Chris Heuer (00:00) she posted in the community on a Tuesday and she was patient about it. Two weeks later, she emailed support, and by then she wasn’t. The CRM had her as an account in good standing. Renewals in 90 days, there’s no flags, she’s gonna renew.
And somewhere in a call recording, somebody had no time to listen to, Her CSM had promised her a fix by the end of the quarter. But the thing is, is that all these different versions of her and these different elements of the conversation and the relationship were in four different systems, four different records, and essentially four strangers who happened to be the same person. Now, nobody inside the organization failed here. This is a systems problem. And that’s the part we’re sitting with, right?
Chris Heuer (00:38) The community manager and the CSM didn’t miss anything because none of it was theirs to see.
Good morning, good afternoon, good evening, wherever in the world you might be today. Thank you very much for joining us on our second episode of the Community Assistant.ai podcast. I’m joined today, of course, by my co-host Marius Ciortea CCO, the Chief Community Officer, for those of you who don’t know that, of Higher Logic. How are you doing today, Marius?
Marius Ciortea (01:31) I’m doing great and I I keep forgetting what CCO stands for. so thanks for reminding everyone.
Chris Heuer (01:38) Well, it well, you know,
It’s all immaterial to the work we actually do, except to the people outside who need to be able to understand the context in which we work, which by the way is the topic of today’s episode. But before we get into that, Marius, we found an interesting piece of news today that we should probably talk about.
Marius Ciortea (01:55) Yeah, I I I love the message from Anthropic today that we should take it a little slow on the AI development side. maybe we’re moving slightly too fast, especially when we cannot control what AI agents do once they escape the sandbox. so kind of scary stuff, but good to see that the industry is recognizing that they’ve may lost a little bit of control on how fast their
building and they’re building things faster that they can actually create
Chris Heuer (02:23) Yep.
Marius Ciortea (02:24) safeguard rails around it.
Chris Heuer (02:26) Well well given the story that was around Anthropic with the researcher quitting, saying that it’s too dangerous, that’s a really good time for them to respond to that. And of course, didn’t Sam Altman say something similar recently?
Marius Ciortea (02:39) Yeah, I think the same story is is across the industry. Now I don’t know, you know, as they have the pressure of going public, how much of this will they actually follow? So hopefully some policy will follow as well that can create the right balance between innovation speed and and safety. And of course, we have competitors around the world that
may not follow that policy. So that has to be kept
Chris Heuer (03:07) Yeah.
Marius Ciortea (03:08) in account as well. So it’s a it’s a challenging time. It’s it’s great that it’s being recognized. I don’t know that there’s a clear solution for any of it. but but I’m happy that it’s at least being discussed now.
Chris Heuer (03:18) Well
well, this is where and why, you know, they’re not just doing it unilaterally. because they’re still in a race to super AGI and the winner gets trillions and trillions of dollars. And so that’s that’s where they’re not willing to just do it on their own. But hey, that’s what the article from the anthropic CEO, Dario Amodei came out and said yesterday or the other day, I guess, is that government really needs to slow down the development, not stop it, but
Marius Ciortea (03:44) Right.
Chris Heuer (03:44) slow down and regulate it. And of course.
Our current government is not going to do that. And we don’t need to get into that any more than that. It’s just the reality. and they’re not gonna self-police. So where are we left as a society if the gatekeepers aren’t gonna keep the gate and the individuals inside the asylum aren’t gonna like lock themselves in?
Marius Ciortea (04:05) Right. Well, for the good news is on our end, you know, we don’t have that much control. So it’s good to kind of watch it from the outside, but we are in control enough our communities. So I think that the reason we have the show is like to you know work on how do we make community assistant
be our assistant, not actually running the community. And human interactions are become more and more valuable. so we need to in a way protect it, and in another way, you know, enhance it. and I think the enhancement part AI can absolutely help.
Chris Heuer (04:39) Well, and and that’s the key. One of the things, this is the double edged sword of the whole thing and why I’m using it while still trying to advocate for human-centered AI, is that it is a tool and it does open up an infinite number of new possibilities of things that we can do. and as somebody who’s been a bootstrapped entrepreneur for most of my life, the fact that I could rebuild an application that once cost me $30,000 and took 68 days to build, that I can do that in about three hours.
is quite amazing and the compression of time is really interesting. But it’s about the technologies making things more affordable and accessible. And I think that’s really the thing that we have to focus on here a little bit more. The affordability that this creates for us to do things that
we couldn’t before without spending a lot of money and a lot of time on. and that opens up new possibilities. And in fact, just looking at the episode page the other day, I was able to go in and use the new.
what what’s the model called from GPT again? Astra? No. Yeah.
