
Community IT Innovators Nonprofit Technology Topics
Community IT offers free webinars monthly to promote learning within our nonprofit technology community. Our podcast is appropriate for a varied level of technology expertise. Community IT is vendor-agnostic and our webinars cover a range of topics and discussions. Something on your mind you don’t see covered here? Contact us to suggest a topic! http://www.communityit.com
Community IT Innovators Nonprofit Technology Topics
2025 Nonprofit Tech Round Table pt 2
Expert Tips on Current Trends:
Webinar 2025 Nonprofit Tech Round Table
- How is your organization using AI? What policies do you need? What should you be worried about?
Download the Community IT Acceptable Use Policy Template for AI Tools in the Nonprofit Workplace. - Steve and Matt also share a veritable "grab bag" of new tools, new chips, and other new issues to keep an eye on that will impact the nonprofit sector using IT in 2025.
Join CTO Matthew Eshleman and Director of IT Consulting Steve Longenecker, moderated by Carolyn Woodard from Community IT, in a lively and specific discussion of all things nonprofit tech for 2025 and beyond.
It’s like listening in on your smart friends talking about stuff you need to know about but don’t know who to ask.
Kick off the new year with a new understanding of trends and practices that can help your nonprofit succeed. This is one of our most popular webinars and podcasts year after year for a reason. We don’t believe a lot of lingo or jargon is necessary to understand what you need to know to manage IT.
As with all our webinars, this presentation is appropriate for an audience of varied IT experience. Part 1 covers our introductions, a discussion of hybrid work in the nonprofit context, and tips on new cybersecurity from our expert. Part 2 covers AI, a "grab bag" of new tech and new issues, and audience Q&A.
Community IT is proudly vendor-agnostic and our webinars cover a range of topics and discussions. Webinars are never a sales pitch, always a way to share our knowledge with our community.
Many questions asked at registration or live at the virtual event will be answered in the transcript.
_______________________________
Start a conversation :)
- Register to attend a webinar in real time, and find all past transcripts at https://communityit.com/webinars/
- email Carolyn at cwoodard@communityit.com
- on LinkedIn
Thanks for listening.
Carolyn Woodard: Welcome, everyone, to the Community IT Innovators webinar for the 2025 Nonprofit Tech Roundtable. My name is Carolyn Woodard. I'm the Outreach Director for Community IT and the moderator today.
Matt Eshleman: My name is Matthew Eshleman, and I'm the Chief Technology Officer here at Community IT.
Steve Longenecker: My name is Steve Longenecker, and I am the Director of IT Consulting at Community IT.
AI Policies at Nonprofits
Carolyn Woodard: We’re going to switch gears and talk about AI policies.
Another quick little poll, and that poll is, do you have AI policies at your nonprofit?
- The first option is, I don’t think so. I didn’t want to put just no, because it may be that there are some somewhere, but it just hasn’t been publicized, or maybe you don’t have any or such. So, if you don’t know, then just say, I don’t think so.
- Second option is, we are in the process of creating policies. We know that we need them, but we don’t have them fully created yet. We haven’t shared them with all the staff. We’re in the process of doing it.
And if you are looking at creating AI policies, we do have a free template download on our site as well under resources. So, you can download that to get started, but you do have to customize it to your own organization.
- And then the final option is yes, our organization has created an AI acceptable use policy and our staff understand our policy and it’s a living policy that people are using.
Matt Eshleman: So, do you have AI policies? I don’t think so, question mark is what 30% of our audience reported. A little over half reported that they were in the process of creating those policies. And then 15% are in the great shape of saying that they do have an AI acceptable use policy and that their staff are using it. So, kudos to those seven people who were able to answer that way for their organization.
Carolyn Woodard: Congratulations. All right.
AI Tools at Nonprofits
We’re going to go on and talk a little bit more about AI at nonprofits.
I feel like in the beginning of AI, really making this huge, big splash, there were a lot of stories about a few nonprofits using AI to achieve their mission. Using AI to do some kind of analysis that made it faster for them to reach more people or that sort of thing. And now, I’m seeing a lot more stories about nonprofits using AI intentionally or ad hoc to allow staff to do a lot more by using these new productivity tools.
