What are the best practices for designing and conducting knowledge management surveys?

Conducting a survey for your knowledge management help to know the current phase you are in knowledge sharing and collaboration in an organization. The assessing part of the MI helps to know the specific knowledge management part you want to evaluate, the best tool that may work best for it, the knowledge gap and opportunities ahead for the organization.

The thing with surveys is that if it is seen as just another nuisance task, you'll get either pro-forma answers or no replies. (The irony is that making it mandatory is probably the best way to get horrible feedback!) When we did surveys, the successes we had were with short (5-7 questions) and focused surveys (e.g. KM Representatives). Im personally on the fence about the real value of surveys unless they are tied to a specific action (e.g. survey as part of a post-course or post conference After Action Review.).

This sounds like you want to have a database and run analytics on it. Nothing wrong with that in my opinion. Having said that, we are talking about knowledge management not database analytics. In my not so very humble opinion, knowledge management implies data in a context - or teaching the computer to work in concepts as opposed to tuples. Furthermore in my opinion the way the knowledge is related needs to be vastly different. In the 2 programs, the one we call 'Knowledge Master' (KM) contemplates a very large interlinked mesh of concepts. The concepts KM 'knows' about are changeable on the fly.

Surveys can certainly help to measure performance of your knowledge management practices. In addition to defining objectives and scope, there is another important element to consider: repeatability! A one-off survey may serve as a baseline and provide some insights. But the true benefit comes from regular repetition. To identify trends, reinforce what works, and address problem areas. 3 practical tips: Start small: Only take the questions that are truly relevant and repeatable. You can add more questions any time you run it. Keep the effort to answer low: If I had to answer this question 10 times, would it annoy me? Avoid neutral answers: A "not applicable" is fine, but on a scale you should not be able to choose the safe center.

There should be pre steps to that. Understanding the current situation, needs and business goals. We can collect those in various methods, such as shadowing sessions with reps, interviews with team leaders and a general survey sent randomly. After 1st analysis, we can set proper goals and conduct a deeper survey

I've designed knowledge management assessments. Simple Microsoft Office tools like Excel or SharePoint forms can be used, if you are conducting one-off surveys. If you are surveying a large group, a web-based platform like SurveyMonkey can help.

Hier k?nnen Tools wie die "Wissensbilanz" oder der "Knowledge Quick Scan" helfen. Für ersteres gibt es Internet-Tools für zweiteres nutze ich eine Excel-Vorlage. Ich denke, dass es wichtig ist, die Umfrage für die Teilnehmenden schmackhaft zu machen. Klar zu machen, dass es super wichtig ist und ihre Meinung z?hlt. Nur so kann das Unternehmen besser werden und sich weiterentwickeln. Damit habe ich gute Erfahrung gemacht und auch eine 100%-ige Rücklaufquote erzielt. Klar, je kleiner der angesprochene Kreis, desto einfacher vielleicht. Eine Idee für gro?e Unternehmen/Konzerne k?nnte sein, abteilungsweise/teamweise die Umfragen mit einer kurzen direkten Ansprache zu verteilen. Das kostet jedoch mehr Ressource...

It should be a combination of tools, to gather the "big picture" - combination of qualitative and quantitative tools. Face-to-face interviews for insights and deeper questions, and using a survey form for answers from a larger scale of numbers. If the survey tool is part of the platform itself - can be easier with gathering and monitoring the data, while testing the usage of the platform

Choose your survey method and tool: Select a survey method that aligns with your objectives (e.g., online, face-to-face) and a user-friendly tool to ensure ease of participation and accurate data collection.

I like using Microsoft Forms as it allows for cascading questions to keep questionnaire on topic based on answers of your target population. Combination with SharePoint, as other author here proposes, is also a good idea as it allows for easy reuse in documents, presentations and (PowerBI) reports. What I do miss in the Microsoft suite, and maybe SurveyMonkey solved this, is recoding answers or have random placing of dimension extremes in Likert-scale questions. This, in order to lower risk of surveys automatically clicked based on assumption that everything on the outer right extreme is a positive answer or rating. Go and experiment!

In my experience, asking the right questions to get the feedback you are looking for about your KM program is critical. It takes time to curate the right questions and you must have a clear picture of what you are trying to seek out from the survey answers. We do different types of surveys throughout the year to gather feedback to improve and adjust the work we do in our KM program. We do a yearly survey for a broad audience to collect data about the KM program, tools, and activities. We also do small focus group sessions that are on one specific topic to dive deeper. KM survey questions we create tell us what is working well, what needs improved, what our people need, and what direction we need to go in the future.

