professional legacy for the contributor, we would have high quality content in those KM systems.  
It’s time to set up KM retirement accounts for every lawyer in your firm.  In fact, it’s time to set up KM retirement accounts for every knowledge worker within your firm.  No firm can afford to let valuable knowledge slip out the door with its retirees — regardless of whether those retirees are lawyers or non-legal professionals.  

What’s Your O-Ring?

The destruction of the space shuttle Challenger in 1986 was one of those events that seared the nation. After the inquiry into the accident was complete, we learned that a small defective O-ring seal had led to the disaster. Similarly, when the space shuttle Columbia exploded in mid-air in 2003, we learned that a piece of foam insulation had fallen off the spacecraft during the launch and had struck a protective tile on the left wing, thereby damaging the shuttle’s thermal protection system. During Columbia’s re-entry into earth’s atmosphere, the damaged tile allowed hot gases to enter and destroy both the wing and, ultimately, the space shuttle. More recently and a little closer to the ground, Boston learned to its sorrow that using the wrong glue in its Big Dig project could cause several sections of a tunnel’s ceiling to fall — crushing a car and killing its passenger. In New York City, officials are wondering if the local practice of pouring concrete in two days rather than five during the winter months has led to a disproportionate increase in the number of construction site accidents.

In each case, it’s a seemingly simple thing that is the point of failure.

What’s your point of failure? Do you have critical business processes that can be completely undermined by a single design flaw?

One classic KM system design flaw has to do with gathering content. You may have technology that is a thing of beauty, but if content collection depends on the voluntary participation of your knowledge workers, that’s a flaw that could be fatal. They often are (or feel that they are) too busy to contribute. Or, as I discussed in my recent posts on capturing content and hardwiring KM into client engagements, if content gathering is not baked into your client engagement process, but is shunted off to the side as something that is nice to have rather than necessary, your content collection business process will fall apart.

Another classic point of failure in law firm knowledge management is appointing KM “guardians” who have the responsibility of ensuring the purity and currency of the KM collection. What happens when those guardians become bottlenecks? That’s another design flaw.

Yet another point of failure lurks in the Outlook folders of every lawyer in your firm. If they don’t faithfully send copies of their e-mail correspondence to your central records system, how will your firm have an adequate record of your client engagements?

So how do you prevent these potential O-Ring disasters within your systems? The best approach seems to be extremely pessimistic planning. In the midst of project planning it’s so easy to get caught up in the excitement and anticipation. That’s when you make the mistake of believing that reality will follow your planning. The better approach is to ask at each critical juncture of the design phase: What if this doesn’t work as planned? What if users don’t comply? If they do not comply, what are they most likely to do? These questions should help you identify potential points of failure.

After launch, repeatedly tracking user behavior and system results can help identify where reality departed from planning. Sometimes the results are serendipitous. More often, these departures from plan indicate a point of failure that wasn’t properly analyzed and addressed in the planning phase. Then, the next question is: Should we fix this? That’s when you get into some interesting cost-benefit analysis. However, if the design flaw goes to the heart of a critical business process do you really have a choice? Just ask the folks at NASA.


Finding the Right Stuff

The best search engine in the world cannot convert garbage into useful content.

So, before you spend megabucks on the latest cool search tool, think about what repository you’re searching. If your office is anything like most offices, you’ve got tons of ephemera — stuff that probably isn’t going to matter 30 minutes from now — and tons of materials that really ought not to see the light of day. However, a decent search engine has the effect of shining a klieg light on the shortcomings of your content.

In fairness, a great search engine can help filter out much of the suboptimal content, but if there isn’t anything worth finding, what have you really achieved?

One solution is to stack the deck: make sure you put a goodly amount of useful content into your repository before you turn the search engine on. And what constitutes “useful content”? Not just something that someone might possibly need someday. (That’s the sort of rationale the turns a document repository into a cluttered and confusing packrat hell.) “Useful content” is content that is vetted: someone has reviewed it and given it a seal of approval or it’s been used with good results enough times for you to know it’s a great resource.

With that kind of content, you’re going to look like a genius every time the search engine does its thing. How often can you say that about a KM project?

For more information on the benefits of vetted content, see my article “Loading the Deck” in ILTA’s 2007 Knowledge Management White Paper.


Capturing Content

In my last post I discussed Hardwiring KM Into Your Client Work as a way of improving your chances of actually capturing knowledge and making it available for reuse. The focus there was on specifically planning to capture knowledge and then organizing your project around that goal.

But how do you actually capture that knowledge? The good folks at HP have some suggestions in their presentation regarding using project management to advance knowledge management goals:

– Identify “checkpoints” or “milestones” within each project at which you can stop and take stock.
– At each of these points, ask “What lessons did we learn?” And, “How have we documented this?” And, above all, “Is this information in a centrally accessible system where it can benefit the firm and our clients?”
Within a law firm, these checkpoints or milestones arise naturally: when we win a client engagement, when our clients sign a deal, when a deal closes, etc. The key is to remember to pause momentarily at those points and take stock of the knowledge generated or improved up to that point in the project. And then, to make sure that the new or improved knowledge has been documented and contributed to the KM system where it can be of use on other client matters. By baking this step into our deal timetables and responsibility lists, we increase the chances of actually doing it. In addition, we need to hold the partner and senior associate on every matter responsible for this and then ensure they are judged and rewarded based on the contributions of their team to the firm’s learning. It’s their leadership that encourages team members to take the time to identify and contribute content. It’s their leadership that brings about the necessary cultural change — one project at a time.

The key is to make knowledge capture and reuse an integral part of the work flow of every client engagement. You’ll know you’ve accomplished something when it is no longer viewed as an alien, artificial activity, but rather as a perfectly normal and rational thing to do. That’s a KM success.