So that’s something that we’re going to talk a little bit more about. I think, Steve, you were going to take the lead on talking about the hype cycle for AI and nonprofits.
Steve Longenecker: Yes. So, first of all, I wanted to steal a quote that I hear a lot on various tech blogs and so forth that I read. It’s attributed to Bill Gates, but apparently Bill Gates was just re-quoting a guy named Roy Amara, who’s an early computer scientist out of Stanford.
And the quote is basically that people overestimate how much tech will change in one year and vastly underestimate how much it will change in five years.
And I think that that is probably true about AI. So, if we were looking back a year ago and thinking about our predictions about 2024, and what happened in 2024, we may have overestimated how much AI would change the landscape, but we shouldn’t take that mistake and say, oh, well, then the five-year vision is similarly also not going to be very impressive, because that’s probably not the case.
Productivity Tools Using AI
We are definitely seeing growing use of AI. I would say that in general, it continues to be sort of a life hack for the staff, for the users that choose to use it.
It is very possible to do your job without AI, and most people are probably doing their job without AI. But the folks that have sort of figured out, “oh, I can use AI for this, and I can use AI for that,” are either being a lot more productive, or getting their jobs done a little bit earlier, and, knocking out a day, an hour early each day or whatever. Probably, since we’re all talking about nonprofits that are working hard for the good of their organization’s mission, they get more done in eight hours than they would have otherwise.
I can speak to this from personal experience. I’ve had a Copilot license courtesy of Community IT for most of 2024, and I have eyed it with suspicion at times, and other times I’ve gone ahead and tried it, and I’ve had mixed results. But the times that it’s really helped me, I have to say, wow, this really helped me.
I’m thinking of a time, for example, that I needed to do some analysis in an Excel spreadsheet, and I could have googled it and read some tutorials and figured it all out. But it was really pretty cool that in Excel directly, I could click on the little Copilot button and ask my questions in natural language. And then Copilot just gave me this formula that was several lines long, that I could copy out of the Copilot window and paste it into the cell. And then boom, I had the analysis I was looking for. And that was pretty cool.
That is what we’re seeing a lot of. We’re seeing AI built into the tools. If you have Teams or Zoom or whatever virtual meeting platform you use, you’re probably used to seeing artificial-intelligence-generated meeting notes. That’s wonderful. Taking notes while you’re leading a meeting is difficult. Having someone who is a formally assigned note taker is great, but if they’re trying to also participate in the meeting, that’s hard to do that double duty. So having an agent do it and do it decently is pretty cool. The expertise with which that agent does it varies. But I do think that they’re getting better all the time.
The other place where we’ve really seen incredible progress in the sophistication of what the tools are delivering is in the generation of graphics and video. It used to be that when I would do a PowerPoint, I really valued having some nice illustrations or graphics or whatever that went with it. I was always very careful about not taking things that weren’t in the public domain that I was allowed to put into my PowerPoints if I was giving a presentation to clients or whatever. But I ended up spending a fair amount of time scouring media wiki, places where these graphics were available free of charge.
If I had had a budget, maybe I would have gone to a graphic artist and said, can I have a picture of this or a picture of that? A key unlocking a lock, you know, something like that to illustrate my thing. Now you can just ask AI to do it. And people whose livelihoods depend on producing these graphics and videos as human beings are reasonably concerned that their productivity is being potentially replaced. Although, at this point still maybe they’re able to leverage that to make themselves more productive. I don’t know, but that’s definitely improved. That improved a lot in 2024, and I expect it will continue to improve in 2025.
What we have not seen much yet, and that’s what this bullet point Chatbots from Salesforce is, is this idea of top-down management saying, “we’re going to try to replace human beings with AI.” I think that that might be more of that, don’t underestimate what will happen in five years or a longer horizon. I think that these changes are just getting started, and I don’t know where it’s all going to lead, but we’re not seeing that yet.
