Skip to main content

What is knowledge management?

Knowledge management simply means decision-making based on accurate information, obtained by analyzing data. Decision-making can be strategic, tactical, or operational, and often operational decision-making is highly automated.

How does knowledge management help create better business, what does it require, and how can I get started? After reading this page, you'll know – so take about 15 minutes, find a comfortable position, and start browsing!

What knowledge management doesn't mean (although it is often misunderstood)?

Knowledge management is still defined somewhat lightly as decision-making supported by decision support systems—reports and dashboards—producing information for decision-making. According to this definition, advanced analytics would be something that comes in addition to knowledge management.

When knowledge management is defined as reports and dashboards, the challenge lies primarily in the perspective: the focus is in the rearview mirror, and diagnostics provide reasons only for why something already happened. In general, it is useless to do flashy tricks when the horse has already left the barn.

What if the perspective could be turned towards the future: what does analytics tell us from which we could learn and predict the effects of actions and events?

For us, knowledge management encompasses all analytics and artificial intelligence.

Analytics refers to the analysis of data to solve a question. Artificial intelligence, on the other hand, can be defined as analytics based on machine learning combined with automation.

For us, there is no such thing as knowledge management that requires additional analytics or artificial intelligence.

Why does knowledge management result in better business?

The business benefits in knowledge management are found in the term itself: knowledge management shifts business from intuition-based management to knowledge and data based. Data-driven business and its development are not based on guesswork, luck, or opinions, but always rely on data and the information derived from it, processed into understanding on which decision-making is based.

All business is ultimately human, and because of that, the temptation to make decisions based on emotions is significant. When there is accurate historical data and an understanding of what has worked and what hasn't behind business decisions, decision-making shifts from emotion-based to fact-based.

As a result, the accuracy of decisions improves, making the business more effective.

knowledge management_horizon

Eyes on the Horizon

When decision-making is real-time and forward-looking, instead of mere ad hoc reactions, the perspective of business shifts from the rearview mirror to the crystal ball.

For example, rather than looking at a map of where maintenance workers are currently, advanced location analytics could be used to optimize maintenance targets automatically.

Accurate Information, Smarter Decisions

So, knowledge management delivers relevant information to the right processes at the right time, resulting in better decisions for business development.

The decision-maker can be a human or a machine. Data-driven management can provide business with, for example, additional sales opportunities, streamline processes, create entirely new business possibilities, and generate significant cost savings.

knowlegde management_accurate information
knowledge management_customer experience

Example: Better Customer Experience

If a customer service representative isn't even aware that a customer is dissatisfied, how can customer retention be prevented?

When decision-making in practical work is enhanced with advanced analytics, it becomes easy to choose the right actions and words, leading to increased customer satisfaction—and sales as a result.

The issues of traditional knowledge management

Knowledge management enables various revolutions in business, but when used incorrectly (or more precisely, misunderstood), the benefits get lost.

Here are a few examples of pitfalls in knowledge management:

knowledge management_outdated data

Data is outdated, and it is examined too infrequently.

For instance, if the only analytics on sales situations come from monthly performance reports, quick responses to changes become impossible, and predicting the future relies largely on guesswork.

The collected data is not utilized for management but is only used to create dashboards.

If dashboards are created solely for the purpose of creating a visually appealing representation of data, the utilization of data in data-driven management remains rather thin.

knowledge management_unutilized data
knowledge management_humans

Decisions are made by humans even when it wouldn't be genuinely rational.

In surprisingly many situations, a machine's ability to make decisions surpasses human skills—and yet in many situations, a human is still the one pressing the OK button.

For some reason, mistakes made by humans, which happen constantly, are accepted, but a machine is not allowed to make a mistake ever.

Only internal data is actively used, and external and open data remain untapped.

Utilizing open data effectively enriches the information obtained from a company's own systems.

For example, if traffic congestion is predicted solely based on the traffic volume data collected by the company, neglecting the use of open weather data, the estimate may not be very accurate.

knowledge management_external data
knowledge management_crystal ball

The focus is on the rearview mirror instead of the crystal ball.

A map view of the destruction caused by a forest fire only tells what has already been destroyed.

It is more useful to generate a predictive model of the fire's spread and provide pre-digested information to firefighting units.

Cornerstones of modern knowledge management

What does modern knowledge management mean – what new tools have been introduced into the toolkit to avoid old pitfalls? In our view, modern knowledge management encompasses all analytics – up to prescriptive analytics predicting optimal practices.