When preparing the questions, it is important to look at them from different perspectives to avoid misunderstandings. Before the survey is sent out, at least one test run should be done. Let the test-group pay particular attention to the wording and check if they interpret them as desired. Not only the questions are important, but also the possible answers. I prefer answers on a Likert scale for things like acceptance or satisfaction. Example: fully agree (1), agree (2), slightly agree (3), slightly disagree (4), disagree (5), fully disagree (6) Ideally without a neutral value. In the evaluation, the mean value can simply be calculated from the Likert scale data in order to decide whether measures should be derived or not.

Design your survey questions: Craft clear and concise questions that align with the survey's objectives, ensuring they are specific, measurable, and relevant to gather meaningful insights from participants.

I recently used ChatGPT to generate questions. First familiarize the LLM with KM research around specific models. Then prompt it towards survey questions and iterate a few times on the results by guiding it to be more precise around constructs of your research. After few iterations I was amazed by the question set I got regarding Social Learning Cycles. 3 questions (1 per dimension), 6 phases. DM me if you want to know also about what happened when I prompted to familiarize with KYC bank activities and roles, and then asked to provide answers as if the respondent was a senior KYC analist. Wow!

Having burnt my fingers quite early in my career with surveys, I would highly recommend that they are kept short, easy to read and comprehend, and aligned with the chosen objectives. If you are in need of comprehensive information that a short survey will not suffice, consider a different method like an interview, or revisit the objectives of the exercise. Asking respondents to go through 5 pages of questions on their own is setting the study up for failure.

One thing that I have done in some projects is suggest meeting with managers. After defining KM, publishe questionnaire and describe each aspect. After that, set a deadline for each manager to publish this questionnaire in his/her department. In the next meeting, the departments performance was checked, and this road map made effective cooperation.

Participation is key to ensure the current knowledge management activities are implemented. This includes roundtable discussions with various roles and types of users from content creators to consumers. During these discussions you have the opportunity to learn about the importance of KM and the different needs from audiences. The users are then more likely to engage with the solution in addition to having a deployment more suited to their business needs and requirements.

Identify a diverse group of participants representing various roles and expertise levels within the organization. Ensure clear communication about the survey's purpose, emphasizing the importance of their insights in shaping knowledge management strategies.

After conducting knowledge management surveys, it’s crucial to analyze data effectively and present actionable insights. Utilize tools like charts, graphs, and summaries to highlight key findings. For instance, survey results should be analyzed using both quantitative metrics (like response rates) and qualitative feedback (themes from open-ended questions), enabling the team to identify training gaps and improve knowledge sharing by a significant rate within teams.

What I find useful is visualizing Forms output (your questionnairs results) in PowerBI. By this I can design a research project and have actuals included automatically in a report. The report might as well be a set of webpages, making interaction on findings possible and directly linked. Though the visualization library is well packed, I am still looking for visualizing the Social Learning Cycle (Boisot) based on values alongside 3 dimension (codification, abstraction, diffusion). Think of a swirled carpet in a cube. I now connect the dots from an Excel visualization. Maybe someone has an alternative suggestion?

I have to question the fundamental premise of this whole article. Do you call making a survey and running analytics on it 'Knowledge Management"? Thats a philosophical question I would think. I dont consider that knowledge management. Thats my opinion - worth just about what it cost you.

Make the survey short. Even well-meaning people will skip a long survey. That will be the last thing on anyone's priority list.

Some quick tips from my side: Keep it concise and engaging! Use clear questions, mix types, keep it anonymous, and offer incentives. Keep your tone friendly and professional. Test before launch and analyse results thoroughly.

The real "KBA" database that I worked in had to be 508 compliant, and had to be vetted on a case by case basis by subject matter experts. Also, contains 2200+ articles that were audited on a cycle (for my org, it wad 2 years) this required constant feedback loops and a process more akin to "change management" and CAB needed to be in place to make sure the most frequently used KBs where current and English readable, not IT readable. In short, if you are planning on doing a survey, you're flat out doing it wrong. Now if your Knowledge Management is at ground zero than this may help identify "heat areas" where documentation needs to start from, but if you think you need to plan a survey, just start a separate KM team.

I prefer focus groups to surveys and use them to identify community pain points particularly early on in an organization’s KM program maturity.