Microsoft Copilot expands availability! Here’s 6 tips to get started
No more minimums, more plans than ever – it’s a pivotal moment for generative AI.
Today, Microsoft is taking the speed brakes off Copilot for Microsoft 365. With a deluge of articles and announcements flooding your feed about this, why read this one?
I’ve been a Microsoft partner since 1996, so I’ve been at this for a while. A significant part of my work involves advising clients on their Microsoft agreements, often drafting future-state roadmaps for their Microsoft stack utilization. Since its announcement, Copilot has been a major topic in these consultations. I’ve had in-depth discussions about Copilot with some of the world’s leading organizations, encompassing a broad range of industries in the public and private sectors.
Am I an expert? Maybe – but the clients I’ve interacted with certainly are, and I aim to channel their voices in this article while adding a dash of my own humble perspectives. I hope to give you three things:
- To outline the changes – the factual landscape.
- To share the feedback I’ve collected from numerous profound discussions with early adopters, who are incredibly insightful and forward-thinking.
- To explain why, in my view, we’ll likely reflect on this announcement in 18 months as a pivotal moment in the evolution of generative AI.
Part I: What’s changed?
Since the Copilot early adopter days, Microsoft placed constraints to limit customer exposure while they bolstered the necessary infrastructure to facilitate mass adoption. Today, all these limitations are being lifted. This means:
No more minimum quantities
Previously: Microsoft required a minimum purchase of 300 Copilot licenses.
Now: No minimum. Go solo and buy one, or go nuts and buy 100,000. Your choice.
Expanded availability across licensing programs
Previously: Copilot was only available by request as a special EA SKU, and Microsoft had to approve it every time.
Now: Copilot is available as a normal EA SKU, meaning you can add it to your EA and reserve it as needed. Even better, it’s now available on CSP as an annual-commit SKU.
Widespread availability as an add-on for more plans than ever
Previously: You could only add Copilot to M365 E3 or M365 E5.
Now: You can add Copilot to O365 E3, O365 E5, M365 Business Standard, M365 Business Premium, M365 E3, and M365 E5.
In short, Copilot has become as accessible as OneDrive, SharePoint, Exchange, or any other available SKU. For more detailed information, or if you have questions about your specific scenario, feel free to contact us.
Part II: What’s the word on Microsoft Copilot among your peers?
Since the initial announcement of Copilot, I’ve had intriguing discussions and even held informal roundtable dialogues with numerous customers (Remember the first rule of Copilot club: No one talks about Copilot club!). Out of these, six common themes have emerged that are worth noting:
1. Almost everyone is planning a partial rollout due to the substantial investment required for an enterprise-wide implementation.
The big question is: Who should be the target of the initial rollout? IT, executives, power users? Opinions vary, but common suggestions include targeting a few entire groups to observe the effects when a full team has access. Moreover, it’s important to spread it across the organization, not just to executives. There is an element of serendipity to Copilot – it can surprise you, and the broader the user base, the more likely you are to stumble upon unexpected findings.
Another critical question is whether it’s feasible to limit this technology to a select few. In some ways, it can give your users information worker superpowers – and who isn’t going to want that?
Microsoft’s own Work Trend Index research reveals that 77% of those surveyed did not want to give Copilot up after using it. They’ll certainly tell their colleagues about their exciting new productivity gains.
Right now, there’s a tension between the desire to democratize this technology and the need to bear its cost. The limited approach is winning for now, but will this remain the case after today’s announcement?
2. The early consensus is that Copilot needs fine-tuning, but there is also consensus that the work will be done in short order.
It’s well-known that Microsoft has a penchant for launching technology and enhancing it over time. Copilot seems to be following this pattern, but there’s a keen sense that improvements will be swift. This technology is too big to fail, so Microsoft will do whatever it takes to push the evolution envelope. Today’s announcement underscores this; personally, I didn’t think they’d be ready for a wide release for some time, yet here we are.
There’s an aspect of AI that resonates with the saying that advanced technology can seem like magic. Where we once had Clippy, now we have AI composing poetry – sometimes even good poetry. This technology has the capability to refine itself – an advantage Teams and SharePoint never had.
