BẢN CHÉP ÂM THANH 🔗
WEBVTT
00:00:00.479 --> 00:00:10.400
If you've ever opened a SharePoint library and found final version 7, really final, this episode is for you.
00:00:19.920 --> 00:00:29.359
Hey everyone, welcome back to the Simply SharePoint podcast, where we fix the mess, simplify the chaos, and make SharePoint actually work for real people.
00:00:30.239 --> 00:00:34.240
This week, on my blog, I shared a big project story.
00:00:34.479 --> 00:00:40.320
The policy and procedure hub that became the single source of truth across an entire organization.
00:00:40.719 --> 00:00:49.200
We use nothing fancy, no custom code, no third-party tools, just good, clean SharePoint fundamentals done well.
00:00:49.600 --> 00:00:52.000
But this week we're getting personal.
00:00:52.240 --> 00:01:01.759
Because while it's great to talk about organizational transformations, sometimes the real battle happens inside one very messy document library.
00:01:02.079 --> 00:01:10.560
The kind of library buried under files called final underscore version underscore seven underscore really underscore final.
00:01:11.200 --> 00:01:12.319
Yeah, that one.
00:01:12.560 --> 00:01:16.239
So today we're staging a SharePoint intervention.
00:01:25.439 --> 00:01:27.120
Every team has one.
00:01:27.280 --> 00:01:31.760
A document library that's basically the digital equivalent of a junk drawer.
00:01:32.000 --> 00:01:33.519
It starts innocently.
00:01:33.680 --> 00:01:40.560
Someone creates a folder called Dras, someone else adds archive, then someone decides to rename everything final.
00:01:41.359 --> 00:01:50.239
Before long, you've got nested folders, hundreds of duplicates, and a mysterious folder called old that no one's opened since 2018.
00:01:50.400 --> 00:01:51.599
And the worst part?
00:01:51.840 --> 00:01:54.959
People still depend on it to make business decisions.
00:01:55.280 --> 00:02:02.079
When I was brought in to fix this particular library, no one could tell me where the latest version of anything licked.
00:02:02.239 --> 00:02:07.840
The team had actually started emailing attachments again because search results were useless.
00:02:08.080 --> 00:02:09.759
Does this sound familiar?
00:02:10.000 --> 00:02:15.439
That's when I realized this wasn't just a cleanup job, this was an intervention.
00:02:22.719 --> 00:02:30.719
So once we'd taken a deep breath and agreed not to touch anything just yet, I started the real intervention process.
00:02:30.960 --> 00:02:35.759
The first thing I did was map the mess, and I mean literally.
00:02:35.919 --> 00:02:39.439
I exported an inventory of every file in that library.
00:02:39.599 --> 00:02:44.319
The names, owners, last modified dates, and sizes.
00:02:44.639 --> 00:02:51.520
This gave me visibility into what was active, what was abandoned, and who was responsible for what.
00:02:52.000 --> 00:02:55.759
Next, I asked one simple question to the team.
00:02:56.159 --> 00:02:59.439
What do you actually need this library to do for you?
00:03:00.000 --> 00:03:03.360
And that question completely changed the conversation.
00:03:03.840 --> 00:03:12.800
Because instead of we just need better search, what they actually wanted was confidence to know they were using the right document every time.
00:03:13.360 --> 00:03:20.639
That's when I explained my golden rule, which is you don't fix search, you fix the architecture.
00:03:20.879 --> 00:03:27.759
If you can't find what you need in a few clicks or by filtering a view, your library isn't structured properly.
00:03:27.919 --> 00:03:30.960
It's not a search problem, it's a structure problem.
00:03:31.199 --> 00:03:33.039
So that's where we started.
00:03:34.400 --> 00:03:38.400
I built the new structure around metadata, not folders.
00:03:38.639 --> 00:03:43.120
I know people love their folders, but in this case the folders were the problem.
00:03:43.280 --> 00:03:45.120
They were hiding context.
00:03:45.280 --> 00:03:49.759
So I flattened the entire structure and replaced it with clean metadata.
00:03:50.080 --> 00:03:52.479
We used four simple columns.
00:03:52.719 --> 00:03:55.840
First was department, so who owns it.
00:03:56.080 --> 00:03:57.919
Second was document type.
00:03:58.080 --> 00:04:01.680
Is it a policy, a form, a procedure, etc.?
00:04:02.240 --> 00:04:06.800
The status, is it working, approved, or archived?