Marius Ciortea (05:41) Astra, yes. Or GPT six, you can call it like that
Chris Heuer (05:45) Got it.
Marius Ciortea (05:46) as well.
Chris Heuer (05:46) Okay. And so I was able to use that, give it a very simple prompt, and it gave me a complete page that it spun up in its own container, and it let me look at a new alternative view for how we might present our episode in a different context. And so that’s one of the ways that got us thinking about the topic for this week. but it’s more than that and it’s deeper. you know, Marius, do you remember back in the day when we were talking about the three C’s of the internet?
content commerce community. And that was essentially how we framed our approach to looking at the internet.
Marius Ciortea (06:19) No, tell me more. I I forgot that part. I’m much younger than you, you know.
Chris Heuer (06:22) Well, in the in like
you you actually are. so
Marius Ciortea (06:27) Just kidding.
Chris Heuer (06:28) I always realized at the time the context was key. And in fact, this was part of the the conversation on the page view a little bit as well in creating different versions of how we present the data. and started thinking about the idea of in the early days when I was working with a client like Sea-Doo, for instance, from Bombardier, one of the ideas we had was that the home page would have a path for the adventurer, for the racer, and for the family.
And in this way, create a context of the story we wanted to share with them and to create for their world that they could relate to so that they could get the information they needed from us and we could communicate with them in context. So that’s why, while everyone was saying content is king, I was trying to push the idea that context is king.
Marius Ciortea (07:10) Yeah, I think that’s super important and I think that theme is reoccurring over the years. And then having grown up in marketing, you know, everybody always talks about personalization and
And how in personas, right? I mean, those are just ways of describing sort of like context, right? The the right content to the right person, it becomes in context and relevant. The best possible content to somebody that doesn’t understand it, it’s completely irrelevant, right? So that’s why I think context is always king, like a badly written article to the right person that understands it.
than the best possible article to somebody that has no interest or doesn’t understand it.
Chris Heuer (07:55) Yeah, context also for me brings to mind a really important word that most people you’ll get what it means, but you know, most people probably haven’t even heard, which is the idea of situatedness, which is the nature of my situation and where do I sit within it. you know, I’m watching this show Dark Matter on Apple, and they talk about superposition from a physics and and metaphysics, quantum
Marius Ciortea (08:15) Mm-hmm.
Chris Heuer (08:16) physics sort of thing. And it’s about the the individual where they sit in the world of all things, whether you look at it on XYZ coordinates.
But it’s about like what’s around me. All these different facets and elements create the whole context. And so if we understand the context, we can do more for an individual to meet them where they’re at and to serve their individual interests. But today, most people, when they’re talking context engineering,
particularly in the application development space and and in the AI space, they’re talking about a package. They’re talking about it as a bundle of data itself, and they’re missing the root of this whole thing. so I went ahead, I’m I’m going a little long here, but I want to wanna go ahead and make sure we understand this. The origin of the word context is actually contextier. Con meaning together, and textir meaning literally to weave.
So the idea of context, the the root, the etymology of it is to weave together. And I think that’s a really important thing that sets us up for understanding how this applies to our community world. But first, before we get too much deeper into that here, you know, isn’t the whole point here about being able to build the relationship, and of course our show is about that human element of this and how we maintain it.
Marius Ciortea (09:35) Yeah, so well first of all, I want to say I’ve been a context engineer for 20-something years. I think as a community manager, that is what communities are. You you’re context engineering. think about like the time we spent of like
When we’re building a community, w what groups should we have in the community? How should we call those groups? What are the categories? How do people find each other in context of what they do? Do we name it by product? Do we name it by skill? Do we name it by any other attributes that we have? because our main goal always has been in a community is to connect people. And you connect people by finding the right attributes that are a match, right?
So we’ve been community managers have been context engineering from the second they they’ve became a community manager.
Chris Heuer (10:26) Yeah. Well and and I believe the the one of the main roles of a community manager is to weave the community together.
the idea of how do we actually bring together the fabric or the tapestry and the rich tapestry of our community behind it.