3. There are mic drop moments – be sure to collect them.
Here’s one of mine. When we first activated Copilot, we set up a small pilot group to test it.
For context, there’s a shift in our industry where partners like Microsoft and VMware are taking some customers direct. It’s compelling us to redefine our business models. There’s no malice in it, just another in a long line of evolve-or-die moments that are the hallmark of the tech industry.
We were interviewing for a consultant role on my team, and I asked Copilot to suggest a challenging question. In seconds it returned: “How would you approach a customer that is reluctant to now pay for services they’ve been receiving in the past for free?” Stunned silence. Just twenty-four hours after activation, it had drilled down to the core of the crisis we face.
Collect your mic drop moments and study them – they are precious cargo in decoding this technology.
4. Lean in on prompt theory.
One brilliant CSO I spoke with had empirical evidence that you get better quality responses from generative AI tools if you say please. AI is enigmatic – Copilot will give you breadcrumbs, but you still have no real idea how it gets from A to B. The way you phrase your prompts is crucial to the quality of the response you receive. A shared concern emerged from my conversations: a lack of education on prompt theory could lead to initial disappointment, potentially hindering long-term adoption. I experienced this firsthand.
When Copilot appeared in PowerPoint, I was eager to have it generate a presentation. The results were so awful that I didn’t try again until I saw a colleague do it and produce a brilliant result. It all hinged on the prompt – in this case, the prompt also involved structuring the underlying document to facilitate AI’s comprehension of the data. Upon revisiting my initial use case, I realized my poor structuring and lackluster prompt were to blame. The takeaway is that I didn’t attempt again until I witnessed a successful usage.
5. Everyone has security concerns – so think about your base license.
This is a big topic because everyone fears Copilot turning their data estate into a leaky faucet. While there are more questions than answers at this point, everyone agrees it’s vital to keep asking the questions. Consider your base license – it may be time to assess those InfoSec elements like Microsoft’s full data loss prevention stack, sensitivity labeling, and risk-based access controls. Different licensing plans offer different capabilities; the higher up the suite ladder you go, the more refined those tools become. If you already own M365 E5 as your base, take a second look at what you own because the toolset is substantial.
6. Everyone also has adoption concerns – so have a learning plan for broad implementation.
You’ll have power users who will quickly harness their Copilot superpowers. However, there will also be users who struggle to integrate Copilot into their workflows. Lack of time to learn new things, a poor initial experience due to ineffective prompts, or simple resistance to change could undercut the value of your Copilot investment.
Nearly everyone I’ve spoken to agrees that a well-structured, proactive learning plan is crucial for widespread success. Copilot itself provides some hints; it suggests logical follow-up prompts, and it’s quite adept at drafting training manual outlines. But no best practices have emerged yet – and even the formulation of a learning plan necessitates innovative thinking in the era of Gen AI.
Part III: What I think this announcement really means
Microsoft’s announcement releases Copilot into the wild. The barriers are down, and every business, government agency, and academic institution is now free to test small and think big. Microsoft has always been at its best when it builds a platform and lets us figure out how best to use it. Office, SharePoint, Teams – Microsoft built them and said, “Go!”
Copilot is woven through the fabric of the M365 ecosystem, but in many ways it is like a new platform. Wide-scale availability is going to lead us to interesting new places. Will adoption be measured or a steamroller? Whatever happens next, putting this powerful technology into so many hands is a milestone. I remember vividly when Steve Ballmer stormed the stage at the 2011 Worldwide Partner Conference, thundering, “We’re all in!” on cloud.
He wasn’t kidding – and Office 365 changed everything. Microsoft is all in on generative AI with this announcement.
Only this time, the technology is going to speak for itself.
SHI is “all in!” on Microsoft Copilot
- Evaluate AI priority use cases and spin up Copilot test groups.
- Understand how to integrate and adopt Copilot across your organization.
- Realize rapid ROI with Copilot deployment and integration.
Ready to explore how Copilot can transform how you work? Sign up for SHI’s Generative AI In The Microsoft Ecoverse Briefing or our Microsoft 365 Copilot enablement workshop today. And if you’re still in the information-gathering stage, contact us to speak to an expert about how SHI and Microsoft Copilot will help you solve what’s next.