AI is still very much an individual’s life hack, if you will, to being able to be more productive personally, and they’re not really having to show their work to anyone, and it’s not. So, you could make your staff potentially more productive if they’re willing to use these tools and are able to use these tools. But it’s not at a point yet where you can mandate the use of the tools or find that you can cut your work force by 10% because you’re going to have AI doing that other 10%.
Matt, do you have thoughts on what I said?
AI Chatbots Nonprofit Use Case
Matt Eshleman: I wanted to jump in there because I was at a conference last week, and actually, there was an interesting use case or a situation about this AI chat bot that came up. And one participant was saying, as a legal services organization, one of the services that they provide is a staff interactive chat service. People that have legal questions can use their chat service, to talk to qualified attorneys or paralegals to help get information about whatever legal services that they’re interested in.
And the organization has noticed a drop-off in their engagement on chat because they’re finding that people will just go talk to Gemini or Copilot about their legal question and get it answered that way, as opposed to using a more well-curated resource.
I think the challenge for them as an organization is going to be, now how do we respond to this changing demand and competition from these more bland, generic AI services that are just spitting out results curated from Reddit and all these other places? Maybe we need to look at building AI chatbot services on our curated knowledge as a way to do that.
I think that was an interesting window into some of the competition that exists, if you’re an organization that provides resources as part of your service delivery.
I do think most of the AI use and adoption that we’ve seen has been largely an individual productivity improvement, as opposed to organization-wide process improvement, or chatbot, or data analytics. I think that’s maybe an area where nonprofits are behind, because they don’t maybe have quite the same financial incentives to help justify the investment, to figure that out, because I do think it’s going to be expensive.
Steve Longenecker: I don’t think even in the commercial world that that many corporations have gotten that far along in more than the individual productivity.
Matt Eshleman: Well, I do think the numbers that you see in terms of developers like Google and Microsoft, I think the productivity of code creation, even though it still requires human review, but I think with processes like that, I think those big companies are able to see if they invest in this platform, now they can hire 50% fewer developers, and they can increase their speed to market and do all this other stuff. I think it’s something that there’s a scale question as well. But yeah, I’ll be curious to see how the trickle-down effect impacts smaller nonprofit organizations.
Grab Bag of New Nonprofit IT Tech and Trends
Carolyn Woodard: Well, I wanted to segue, I’m sorry to cut off this great conversation. But we have this great grab bag as well.
Google Storage vs Microsoft Storage
Google Workspace, I think you made this bullet point that they have added storage.
Steve Longenecker: Yeah, so I did it in the review of last year’s predictions. Last year we predicted that there’d be continuing tension between do I want to be a Google organization or a Microsoft organization? And I said at the beginning of this presentation today that, I feel like Microsoft continues to really be a strong investor in the Microsoft 365 platform, particularly on the security side, obviously the Copilot AI side as well. It makes sense. That’s their bread-and-butter business. That’s what they are. They’re a services company.
Whereas Google is driven by search first and foremost, is driving their revenue by selling ads against those search results. So, it’s just a different business model and cultural incentive structure for them. I don’t see them as investing in Google Workspace as much.
That said, we see it is really interesting how I think it was in 2024, if not, it was in late 2023, that Google started rolling out 100 terabytes of free storage to their nonprofit charity customers. So, if you had the charity level license, you suddenly were able to avail yourselves of 100 terabytes of storage. And if you’re not a tech person and not sure what a terabyte is, it’s a lot. It’s a lot of storage. It’s an awful lot of storage.
In contrast, a lot of our small nonprofit customers have something like, one, one-and-a-half, two terabytes of SharePoint storage. And we are seeing our clients running out of SharePoint storage. It’s not a daily event for us. It’s not even a weekly event. But we see clients’ tickets come in, running out of storage. What should we do? And then, you’re scrambling to try to delete old files that no one needs. The value of doing that, it takes labor if you make decisions. Do I keep this? Do I throw this away? Most of the time, it’s cheaper just to buy more storage.
But Microsoft is not making it free. And it does have, ultimately, it’s a never-ending cost. And our clients that are having to do that obviously don’t like it.