Marius Ciortea (10:40) Because that then leads to the engagement part, right? It’s you know, it’s all about connecting people and you need to have the the right context. And I think that’s where AI is is really helpful, right? one thing as as an example, if you own a community, you know, have AI look at your context, how it’s built, and let it see if there’s other opportunities that you have not thought about
Right. Maybe they they are different interest group that AI can surface to you because it finds actually patterns of discussions. Maybe it discovers that a lot of people like to talk about a topic that there is no group for it. and now you have the opportunity to create that group and bring those people together. that’s a good example of how AI can identify look at patterns and create sort of like contextual things that we may have not seen beforehand.
Chris Heuer (11:34) And and this is actually one of the core theses of the show. And that’s why we call it the assistant, because it’s a tool that assists you in doing things. It is not the look, AI is not going to necessarily improve the relationship. What it does is it makes the ability for you to improve the relationship by weaving together this context more affordable. And that’s really the key differentiator in all this stuff. And it’s important for you to take on that mindset to be able to get the most value out of using your AI, which is why we’re diving into this topic today.
Marius Ciortea (12:04) Yeah. So Chris, you had a an example of like context beyond the community and how it can matter in your job and the things that we we have challenged today.
Chris Heuer (12:15) Well, you know, it it’s funny because this is what I did with Palm as well. and looking at the Palm economy and and even in looking at the work that I did with the American Heart Association and helping them rethink their engagement strategies was identifying all the different stakeholders in this sort of thing. And what we found with Palm, for instance, was that there was an individual who was both a a customer, they were a developer, because they built their own apps for Palm, and then they were also a reseller.
So they literally wore three different hats, and we had to understand their unique context. And it was really hard to actually reach them.
so context was never just the stuff sitting around the thing and the facets in the data itself. It’s the weaving of it. And this is actually, I believe, one of the new jobs that the community manager can take on.
Marius Ciortea (13:04) Yeah, I think it’s it’s I think it’s always been part of their job, as you said. it’s just that the it was not at their fingertips, right? It was just very difficult to actually get. because everybody does sit in a different department, we don’t have access to the same tools. so
having that one place or having the ability to pull this all together just didn’t exist before. I think part of the responsibility of a community manager now, so I think responsibility would be better than job, but is to when we want to get involved in a conversation to understand where the person is coming from, right? And a simple prompt in the AI would be something like, hey, you know, before I answer this question, can you tell me
All about the customers and a customer journey. And AI will go and pull from the CRM data, from the support data, from the CSM data. Maybe they use Gong or something to record the calls and provide the community manager the context of the conversation, why the conversation even was started, and then answer the answer becomes so much more meaningful because we now understand the person asking the question.
We it’s only like we we build the relationship ahead of time because we understand who we’re talking to. and by understanding who you’re talking to, you are far more authentic and real versus having some cold generic answer, like, Hey, I will look into it.
Chris Heuer (14:35) So, this is why we end up with this show topic today here, because as we’ve just kind of really stacked the whole story together, you know, AI doesn’t improve the relationship. Context is what improves the relationship, and the relationships that communities have on a human level improve the relationships. But what AI does, and why it’s important, why we need to look at it as community managers, is it makes that context more affordable.
It allows us to be able to look at those four people as one and understand that individual, that human, not those four different views we have on them through the different lenses and applications. And that’s why today we’re we’re going in to talk about context. Now
Chris Heuer (15:15) Hey, sorry to interrupt, but I need to thank our sponsor, Higher Logic. This episode and all episodes of this season one is supported by Higher Logic. The Vanilla platform is where customer communities live. Discussions, Q&A, knowledge base, ideation, moderation, all in one place and connected to the support and CRM tools your team already runs on. Higher Logic is community first. It’s not a feature bolted onto something else. It’s a whole product built by people who have been doing this for more than fifteen years. Here’s why that matters on this show.
When AI answers the question, vanilla is where the humans show up next, where members find each other and where trust actually gets built. Visit them at higherlogic.com and tell them that you saw them on communityassistant.ai. Now back to the show.
Chris Heuer (16:03) Marius, you were talking to me about a few things that really kind of led to this episode and and talking about this.
in terms of the different things that you had. And I think the one that really struck me that that, you know, kind of ignited this whole thing was just getting the daily brief. And could you tell me a little bit about how that daily brief works for you and how you’re looking at aggregating all these different versions of an individual to create the context so you can understand who they are and then act upon that accordingly.
Marius Ciortea (16:31) Yeah, so I think Daily Brief is if you as a viewer not familiar with it
highly encourage you to put that inside Claude or ChatGPT or your LLM of choice because it’s something that gets you used to on how to use AI and provides context to your personal life, to your work, to whoever makes sense. In this case, it’s the community, right? what I do and my team has every morning they’re getting a brief on what is current in the community.
It looks at and and this is all you know individual, like I’m sure every community may be different, but for me, what I want to know is what are the latest discussions? You know, are there any discussions that are negative and I should pay attention to? So prioritize the latest conversations and the active conversations for me by a priority of is it negative, is it contentious?
Is it something I should answer immediately because the customer warrants it? and the way the AI does things in the in the back end, and I know it’s kind of long-winded answer here because it’s it’s more complex than just one sentence, but before the AI kind of puts the prioritization together, it actually goes out, looks at the individual, and looks at all the various data sources available to us to get context on the individual, right?
So as an example, somebody may have a negative or a question of something that’s broken. or and there are two people that have the same same issue
Chris Heuer (18:12) Well, I think
Marius Ciortea (18:15) of something being broken. The AI goes out and says that individually A is an account that is ten million dollars ARR, just coming out with a number. And customer number two is an ARR of hundred thousand. Now
It’s hard for you to say like customer two of hundreds of thousands is less important than customer A with a ten million. But the reality is that as an owner of a business, I need to pay attention where the revenue is coming from. So when it comes to the priority of things and the context of things, I want to know immediately that the guy who is a very big account has an issue and how I can potentially solve it. Secondary is I wanna know, and the AI goes and looks at
What were the latest conversations with the CSMs? What is about the same topic? Is it a f something that is a frustration? Meaning is repeatedly coming up, or is it the first time that they announce that? And then lastly, is anybody doing anything about it? Right? it could well be that it’s like, hey, this something that he already had a ticket open and the ticket has not been solved for two weeks, and now he’s coming to the community to get an answer.
Or it’s the first time he’s asking and he has not opened a ticket. And the recommendation may be to open a ticket to expedite this.
Chris Heuer (19:30) Yeah.
Marius Ciortea (19:31) And not only give the recommendation or open a ticket, but maybe he’s even giving me information of to whom to open the ticket, because that person has a certain expertise in the area, the problem happens, right? So imagine now if you would have the issue, you put something in the community, I get a daily brief saying.
Chris is frustrated about this particular feature. I now go in and not only help answer, but I can say, like, hey Chris, you know, I contacted John and who is the expert in this area. he’s actually now looking at and has opened a ticket up. I expect that to be open fixed in the next couple of days. let me know if I can help in another way. So now I only not only know like I know that it’s the first time you opened it.
Chris Heuer (20:19) Okay.
Marius Ciortea (20:20) I know
your account and how important it is to the company. And I know who’s the best person in the company to solve it, right? That’s an that’s a super important
Chris Heuer (20:28) Yeah.
Marius Ciortea (20:28) thing. It’s not just like I asked somebody in engineering to fix it. I have actually a real person. And that real person is now mentioned in the thread. And he may go back in the community and have the conversation and and help solve this, not just for the customer, but maybe for others as well. So because I now have this context in in the daily brief, I can do so much more.
Where before it would be just really Chris has an issue. And that’s where it ended, right? And
Chris Heuer (20:55) Well, that’s not where
it ended because from there you’d then have to search out what is that hour conversation with the CSM? Did they put notes in or not? Are there notes there?
Marius Ciortea (21:01) Right. I to be honest,
I would have not even done that. ’cause
Chris Heuer (21:06) Good.
Marius Ciortea (21:06) I don’t have the time.
now it’s just served to me in the morning without even having to do any work, right? It’s kind of it it tells the AI tells me what the best next action is.
Chris Heuer (21:18) But but even then you you would have had to go into different systems. It would have taken you 30 minutes just to assemble your own context in order to be able
Marius Ciortea (21:25) Yes.
Chris Heuer (21:25) to have the call to resolve the issue. So again, this is where AI makes context more affordable because we’re able to actually build systems that do it. And this is one of the reasons we talked about this in the pilot episode, and we’re gonna keep talking about this. Why w community managers really need to adopt this well, every professional, but community managers especially given their purview, need to
Have a systems thinking sort of perspective on this stuff. Need to think like an architect in how they build these systems. And that’s where the technology of AI makes this more affordable. And that’s where we get into something. We’re not going to get too deep in the technical, but that’s where we get into the MCP server and why it’s so important that community platforms actually enable the accessibility of the data and those conversations vis-à-vis an MCP server that makes it available to an to an AI.