00:04:07.039 --> 00:04:10.319
And owner, the person who is responsible.
00:04:10.560 --> 00:04:11.520
That's it.
00:04:11.840 --> 00:04:13.360
Four fields.
00:04:13.759 --> 00:04:20.240
Every single document had to have those four pieces of information before it could live in the library.
00:04:20.560 --> 00:04:29.199
Then I created a few smart views, the first being my department's documents, then approved documents only.
00:04:30.560 --> 00:04:36.079
Another one called Drafts in Progress and then recently updated.
00:04:36.879 --> 00:04:42.079
These views became the magic filters that replaced folders entirely.
00:04:42.959 --> 00:04:49.839
People could now click once or twice and instantly see what they needed, but no digging, no guessing.
00:04:50.319 --> 00:04:53.439
And the beauty of it was it looked simple.
00:04:54.079 --> 00:04:56.800
We didn't build some big new system.
00:04:57.040 --> 00:04:59.600
We just used SharePoint properly.
00:05:01.279 --> 00:05:05.839
Once we had the structure in place, we started cleaning up the actual content.
00:05:06.319 --> 00:05:08.720
Every final version went through a check.
00:05:08.879 --> 00:05:15.199
We reviewed it, tagged it with the correct metadata, and deleted or archived the duplicates.
00:05:15.439 --> 00:05:20.959
We agreed on a simple rule: one current version per document type.
00:05:21.279 --> 00:05:26.000
No duplicates, no half-edited drafts hiding in folders.
00:05:26.319 --> 00:05:31.439
I also created a working documents view where people could collaborate and draft.
00:05:31.680 --> 00:05:36.800
Once approved, the file was converted to PDF and moved
Load full transcript
into a published view.
00:05:37.040 --> 00:05:41.839
That one change alone stopped all the confusion between draft and final.
00:05:42.480 --> 00:05:52.079
We didn't need an expensive approval workflow or a power automate circus, just a consistent pattern that everyone followed.
00:05:54.000 --> 00:05:55.360
So what happened next?
00:05:55.600 --> 00:05:57.680
Well, the difference was day and night.
00:05:57.839 --> 00:06:06.639
Suddenly, search results made sense again, not because we fixed search, but because SharePoint finally had something clean and structured to search through.
00:06:06.800 --> 00:06:11.439
The team could filter by department or status and instantly find what they needed.
00:06:11.680 --> 00:06:14.319
No one had to guess which version was the latest.
00:06:14.800 --> 00:06:19.920
And for the first time in years, the library actually reflected how people worked.
00:06:20.399 --> 00:06:22.639
Within a week, the noise stopped.
00:06:22.879 --> 00:06:32.480
No more where's that final version emails, no more back and forth attachments in Teams chats, people started trusting the system again.
00:06:41.360 --> 00:06:42.639
Now here's the thing.
00:06:42.879 --> 00:06:50.160
We'd built a really clean structure, but you can't just hand people a new way of working and expect magic to happen.
00:06:50.399 --> 00:06:52.639
That's where user training comes in.
00:06:52.879 --> 00:07:03.920
We had to make sure the team understood the why as much as the how, because structure only works if people know how to use it, and more importantly, why it matters.
00:07:04.160 --> 00:07:07.199
So I ran short practical training sessions.
00:07:07.360 --> 00:07:10.319
I also recorded them and put them up on their intranet.
00:07:10.560 --> 00:07:22.720
They were nothing formal, just 30-minute drop-ins where I was able to show them some quick tips so they could start working in that library right away, but also properly.
00:07:22.959 --> 00:07:29.439
The first one was how to edit metadata quickly, right from the document library grid view.
00:07:30.000 --> 00:07:36.959
The second one was how to filter or group by department or status to find what they needed fast.
00:07:37.360 --> 00:07:43.920
And thirdly, how to spot gaps using a special view I'd built called missing metadata.
00:07:44.560 --> 00:07:46.639
Now, a quick note on that.
00:07:50.079 --> 00:07:57.439
When you do, people get blocked from saving, they get frustrated, and they start working around the system.
00:07:57.839 --> 00:08:05.519
Usually by saving locally or renaming files temp.doc, which is what we were trying to get away from in the first place.
00:08:05.920 --> 00:08:10.160
Instead, I make it visible and not painful.
00:08:10.800 --> 00:08:17.920
That missing metadata view became the friendly nudge that said, hey, you've got a few things to tidy up.
00:08:18.480 --> 00:08:22.000
It empowered users instead of punishing them.
00:08:22.800 --> 00:08:29.040
We also covered version history because most people have never really explored what that does.
00:08:29.279 --> 00:08:39.679
I showed them how they could roll back a document, see who made changes, and use that instead of final underscore version underscore h dot doc.
00:08:40.480 --> 00:08:44.320
Then we talked about the draft to final process.
00:08:44.720 --> 00:08:51.840
Once a file was ready, it was approved, converted to PDF, and surfaced in the published view.
00:08:52.320 --> 00:09:00.159
That clarity helped everyone understand what final actually meant, not just I stopped editing this.
00:09:00.480 --> 00:09:03.360
And this wasn't a one-and-done thing.
00:09:03.679 --> 00:09:06.799
We built ongoing learning into the process.
00:09:07.279 --> 00:09:13.360
Anytime a new team member joined or the library evolved, we did a refresher.
00:09:13.679 --> 00:09:18.639
Because good information management isn't a project, it's a habit.
00:09:33.519 --> 00:09:41.440
You just need to clean them up, give them structure, and make sure the people using them actually understand how it all fits together.
00:09:41.679 --> 00:09:44.559
And yet, lately I've noticed something interesting.
00:09:44.799 --> 00:09:52.559
There's this growing belief that you can just build an AI agent or switch on Copilot and it will somehow do the heavy lifting for you.
00:09:52.799 --> 00:09:57.759
Like AI is just going to swoop in and magically organize all those years of chaos.
00:09:57.919 --> 00:10:00.720
I wish it worked that way, but it doesn't.
00:10:01.039 --> 00:10:05.360
Copilot or any agent you build is only as good as the information you feed it.
00:10:05.600 --> 00:10:09.440
If your SharePoint is a mess, your AI will be confused too.
00:10:09.600 --> 00:10:17.519
It'll pull outdated documents, show draft versions, and surface content that should have been archived years ago.
00:10:19.039 --> 00:10:24.720
So even though we're living in this shiny new AI era, the same fundamentals still matter.
00:10:24.960 --> 00:10:34.480
Your information architecture, the way you name files, apply metadata, design content types, and structure libraries, that's what makes AI smart.
00:10:34.639 --> 00:10:38.480
That's what helps Copilot give meaningful, accurate answers.
00:10:38.720 --> 00:10:43.840
When I talk about AI readiness, I'm really talking about data discipline.
00:10:44.080 --> 00:10:49.200
Every column, every term, every title contributes to that readiness.
00:10:49.440 --> 00:10:54.240
If your metadata is missing or inconsistent, you can have the best agent in the world.
00:10:54.399 --> 00:10:56.639
It will still give bad results.
00:10:56.879 --> 00:11:06.480
So before you start building that next co-pilot extension or fancy agent, ask yourself: can a human find what they need in three clicks?
00:11:06.720 --> 00:11:11.360
If the answer is no, how is Co-Copilot going to find it?
00:11:11.519 --> 00:11:16.720
So that's why I always say, and yes, this has become a bit of a signature line.
00:11:16.960 --> 00:11:21.759
The fix isn't more AI, it's better organization.
00:11:22.480 --> 00:11:25.440
Because here's what happens when you get the foundations right.
00:11:25.759 --> 00:11:29.360
You don't waste time redesigning your entire SharePoint site.
00:11:29.519 --> 00:11:36.080
You refine it, you reuse what already works: the libraries, the lists, the metadata columns.
00:11:36.240 --> 00:11:44.159
You make it scalable and sustainable so every future project builds on the same logic instead of reinventing it.
00:11:44.480 --> 00:11:47.919
You don't need a big rebuild or a new app.
00:11:48.159 --> 00:11:49.360
You need clarity.
00:11:49.679 --> 00:11:56.000
You need good information architecture and a team that understands how to keep it clean.
00:11:56.639 --> 00:12:07.440
When those pieces come together, SharePoint stops being that annoying system people avoid and becomes the foundation that makes every AI tool around it work better.
00:12:07.679 --> 00:12:20.480
So remember, before you automate, before you integrate, before you add co-pilot, get your content right first, because the smartest AI in the world can't fix a messy library.
00:12:31.600 --> 00:12:36.799
Once the structure was in place and the training kicked in, the change was obvious.
00:12:37.039 --> 00:12:41.919
That same team who used to roll their eyes at SharePoints suddenly started using it every day.
00:12:42.000 --> 00:12:43.759
And not because they were forced to.
00:12:44.000 --> 00:12:45.759
They used it because it worked.
00:12:46.000 --> 00:12:50.240
There was no more where's the final version messages in teams.
00:12:50.399 --> 00:12:54.879
No more desperate searches through folders called old or backups.
00:12:55.120 --> 00:13:01.279
Instead, people could filter a view, click twice, and land exactly where they needed to be.
00:13:01.919 --> 00:13:04.879
The results went beyond tidy folders.
00:13:05.120 --> 00:13:08.159
Meetings stopped wasting 15 minutes trying to find the right document.
00:13:08.320 --> 00:13:15.120
New starters could get up to speed within hours instead of weeks because everything they needed was already surfaced with their department.
00:13:15.360 --> 00:13:19.679
And the best part, your team actually started replicating the model.
00:13:20.159 --> 00:13:24.399
Other departments saw how well it worked and asked, can we have that too?
00:13:24.639 --> 00:13:29.200
That's when I knew it had stuck, when the design became a pattern.
00:13:29.679 --> 00:13:31.840
But here's where it got even more interesting.
00:13:32.080 --> 00:13:36.320
Once everything was clean and consistent, we introduced Copilot.
00:13:36.399 --> 00:13:39.120
And for the first time, it actually made sense.
00:13:39.360 --> 00:13:47.759
When people asked Copilot to summarize all approved safety procedures updated this year, it could do it perfectly.
00:13:48.080 --> 00:13:48.720
Why?
00:13:48.960 --> 00:13:52.960
Because the metadata, content types, and the file names were right.
00:13:53.200 --> 00:13:56.320
That's the hidden benefit of doing this groundwork.
00:13:56.559 --> 00:14:03.120
You're not just building a better library, you're building a foundation that every AI tool depends on.
00:14:03.679 --> 00:14:08.720
Because here's the reality: AI isn't replacing your information architecture.
00:14:08.960 --> 00:14:10.639
It's powered by it.
00:14:10.879 --> 00:14:13.759
If your structure is weak, Copilot guesses.
00:14:13.919 --> 00:14:17.440
If your metadata is clean, copilot shines.
00:14:17.600 --> 00:14:20.799
And that's why I love seeing these interventions in action.
00:14:20.960 --> 00:14:25.360
They prove that SharePoint doesn't need to be rebuilt or replaced.
00:14:25.679 --> 00:14:29.759
It just needs to be understood and maintained.
00:14:30.480 --> 00:14:34.080
So the outcome of this project wasn't just a cleaner library.
00:14:34.240 --> 00:14:39.759
It was a smarter, calmer, more confident team, a team that trusted their content again.
00:14:39.919 --> 00:14:42.879
And as a result, trusted the technology.
00:14:43.120 --> 00:14:54.240
When you get that right, SharePoint stops being the system that IT made us use and becomes the tool that quietly powers everything else, including your AI.
00:15:11.440 --> 00:15:18.159
What started as a cleanup project ended up changing the way the team worked and even how they saw SharePoint.
00:15:18.399 --> 00:15:26.559
Because when your content is organized, everything else, from collaboration to co-pilot, finally clicks into place.
00:15:27.440 --> 00:15:29.360
And that's the real lesson here.
00:15:29.679 --> 00:15:34.080
AI doesn't replace good information management, it rewards it.
00:15:34.320 --> 00:15:43.919
When your content is well structured, your AI becomes smarter, your users become more confident, and your organization finally starts working the way it was meant to.
00:15:44.159 --> 00:15:48.240
So if you take one thing away from today's episode, let it be this.
00:15:48.639 --> 00:15:52.879
The fix isn't more AI, it's better organization.
00:15:53.039 --> 00:15:55.600
Next week, we're going a level deeper.
00:15:55.759 --> 00:16:05.360
I'll be sharing exactly how I design metadata and information architecture from scratch when I'm building new SharePoint solutions.
00:16:05.600 --> 00:16:15.279
I'll walk you through my go-to columns, my naming conventions, and the patterns that make every project scalable and co pilot ready from day one.
00:16:15.679 --> 00:16:18.720
Until then, I'm Liza from Simply SharePoint.
00:16:19.039 --> 00:16:26.399
Thanks for listening, and remember if it takes more than a few clicks to find what you need, you don't need better search.
00:16:26.559 --> 00:16:28.399
You need better structure.
00:16:28.639 --> 00:16:30.240
See you next time.