So, it’s really interesting, this contrast. I don’t know whether that means that Microsoft is just at some point going to have to have parity with Google. “We’re going to give away a lot of space, too.”
It used to be that we talked about storage being free. But, and of course, it was never free, but it felt like it was free. And we’re now seeing on the Microsoft side, the impact of it, of the fact that there are limitations, and we do have clients running up against it.
That’s just an interesting contrast. Otherwise, I feel like the two platforms continue to exist and continue to be good enough. Microsoft is getting better all the time. Google is holding steady, but it’s okay.
Subscription Model and Nonprofit Discounts
Carolyn Woodard: Thank you so much. Matt, I think you were the one who talked about the annual commitments and the low discounts becoming the subscription model.
Matt Eshleman: Yeah, I think Microsoft finally increased the price of Microsoft 365 after it being basically the same price for a number of years. This is kind of around the edges. The nonprofit community continues to get very good discounts from Microsoft.
But at the same time, we now have made that switch from “we’re going to buy the software one year and use it forever,” to now we’re switching to a monthly recurring cost. Now, we’re seeing vendors switch to say, “hey, you actually need to do an annual commitment to get that same pricing. If you’re on monthly, we’re going to raise the price.”
There’s going to be some differences there. I think to some extent, the original promise of this cloud utilization and it being dynamic, and you just use as much as you want, and you don’t, you just try it out. I think it goes away a little bit because most organizations are not dynamically changing and you have relatively static, staffing loads. I think your software subscription is just going to cost a little bit more, and it’s still better and cheaper than trying to do it yourself.
Carolyn Woodard: For sure. Trying to keep a server running, as we talk a lot about, is recurring costs that are a lot more than what you’re going to pay in subscriptions.
ARM Chip for Windows
Matt, did you want to talk about this ARM chip for Windows?
Matt Eshleman: ARM is a different chip architecture. A couple of years ago Apple made the switch to their M platform, switching away from Intel. They got tremendous battery life gains and performance gains.
Microsoft or Windows is a little bit late to the party. But we’re seeing a similar shift now away from or in addition to Intel based chips, to ARM based chips, which are supposedly going to have much better battery life.
The good news is that it’s probably going to be largely transparent to end users. We’ve tested all of our endpoint management tools, our security tools, everything just works with the ARM chip architecture. We’re waiting to see if the battery life and performance demands pay off. It doesn’t look like it’s going to be that much cheaper. I think it’s going to be a performance side play instead of a cost saving side play.
Carolyn Woodard: So something to keep an eye on.
Steve Longenecker: The chips that are traditionally in phones are now going into computers and they run a lot cooler. They’re totally different architectures. So, the software stack just basically needs to be rewritten. And so that’s why for Windows, it was a longer rollout, I think. They’ve been around for a while, for quite a while, but we can’t really try these computers out because half the software we use doesn’t work on it because it’s not been rewritten for it. And I think we’re viewing 2025 as the first year that we’ll really start. We don’t think we have any of our clients buying ARM chips, in ARM Windows computers yet.
Matt Eshleman: I think there’s one other computer other than our test machine that’s in our management fleet of 8,000, right? So over 6,000 Windows computers.
Steve Longenecker: But the promise of the 12-hour battery life, perhaps, and also just running cooler. I mean, everybody’s experienced how hot a laptop can get, but Windows laptop can get really hot when it’s really running fast and furious. So, it’s a good possibility. I’m excited about it. We’ll have to sort of see how it actually plays out in terms of, does all the software actually run.
The word on the street is just that for gamers it’s a non-starter right now because all the game software doesn’t run on except on the traditional Intel chips. But probably most of us are hoping our staff are not playing games on their organizational laptops as much as they’re word processing and browsing the internet for whatever.
Windows 10 End of Support
Carolyn Woodard: While we’re on Windows, do you want to address this point of the Windows 10 and the end of support, which was a question at registration?
Steve Longenecker: Yes, yes, thanks. I’ll quickly take that one. So, Windows 10 is going to reach the end of support in 2025. I think it’s October of 2025. At that point, Microsoft will stop issuing patches for that operating system. So, any bugs or particularly vulnerabilities that are discovered, and they’re discovered every month, and that’s a lot of what the security patching is, is covering up, correcting vulnerabilities with code updates.
That will stop for Windows 10 in October of this year. We’ve been encouraging our clients buying new computers to buy Windows 11 computers for quite a while now.
The difference for users of Windows 10 and Windows 11 is, they look a little different. There are definitely differences, but we don’t feel like you need to have a long training course on it to make the adjustment. It’s pretty intuitive. So, it’s not that big a deal to have a mix of Windows 10 and Windows 11 computers.
At some point, we’re going to push Windows 11 out to the Windows 10, and when I say at some point, it has to be soon because we have until October. So, that will be a priority in the next few months for our clients.
The question that came with the registration, Carolyn passed along to us was, “What do I do with the Windows 10 computers that are not able to be upgraded to Windows 11?”
And there are a few things that computers have to have first and foremost among them, I think, and Matt, you can correct me if I’m wrong with this, but they need to have a TPM chip in them for Windows 11, for them to be able to be upgraded to Windows 11.
If you have a computer that can’t be upgraded to Windows 11, the question is, what should I do with it? I would tell you to make sure that you recycle it as responsibly as you can.
Don’t just put it in a landfill. But you are done. That computer is over. I’m sorry to tell you that, but it has reached the end of its life.
Mac Parity with Windows Hello and Platform SSO
Carolyn Woodard: Mac parity with Windows Hello, and the platform single sign on. I can’t remember now who wanted to take that.
Matt Eshleman: This is a little bit of a tech geek out thing, because it ties into a couple of underlying technologies. Steve already mentioned one of the acronyms, which was TPM, which is a security chip that is in Windows computers. So, in the Windows world, Microsoft uses that TPM chip to secure and generate unique security IDs for you. Then you can sign in with your Microsoft 365 credentials, and then that TPM can store some information and is a secure phish-resistant MFA method that can be used then to log into websites and other applications.
That’s been a super handy feature for Windows users and Microsoft 365 to basically have one username, strong password, and it’s all self-contained within the Windows platform. That is not something that was available in Mac.
If you’re a Mac user, I actually own a Mac right now, and logging into Microsoft 365, you have your local ID that you use to log in to your Mac, and then you have a separate login that you use to log in your Microsoft 365 credentials.
Microsoft has now released a feature. It’s a preview, technically, called Platform SSO that basically allows you to use the secure chip on the Mac called Secure Enclave to basically do the same thing. And so you can, with Platform SSO configured, have single sign-on to access your Mac, to access all your Microsoft 365, to use the secure fish-resistant NFA method.
It’s going to be a really tremendous security benefit to streamline things, to streamline administration and make things a lot more secure for your Mac users and to provide that feature parity. It’s the same basic things that you can do on a Windows computer you can do on a Mac computer as well.
Carolyn Woodard: Wow. That is so great. I know that’s different.
I just want to shout out to our amazing chat. We had a question in chat about how to find out if you’re on Windows 10 or 11 after the discussion about the computers that are going to be useless. Someone in the chat actually helped that person, so I appreciate it. Thank you very much.
Data Silos at Nonprofits
We wanted to talk a little bit about data source silos as an ongoing issue, as data and I guess AI search as well as becoming a bigger thing. We still see a lot of clients that have silos. So, can you talk a little bit about that?
Steve Longenecker: I think we put this on the list mostly to acknowledge that most small nonprofit organizations do have their data in individual silos. And while there is a move afoot in the greater technology world to put these things into data lakes and make it possible to sort of mingle all your data and bring it all out with AI queries and other kinds of queries to get the data out, that most of us are not at that place and probably won’t be at that place in 2025 either. So, it’s more of a note that there’s some cool stuff on the horizon coming, but probably not something that the small nonprofits are going to be able to do. It hasn’t trickled down yet to the smaller organizations. That’s my takeaway from on this topic. Matt, do you want to add anything?
Matt Eshleman: Yeah, I do think that this is an area probably where nonprofit clients are probably ahead of us even, because I think organizations really know their data and know it well. Over the past couple of years lots of investment has gone into this area to take advantage of the data that is there.
As organizations are prepping to use AI services beyond just individual productivity enhancements, having a good foundation of data to access and use and to feed into these systems is really valuable.
If you’re thinking about adopting AI internally, it’s really helpful to know the Copilot world, right? Copilot has the same permissions that you do. And so, if your SharePoint site is just completely open and you’ve relied on security by obscurity, and now somebody is searching for, “hey, what’s the payroll information for my organization?” that information would be surfaced if you haven’t adequately restricted your data.
Organizations that are trying to think about, well, how can we use data or AI effectively? We don’t really have a capacity to do that right now, but maybe we have capacity to really take an audit of what data systems we have, where does it live? What is security? What kind of data is contained there? Do we need to make any additional controls or restrictions?
I think that’s a good place to start, just understanding your data, where it lives, what’s contained within, so that then you can take advantage of it, and be more effective with it later on.
Carolyn Woodard: I think as long as I’ve been in nonprofits, which is 25 years now, that’s been the killer app we’ve waited for is how can we easily find data without having to clean it up?
I just listened to an interesting podcast a couple of months ago from a company that does this for nonprofits and talked about data lakes, data warehouses. There is no magic AI tool that you can just give it your confusing, disoriented, not labeled correctly data and have it tell you what you have and have it find all of the things that you need it to find. I’m sure it does seem like that is something that the next few years is going to be really talked about and the tools are going to come out, especially with AI being able to do more and more. But I still think you have to do the work. You have to look at what your data is, where it lives, who has permission to it, and make some of those decisions about what you’re going to archive.
Again, I want to shout out the chat. There was some really good advice on being able to figure out what you need to archive and how to do that. So, thank you again to the chatters.
Policies that Help Protect Organizations or Staff Who May Be Targeted Now
I think we have one more question, which I thought was a pretty good one that came in through Q&A. What other considerations should an organization consider with AI and other use policies that help an org protect its staff who may be targeted in this administration or this political climate, shall we say?
Do you guys have some ideas about where to look for support on protecting staff?
Matt Eshleman: I think there’s a couple of different things in there in terms of protection. I’ll probably start with a new thing. Maybe this could have made it onto the trend list.
I think a new thing that we’ve started to see in 2024, and I think we’ll continue in 2025, is protecting an individual’s personal identity against doxing or harassment, online threats. I think a lot of the work that we’ve done with Community IT and other kinds of cybersecurity stuff has been around protecting your organization’s digital identity, your work computer, your work email, all the stuff that is around your work identity. And we know you have a personal email, but that’s separate and we don’t really think about that.
I think from a cybersecurity threat model perspective, certainly for high-profile organizations or policy organizations, your personal identity is kind of in play as well, right? Threat actors would target your personal email, your home, your spouse, other people in your house.
Carolyn Woodard: Find your home address.
Matt Eshleman: Right. They aren’t just targeting your work email; they’re targeting you as a person. That is now extended to online bad actors. It’s not just Chinese state-sponsored hackers that are going after people individually, but we’ve certainly seen it, online doxing people that have a different political identity as you, going after you as an individual, because of your politics or your identity.
We are seeing more organizations and more services that are focused on doing some basic stuff like removing your personal digital identity from web services. And then, individuals that may be targeted beyond that, these services can remove all aspects of your digital identity from the web. That could include things like putting your house in a trust to remove your name so your address can’t be located.
I do think that there is a lot of movement in that space in terms of really protecting your personal identity against that malicious online actor targeting or doxxing you as an individual.
Policy Templates/Fundamental Policies Nonprofits Need
Carolyn Woodard: We had a question come in about where to find policy templates and which policies are necessary. I’m going to share a link to our governance webinar that we did last year that has a bunch of links in it to other sites that have good templates and good information on setting up those governance policies.
What’s a Data Lake?
And we have a quick question. What is a data lake? Anyone want to take that?