And from there, we need the same thing from our CRM systems and from elsewhere so that we can get that whole view together. And if we just establish those connections and we build the systems design in that way, then we can actually query and access and build that report that comes out every day. We can build an automatic notification on the update when it changes, or if it doesn’t change by a given amount of time, and we can stay on top of it and do the work that we’re uniquely supposed to do.
Because we have the data and the tools within the system that enables us to actually build strong relationships, serve the needs of the customer, and deliver an exceptional customer experience.
Marius Ciortea (22:56) Yeah. To that point, just real quick. One, I think that MCPs are becoming table stakes.
Chris Heuer (23:02) Hmm.
Marius Ciortea (23:03) I don’t think there are any components to a or
tools within a company that will not have MCP. I think it’s just table stakes.
and yeah, and I think it’s gonna make this context thing really valuable, not only for the company, but for the customer. Because me as a customer, I wanna be understood, right? it would be so much better when I’m stuck in an airline, my flight is delayed and I call United or Delta or whoever.
and want to rebook that the first thing they would say to me was like, Hey Marius, I’m so sorry you’re stuck in whatever city. I already see a flight that you should be taking, versus me spending the first 20 minutes explaining that I’m stuck because of this and this and this, right? This is where AI
Chris Heuer (23:58) Yeah.
Marius Ciortea (23:59) can help us. Like it’s it’s gonna predict things almost. Like when you’re making this first call or you’re making this first post, it already knows what.
about you and what your situation may be and how it can help.
Chris Heuer (24:11) And and of course in that situation, the first element of the context is the fact that you call the medallion or the platinum line. And so
Marius Ciortea (24:18) Right.
Chris Heuer (24:19) that is actually that gateway model that we were talking about before with these home pages and having different gateways on the homepage for who I am and and and what am I trying to accomplish. But then we’re talking about bringing this other stuff in. But wait, before I get too far, let’s let’s let’s go back and and apply a more practical systems thinking approach to how we’re expressing this as well.
And for that, I come back to this idea of how do we build that context? And it requires you to have vendors who actually enable this stuff, who understand the bigger picture, and so much more internally from politics for people to not try to hoard their data as they’ve done over the years. But when we open it up, we can co-create so much value and serve the mission of the organization so much better as a result.
And the way that I’ve been looking at this the last couple of days, as Marius and I have been talking about this, is that there’s really four steps or three steps and four different components.
We actually go ahead and aggregate
And process. So we get the data and we get it normalized. We actually understand what’s going on. So we have to analyze it. And of course, we’re getting some assistance because the AI layer is providing us some understanding and building a more richer context from all these different sources that it’s building in. So we can understand it more easily. And then we can finally act upon it. And so I think that’s a really interesting model to look at in terms of what you need to do in adopting a systems thinking and architect approach.
as a community manager and what you can do for the whole of the company. And of course, if you’re able to lead this charge, then you have a h much higher value inside the organization. And of course your political capital will rise as well.
Marius Ciortea (25:57) Yeah, let me give a practical example on something every community manager can do in the role of what I call the liaison of the community manager. So we get a question in the community and it is about one of our products and it’s technical, right? And
We don’t often know who the right person to answer is. So what I’ve done, or what any community manager here can do is says, at your company, go and find the people that feel like they or you know are engaging in the community, right? Set up a call with them and make sure you record that call on Zoom, but ask specific questions of their expertise. So you ask them, hey.
What’s your name? What’s your job title? What is your expertise in the product? Where do you usually engage when it comes to questioning the product? What type of questions do you like answering and you you know you have the right answers? Which some of the questions is that you don’t have expertise and you don’t ask ask the question. So before you have the interviews, think about the questions you want to ask. You have it then all recorded, right? That is the aggregation part. You do this with 10 different people within your company, you aggregate.
Now let’s go into the process part. You give the transcripts to the LLM and ask the LLM categorize these 10 individuals, based, analyze their skills and understanding of the product and where they can help in the community, and create a knowledge directory for each of them and expertise. So we are processing. So first thing we did 10 interviews, we took the transcripts, now we’re processing it with with AI.
And now where we’re gonna make this actionable, a
Chris Heuer (27:45) Yeah.