Looking to the future with predictive analytics
Predictive analytics involves forecast models based on historical data. Predictive analytics tells what is going to happen.
As an example, in HR analytics, predictive analytics can assess—depending on the quality of the data, but often quite accurately—how sick leave will develop this or next year. Traditional analytics only provides, at most, the number of absences and the reasons for absences up to the present moment.
Another example of predictive analytics from the transportation sector: a consumer is unlikely to be interested in an overall view of train traffic when waiting on a platform in winter, with cold feet and thoughts of how quickly they can run from Pasila station to the customer's office. They don't want to see trains on a map but rather get precise and reliable information about the arrival time of their own train and when the train will be in Pasila.
Predictive analytics helps: an algorithm can take into account the current situation of train traffic, weather conditions, and other factors, providing an accurate estimate of the arrival time.
New machine learning methods supporting decision-making
Machine learning is a method, or more accurately, a set of methods, where analytics is implemented by providing the machine with all the necessary data and instructing it to come up with the best way to answer a given question. In this case, the machine learns to recognize significant patterns for the question without being programmed separately when the problem is too complex or laborious for human definition.
AI based on machine learning enables more accurate and highly automated decision-making. For example, image recognition, voice recognition, and text analytics can help streamline and automate decision-making.
One of the most interesting new technology trends is Edge AI, which refers to the combination of edge computing and AI: AI processes data as close to the data source as possible, either directly on the device or in a server located near the device. The device can perform analysis and make independent decisions in milliseconds without needing to be connected to the internet or the cloud. The possibilities for utilizing Edge AI are nearly limitless.
A machine learning-based model can relatively easily expedite certain insurance decisions. The machine can be taught to identify, from an image, whether the car windshield is broken, read the car's license plate, and based on this information, prepare the paperwork for an insurance decision.
Utilizing location data
Location itself is just three attributes among others, but when used correctly, location data can enrich various types of data and bring new dimensions to analytics. In fact, nearly all data elements are associated with location information, which holds tremendous potential for business development.
When location data is incorporated into the entirety of data-driven management, there is also a significantly better understanding of what and where things have actually happened and how things could be anticipated in the future.
For example, predicting maintenance needs and optimizing maintenance visits by analyzing location-enriched maintenance target data saves fuel, time, and the nerves of staff. How are maintenance visits optimized? Could multiple nearby targets be addressed proactively during the same visit?
Efficiency in processes with advanced analytics and automated decision-making
Advanced analytics refers to all analytics that is more advanced than traditional "looking in the rearview mirror."
Advanced analytics is not just an additional component of data-driven management; it is actually the core around which the entirety of data-driven management is formed.
For advanced analytics to truly benefit business, it should not be hoarded in dashboards but boldly integrated into business processes.
For example, in decision-making for banks, advanced analytics provides useful tools: a machine utilizing forecast models can instantly calculate the probability that a person applying for a bank loan will repay the debt on time.
The same decision-making task takes much longer for a human, and the human perspective usually does not bring much added value to the final decision. Also, from the perspective of customer satisfaction, a decision made by a machine is usually better because the customer receives a ready decision quickly.

What does modern knowledge management require?

Modern knowledge management requires the organization to have the ability to evolve and adapt comprehensively. Especially in the following areas, all pieces of the puzzle must be in place for modern knowledge management to be possible in the organization:

knowledge management_culture


The adaptability of culture is the essence of data-driven management.

If business decision-making is based on gut feelings, sophisticated guesses, and a childlike belief that we truly know our customers, a complete shift in mindset is required.

To ensure that data-driven management doesn't appear as an oversized monster to the staff, its benefits should be highlighted through examples.

A key factor in creating a data-driven management culture is for the staff to integrate analytics into their own work and genuinely see the benefits from it.


Data-driven management, of course, requires a lot of technical expertise for building integrations, transforming data from one platform to another, cleaning up the right information, implementing analytical models, and so on.

However, often the most challenging aspect to manage is the ability to understand the business: why data is collected and what is the goal of analytics?

knowledge management_expertise
knowledge management_org structure

Organizational Structure

Especially in many large organizations, data tends to rest in silos.

Each business unit has its own budget, and when it comes time to build a common data platform, the critical question becomes whose budget the funds will come from.

For this reason, it is important for the organization to have a Chief Data Officer (CDO) or a person responsible for analytics who can work across organizational silos.


Technology is an important component but not an end in itself. The technologies used should be chosen based on the situation and business needs, taking into account existing solutions.

Sometimes this requires investing in a completely new solution and abandoning the old, while other times existing technology should simply be utilized more efficiently.

knowledge management_tech

Measuring the maturity of knowledge management

The difference between pioneers and laggards in data-driven management is staggering. Some organizations have embedded data-driven practices since the beginning, while others rely on basic reporting. In many companies, artificial intelligence has been set as a goal by leadership, but the benefits of AI pilots to the business tend to remain thin for several reasons.

Many organizations are not even aware of their level of data-driven management and how they compare to competitors. This can pose significant challenges in measuring the results of data-driven management.

The maturity model for data-driven management helps identify risks and leverage strengths

It is crucial for an organization to understand its strengths and weaknesses in various aspects of data management. This understanding helps identify and respond to risks while leveraging strengths more effectively.

Our built data-driven management maturity model measures the maturity of data-driven management using fifteen indicators in total. The indicators are divided into three categories, which partially overlap.

Here is our model in all its glory (click the image to open it in full screen):

Get started with knowledge management

Success in modern knowledge management is primarily dependent on courage and an open-minded attitude: experiment agilely, dare to fail, learn from your mistakes, and scale your successes into new business processes. Analytics can be utilized with the data that is available, such as data from your own databases combined with information from open data sources.

Usually, a surprising amount of information can be obtained even if the data is siloed—both internal and external data repositories often hold a wealth of untapped potential.

Would you like to discuss more about artificial intelligence and data-driven management? Reserve a free time slot below: