NewB here...Wanting to understand the steps to Implement M365 Copilot

Copper Contributor


I work for a healthcare company, and we are looking to acquire the 300 licenses for Copilot. Before we get too far, I would like to hear some feedback on some best practices on Data tagging, classification, and overall management within the tenant. We are hearing that Data Management is the first deliverable to a successful and supportable Copilot implementation.

5 Replies



So first off right of the bad you need atleast 500+ E3 or E5 Licenses which I'm guessing you have!

As for best practices, I would recommend that you check your data quality! So what I mean with data quality is:

  • That you have files that are relevant! Clean up as much old junk as possible. It is truly bad data in -> bad data out if you don't clean up
  • Look over the whole architecture of your sites and data. Look into Hubsites and Sites in Sharepoint
  • Tag files and documents who as much relevant information as possible in document libraries and lists! Also make use of the term-store
  • You want to implement sensetivity labels and retention policies! This also means that you need to tag the information with the correct tags!
  • You also need to check permissions and that users have access to what they should have access to and not. Implement some form of access review
  • You might want to exclude some data considering you're in healthcare and patient data can be sensitive, look into what laws and stipulations exist.



A lot of these are organisational specific and also industry specific, for example sensetivity labels and retention policies for a health care organisation will differ a lot from manufacturing industry!


I would recommend you to talk to you Microsoft CSAM/CSM and see what support you can get in form of pilot and workshops from partners   

@NicolasKheirallah Appreciate the feedback. That is where we are headed...TY

Hi @CLitz,

As you will see, I agree with NicolasKheirallah,


Implementing Copilot is a substantial project; it's far more complex than simply flipping a switch. A crucial aspect of its deployment involves data and content tagging. To address your query, if the data or content is meaningful for the entire organization, it will similarly benefit Copilot. Like for any of us, Copilot performs optimally when the context is unambiguous. This involves several key steps:

  • Tagging content with relevant keywords.
  • Organizing content into logical folders and sites.
  • Retaining only the most recent versions of documents to avoid duplication.
  • Removing redundant, outdated, or trivial (ROT) content.
  • Adhering to a standardized naming convention for files.

This content could include various formats like files, emails, chats, wikis, pages, databases, etc. Appointing a data steward is a strategic investment for ensuring quality and success.

It appears you are in the process of evaluating the acquisition of licenses. May I suggest to initiate a high-level discussion within your leadership for commitment and establishing a Center of Excellence as foundational steps if not already done ?

The implementation of Copilot for Microsoft 365 involves multiple facets:

  • Establishing a Center of Excellence (CoE).
  • Managing change, as this technology can significantly alter workflows for employees and businesses.
  • Coordinating project management.
  • Conducting thorough analysis.
  • Engaging with partners/providers, who may already be developing solutions.
  • Handling security, access control, and compliance.
  • Providing training and knowledge transfer.
  • Involving programming expertise.
  • Raising awareness about the new system.
  • Popularizing the tool within the organization for maximum adoption and impact.

Source: Prepare your data for searches in Copilot for Microsoft 365 - Training | Microsoft Learn

Feel free for asking if you need further information,

@Andre_Caron ... Appreciate the feedback and information you outlined. I am the Program Manager for our Micrsoft adoption and implementation, so we have project management wrapped around this. Also, we have established an AI Task Force, that includes Senior Leadership. The MS Copilot Implementation is (1) initiative under the AI Task Force Program.

Quick question (maybe), while I would expect that the data scope would vary between companies, in your estimate, what does the LOE look like to get the data in a place where you could move forward with Copilot?

Again, appreciate your feedback and direction here.

Hi @CLitz,

I'm pleased to learn that you have already a team in place.


Regarding Copilot for Microsoft 365, which we will refer to as simply 'Copilot' from now on, its effectiveness hinges on the quality of the data it can access. Microsoft emphasizes several key aspects concerning this data: its security, accessibility, and readiness (get data in place) before progressing with Copilot.


Undoubtedly, the Level of Effort (LOE) required is substantial and will be significantly influenced by the condition of your data, as outlined by @NicolasKheirallah and myself in previous posts. User habits and the specific data available to Copilot also play critical roles. Consider, for instance, if access is confined to SharePoint, OneDrive, and files on each of the 300 employees' computers, in addition to emails and chats. This scope alone is extensive. Should the scope also encompass data from external systems like ERPs and CRMs, I recommend contemplating an additional phase for the project.


A crucial point to note is that if the data quality is subpar, Copilot's output will be substandard, potentially leading to a lack of adoption by your colleagues.


I hope this helps, Kind regards,

Get started with Microsoft 365 Copilot

MS-012 Prepare your organization for Microsoft 365 Copilot