How do you get users to work with metadata? How do you get them familiarized with it so that they are comfortable using it and also see the value in it?
——————————
Lorena Pozuelo Martinez
European Investment Bank
——————————
This is both easy AND hard, in my experience ��.
Easy part is using an example that almost everyone is familiar with (I pick on Amazon and eBay) to demonstrate and explain. Have found VERY few people that can’t relate really quickly. Though my mother is amongst those that can’t….LOL!
Hard part is changing the behaviour. That goes back to Change Management 101. You can only change human behaviour with a) pain for not changing, b) reward for changing, or c) both.
Hi Lorena:
As Lorne mentioned, you have to demonstrate how it can help them. And keep bringing it up and demonstrating it at every opportunity. Unless your office is willing to lock down other options and provide thorough training in the new options and procedures you require them to use.
Or, if you don’t already have the tools, demonstrate to your decision makers how automation can solve the problem for your organization. Essentially, if you have the right software, automation deals with the metadata for them and they get to move on to more productive things.
Example of how automation can take over working with metadata: Stop using paper forms, or Word/Excel templates (most common examples) as starting points for new records. Use a form process that can send the documents into your ERMS repository once a submit button is hit, with all the key information already in metadata fields. This way, users don’t have to do any naming, metadata additions, or even filing of electronic documents; it’s all done by automation. In our case, our repository can show the completed entries in a folder with columns of key metadata that can be sorted to easily find what users are looking for, and those columns can be arranged to create a meaningful report and then exported to an Excel spreadsheet. The forms reporting tools can also create reports automatically and email them to us or save into the repository.
——————————
Connie Prendergast
Records Management Clerk
Flagstaff County
Sedgewick, AB CANADA
——————————
Great thread. I think this is still one of the biggest challenges we have. How to get users to do or make use of categorizations, topics, metadata etc. In the past I would have said it has to be a combination of:
-a good information architecture to start primarily developed by information architect or IIM staff
-way to show users how using metadata helps them get their jobs done
-automate as much as possible to ease the burden
-good user interface to make it super easy for user to select/apply metadata or choose repository
-training the users
To be honest, I haven’t found a way to make this happen without great care and feeding of the metadata and going behind users to ensure good categorization. I think for the future it may boil down to this:
-use AI and IIM staff expertise to create a good taxonomy/metadata
-take user out of the loop for applying metadata with use of AI
-periodic audits/checks by AI and IIM staff
——————————
Laura Downey, PhD
Chief, Applied Architecture
TSA
——————————
Most important in my opinion is:
-take user out of the loop for applying metadata with use of AI
Staff should create business value as much as possible, administrative Tasks should be reduced to the minimum.
As our ECM-Plattform (for business relevant content, final versions) is concerned, our staff is already quite relieved by automated metadating. Still a big concern are media, where you can store your working documents (not final versions, copies of records e.g. on Fileserver, Collaboration, OneNote) and metadata to LifeCycle is missing (no Partner ID, no Contract ID), as these working docs obviously contain PII. To solve this, we consider establishing Tools with cognitive search / semantic web technologies and AI/machine learning capabilities.
——————————
Thomas Pereira Antunes
Center of Competence Records Management
AXA Switzerland
——————————
Whenever I speak about metadata and the challenges with people entering the right metadata, I echo your comments Laura… but there’s one quick analogy I use to help people understand the difficulty of this problem.
Millions upon millions of dollars have been spent to increase hand washing compliance by doctors in acute care settings. (mostly hospitals). Still our compliance rates are in the 80% arena in the US and somewhere above 20% worldwide. No one questions the value of the activity and yet they still don’t do it. There are a lot of factors, but they don’t see hand washing as their job. they see saving lives as their job and hand washing may be appropriate but it’s never what they focus on.
Users are never going to focus on metadata entry. You can encourage it by making it easier, supporting it with automation, etc., however, to expect 100% compliance is fool hearty.
BTW – Handwashing shouldn’t be 100% anyway, there are crisis situations … metadata should be the same way — but hopefully without the same consequences.
——————————
Robert Bogue
President
Thor Projects LLC
——————————
This is a question that many in our profession struggling with. In a large development organization tagging or classifying content is mandatory. I have no idea on compliance but the case has been made that if you do some minimal tagging, the fruits of doing does pay of in the form of reusable and relevant precision of search results. The avenue of applying AI tools, highlighted, in this forum is for sure a path worthwhile to look at to minimize staff burden to classify content. In presentation in the organization I mention that most of you do tagging when posting pictures or stories in FB, IG, or Tweets so why don’t you do it with your professional documents? Why is it we have for financial transactions strict rules but not for documents? It takes a long journey to convey the message and for some co-workers the message falls flat. If you can make tagging extremely easy it may work but we are not there yet.
——————————
Daan Boom
Director
CCLFI
——————————
Daan, hear hear! I agree on all counts:
- Tagging starts to pay off even at relatively low levels of compliance.
- It’s hard even to monitor compliance
- Mere ‘compliance’ in no way guarantees that the tags applied are actually the best ones
- Ways to minimize staff burden become mandatory to consider.
Hopefully the organizations we find ourselves in are not so myopic as to deny the value of investment in semantic AI technology.
——————————
Eric Mullerbeck
Information Management Specialist
UNICEF
——————————
Hello @Lorena Pozuelo Martinez (Et al.),
Good discussion! I enjoyed reading your various inputs; there is a lot of experience bundled in this thread…
I cannot speak much to the AI component, though I hold out hope for that. We are primarily still dependent on the users. Though, up and coming iterations of SharePoint look promising… :0)
To your question:
Ideally set the foundation for metadata with a tool to look through your file shares and derive terms you will need. This should help to start with a good base for building out a taxonomy. I am thinking of a tool like #ActiveNav – https://activenavigation.com
Once docs are in the system then:
- Have leadership buy in to use the Metadata model vice the folder model
- Build/Maintain Metadata categories that are consistent throughout your system
- Work with groups to understand their particular content and requirements
- Use the technology to surface content in various ways based on customer driven need
- Take the time to sit down with individuals, even after training, to assist them in adopting to the technology and making it work for them.
The struggle is real…
Best To You All!
——————————
Ken Huie
Principal Applications Engineer
MANTECH
——————————
Ken,
I disagree with this part of your response: “Once docs are in the system then”. I would assert that those activities should all be done BEFORE content gets moved. Try to do it after content moves is an absolutely losing proposition. Like try to catch an endless piece of string….
If you are depending on individuals to add the metadata after the docs are brought into the repository, then yes, this could be a struggle. Our office does this, however, and we have achieved success by way of constant quality control of incoming documents checking to ensure the individuals responsible did what they were supposed to. One person per department (we are a small org) is trained, monitored, reviewed, given additional training/refreshing of instructions in meetings that include discussions on how we can do better, constantly monitoring. We have them so well trained right now, we have consistent, dependable metadata, and I only have to spot check occasionally now and monitor only new hires more closely.
That being said, we are beginning to take advantage of automation to add metadata, some AS it is coming into the repository, some still after it is already in. Workflows that run as specific times, picking up text (OCR) or adding retention controls based on folder metadata, or creating the paperless processes that automatically add the metadata required just prior to being sent to the repository.
We have not looked at any real AI purchase options, but… something to consider for the future.
——————————
Connie Prendergast
Records Management Clerk
Flagstaff County
Sedgewick, AB CANADA
——————————