Steve Longenecker: I realized I jumped into that question without explaining what a data source silo or a data lake warehouses, data lake houses were. The idea here is that your data lives in a service, a database, if you will. Your Microsoft 365 data, which might be your email and your SharePoint files and your OneDrive files, your Teams chats, that’s all in the Microsoft 365 silo.
Then you might also have a finance system which has a bunch of accounting records, but that’s not probably Microsoft 365, although Microsoft does have Microsoft Dynamics, so it’s possible that it’s also in Microsoft, but more likely it’s a different system.
Then you might also have a system for managing your organization’s customers, or contacts, or stakeholders, it’s called a CRM generally, donors. And that’s a different system.
And if you’re a school, you might have a student information system, and you might have multiple student information systems, because there might be one tied to the charter school board that you report to, and another tied to your grading system, and another tied to attendance, or whatever.
How do you bring all that together? Because obviously the same identities or query points can appear in all the different databases, but since they’re in silos, you have to address each one individually.
The idea of a data lake is that can we dump all that stuff, somewhat unstructured perhaps even, into a single place, and then use AI to connect the dots and bring out the information that we need without worrying about which silo it originated from. And yes, that is happening.
There is a lot of work being done in that space, but at this point, these are massive investments being made by very, very large organizations. And I’m sure there are very large and just medium-large nonprofits that fall into that category that are doing something like this. But most of our smaller nonprofits can’t afford it and aren’t quite there yet.And like I was saying, the trickle down hasn’t happened.
Those tools will probably get more efficient and more inexpensive over time, and it will become an easier thing to do. But at this point, we’re not at that point.
At this point, there’s a lot of value to using Microsoft 365 if you’re a Microsoft 365 organization as much as possible, because then everything’s in one silo. That is, you might not be able to do your, everything in Microsoft 365 because Microsoft 365 doesn’t do everything, but do as much as you can there. Don’t have a separate system for files.
Don’t use Dropbox for files, Microsoft 365 for email, and Zoom for chat, because then there’s three different systems. That’s okay, and maybe that works best for you. But if you want to have it all in one place where one Copilot can access all of it, you might do better to have it all in Microsoft 365 or Google with Gemini if you’re a Google customer.
So that’s a better, more grounded explanation for what I was trying to say earlier. And I apologize for not even saying what it was before.
Carolyn Woodard: But like we said, I think that just demonstrates that data, where your data is, how you access it, how you use it with these new tools to make it more productive and to do clearer analysis on it to connect those dots, that is going to be a continuing story.
So, when we do this next January, we’ll definitely be talking about data, whether it’s in lakes or lake houses or warehouses, I don’t know, but they’ll be important. I mean, everything is changing so quickly.
Learning Objectives Recap
Hope that we did cover these learning objectives, talking about the hybrid work, the security attacks, what to know about AI and then the grab bag.
Before you go, I want to make sure to mention that our next webinar is the Cybersecurity Awareness Training Tips with our team members, Matt and Anna. There are many tools out there to help you move cybersecurity training from a boring and maybe useless once a year video to a dynamic and up-to-date regular mini training that will give your staff real experience identifying scam emails and learning first steps to take if they get something suspicious.
Matt, who’s here today, is our resident cybersecurity expert and our CTO and Anna manages the program that we use with the clients. They have so much experience. If you have been thinking about this, you haven’t implemented it or you’re struggling with having cybersecurity training that your staff are really doing, I’m really excited to have Matt and Anna give us their advice.
That’s going to be at 3 p.m. Eastern Noon Pacific on Wednesday, February 26th.
I want to thank you, Steve and Matt, for staying extra, staying after school for a few extra minutes to answer a few more questions.
This is just always such a great webinar to be able to ask you what’s going on, what are you seeing, what are the trends, what’s happening. I appreciate your time today and to everybody who stayed on also extra, thank you for staying.
We will be putting this video and the transcript and the podcast up as soon as possible so that we can share all of this information with all of you and we’ll see next January what we were right about and what we missed.
Thank you, Matt and Steve for joining us.
Matt Eshleman, and Steve Longenecker: Thank you.