Organizations own a tremendous amount of information–some trivial and some critical. Whether a shoe store or a 10-person startup; an ad agency or a car dealership—there is a ton of data.
But storing shared information is a problem. How do you find what you need? Or, often, how do find out what you need? And if you are lucky enough to know what you need and where it resides, how do you find the right version of it?
As companies scale, this knowledge chaos intensifies. Early-stage priorities like sales and product development take precedence over organizing information. No centralized system exists for managing the exponentially growing volumes of data and documentation. It’s organizational debt that accumulates silently, a ticking time bomb.
Every Manager for Themselves
That debt comes in a variety of forms. Maybe there is no HR function. Payroll and benefits are automated or outsourced, but everything else is ad hoc. So, instead of a universal onboarding process, each manager handles it themselves. Some employees get comprehensive exposure to the strategy, product, and each function’s role—they meet every leader, learn where things are and are trained in how to get what they need.
Other people are added to payroll, meet their direct teammates, and get logins. They have no idea where anything is, or even what anything is.
The manifestations are myriad. Within one of my client’s organizations, each department used different software tools, preventing cross-functional collaboration. Sales couldn’t access the product roadmap, so reps constantly bugged the product team for updates. That was one of innumerable similar gaps.
They undertook a software audit–and discovered that they had licenses to 62 different platforms. They were spending over $43,000 a month on unused or duplicative software.
For another 125-person company spanning 7 countries, hiring itself was a free-for-all. Not only did they lack standardized onboarding, but there weren’t even processes for interviewing or crafting and sending offer letters—despite the conflicting rules within each country in which they operated. For every new hire someone just conjured their own way to do it.
Give Me a Day and I’ll Find It
The challenge is pervasive. Most of my clients use Google Drive. It has no inherent taxonomy. Everyone has their own private repository and can share links or save to a common folder. But unless someone creates a structure for common folders, the default is simply to allow people to “share” documents with each other through links. Those linked documents live in a folder on the recipient’s drive called “Links shared with you”.
Typically, there is no naming convention. Anybody searching for a particular document in its latest version must open multiple folders and files and attempt to glean if it’s what they need.
Customer success can’t access contracts, so they aren’t sure what the sales rep promised the (now) angry client. And the CRM is only nominally populated with notes.
The Growing Pile
Say you live in a house with 15 other people. Every bedroom has its own shoot into the shared basement laundry room into which dirty clothing drops. Every time someone takes off a piece of clothing and drops it in their shoot, it falls on top of whatever was there.
Everything piles up in the laundry room, in descending order from most recent.
Say you have a sudden business trip and need the blue shirt you wore on Monday. You confront the pile of week-old dirty laundry. You start to lift each item, putting them to one side as you eliminate them. Then you simply rummage, trying to spot anything blue. Scanning the now-smaller pile, you try to find shirt-like things. Finally, you spot it. It took ten minutes to find one shirt. And you realize as you look at it that it is stained. You can’t use it.
Don’t exaggerate. It’s Not that bad!
In organizations it is 1000 times worse. Most documents have multiple –even tens or hundreds of versions. How many different iterations are there of sales decks? What about the logo?
As I began thinking about this article, I decided I needed to learn if what I saw was a trend or an anomaly. I launched a survey that got 147 responses, 90% of which were people working in high-growth technology companies. The survey is still available if you’d like to participate.
When rating their satisfaction with their company’s knowledge management (KM), only 19% were very satisfied. The lion’s share – 63% –rated it either 2 or 3 (out of 5). Even allowing for our innate resistance to rating things at the very top or bottom of a scale, the most frequent rating was 2 (tolerable).
Typical comments:
- Information is all over the place.
- The most tricky thing; when colleagues share docs via a link. If you require this document after a while it is impossible to find.
- Usually, instead of hunting for something, I just create it. It’s faster.
Unlike most technical debt, knowledge disorder compounds at an exponential rate. Each new version of each document adds greater obfuscation to finding what you need.
The Cost
This is not a trivial problems. In 2003, a landmark study estimated this knowledge chaos could cost a 1000-person company $2.5 million annually from time squandered on fruitless searches. That’s before considering the immeasurable toll of duplicated efforts, missed opportunities, poor decisions, and lost innovation.
Today, with information even more disaggregated across countless SaaS platforms and cloud repositories, the problem has metastasized.
Wasting 15% or 20% of one’s time searching for information sounds awful. But it’s worse still when you consider all the downstream consequences.
- Using inferior substitutes for the thing they need or duplicating efforts.
- Solving problems that have already been solved.
- Sending inaccurate documents that confuse customers.
- Pricing things incorrectly.
- Failing to give a client the right answer.
- Making poor decisions based on the wrong document or failing to find any information.
Worse, doing all these unproductive activities displaces creating new knowledge or opportunities.
That 2003 study estimated that the lack of KM could cost a 1000-person company as much as $2.5m a year just in wasted time.
Imagine what the numbers might be today.
Now What?
There are some key reasons for the dearth of good KM hygiene.
1) No one owns it.
2) The default tools are terrible.
3) It doesn’t seem to be connected to the critical KPIs like ARR or Growth.
Fixing those issues is not hard.
KM must have specific standards and rules. Those include naming conventions and guidelines about where everything is saved.
Also, ongoing KPIs and employee assessments must include asking about information accessibility—and monitoring information hygiene.
Finally, one person must own the KM of the organization. They are the naming and storage police—as well as the arbiter of how to archive, retire or code current documents that everyone utilizes.
Call Claude! (Or ChatGPT, Or Bard, or Llama or Arctic…)
When I started thinking about this, ChatGPT had just launched. Now that generative AI is here, KM will be easier than it has ever been before.
With just a few guardrails, Generative AI models like ChatGPT can be trained on an organization’s entire corpus of data. Employees can then query it using natural language for specific files or insights, bypassing the scavenger hunt.
Access to Knowledge is NOT a “Nice to Have”
For AI assistance to be viable, leaders have to acknowledge and prioritize the issue. Then, they must centralize and structure their informational repositories – putting the KM groundwork in place.
This is one of those problems that my Founder and CEO clients have at the bottom of their lists.
“We’ll do it when we get the next round of funding…”.
But it is urgent.
The most valuable resource that any organization has is its peoples’ productive time. And it is being frittered away on file searches and Slack channel-surfing. That stands between your employees and their full productivity. Get this done!
PS: I ran across a few products offering generative AI KM systems—good starts. But it may not require a new platform. It may be a DIY project for your team.a DIY project for your team.