Marius Ciortea (27:45) question comes in in a daily brief. The AI has the data source of the 10 employees, understands where the strengths and weaknesses are, and suggests to me, like, hey, this question may be best answered by Chris, because Chris is an expertise on this particular topic, as he mentioned. And now you can contact Chris and say, Chris, I have somebody in the community that has a question that is right up your alley. Or you can even have the AI.
send an automatic message to that individual and say there’s some question in the community that you can your expertise is uniquely qualified to answer. So this is a good example of like, you know, aggregating. We made we did the interviews, just us talking to them, processing, we took the the interview and had a process by AI. And now we have the context and now it’s actionable because the AI actually and you know who to go to based on that information we just did.
And that is something that I’ve built in for us at Vanilla when questions coming in. because I did not know anybody what everybody who they’re who’s expert, who’s not. So it helped
Chris Heuer (28:52) Yeah.
Marius Ciortea (28:53) me understand, but it also created this I call it like the expertise knowledge graph. To sound very fancy.
Chris Heuer (29:01) Yeah.
Well now and that you’ve got me remembering Blue Pages again. And and as we start thinking about who has the expertise, of course, IBM did this thing where they built their own kind of internal Facebook as a social network. And that’s where individuals were able to aggregate and share all the stuff that they were talking about and working on. And so that created the signals for us to be able to understand who was the experts.
unless you have access to those profiles and that expertise and everything’s actually, that’s a really important distinction. I use the word access.
So we’re not talking about a data migration project. We’re not talking about a data lake here. What we’re talking about is enabling access to the data.
So that we can actually deliver on the job that we’ve been hired to do. So there’s a really easy case to make for it in some organizations, easier than others, certainly because of the politics and some of the other BS. But we’re not talking about you having access to a golden record. We’re not talking about you being able to write or overwrite what the CM CSM record was. We’re talking about you being able to aggregate this by accessing it. And that decreases some of the security risks.
Marius Ciortea (30:08) Hundred percent. But but I I again I could what I just give an example, you don’t require any of this and I think should be like the step one. all you need you you don’t even even if you know, I don’t know how you would not have the capabilities of recording a Zoom call. I think that’s we all have it there. But you can send the person an email and ask the three, four questions, and then take those as data points and create your context layer.
Chris Heuer (30:33) And of course, just as a reminder, I’m sure you guys think of this, but as you’re doing these things, you must tell them that you’re recording them. And if you can explain what your use is and your intended use, all the better for it. So people can feel
Marius Ciortea (30:44) Of course. Yeah.
Chris Heuer (30:45) comfortable about that. Because it is the relationship. And at the end of the day, as we talked about on the last episode with what is the cost of AI slop, and that’s trust. And so trust is a really core component to all this stuff. So we we do that sort of disclosure, but that’s also why.
They trust us with all this information when they’re submitting the question to the community, when they’re talking to a CSM and they’re giving us this information with the expectation that the company is receiving it and should be able to respond. And of course, these systems are siloed. And if I call up, like I recently had a problem with AT&T that I’ll be writing about because it was so ridiculous how bad their systems are for customer support. I’m calling as if I’m calling for the first time.
And it’s literally the seventh time I’ve tried to resolve an issue. And I’m now nine hours in trying to get them to remove a charge that shouldn’t be sent to me. You know, I I can’t tell you the amount of deep negative emotion that I have for this company as a result of that experience. And right now, the majority of systems across the world are designed to deliver that kind of experience, not the kind that we’re imagining that’s able to be delivered with the systems we’re describing here.
I get a lot of agita when I think about this sort of thing. what I want to go back to the pipeline because you were starting talking about how you actually act on it. and once you’ve actually got the context established and you’re able to have that there, there’s three ways that you end up acting on it. the first one is the one you were talking about with the brief, it’s being prepared.
of course, in the sales world now, this is one of the key systems that CRMs are are spitting out. Is that before you go into the client, instead of you having to go and research and find all this stuff and spend three hours doing it, you can actually have a report that’s prepared by the AI sent to you right before, or maybe the night before you have that sales meeting. same thing can happen with community, where we’re able to just pull that up on the side, and before we reply to somebody,
not just see that they’re so and so with an issue and they’ve got a new user badge, but in fact that this new user is from that $10 million ARR account, that they have interacted with the CSM, that they have these given products, that they actually have talked about these issues before. and by the way, we looked at their LinkedIn profile and they previously worked at these companies and everything else. Now we have a context in which we can open the relationship more, which will say, hey, I used to work with IBM. That’s cool. did you know so and so?
Now we’re able to move into the relationship development area of this whole thing that makes it really, really exciting. When the AI is able to prompt us with this information, it can go even further.
Marius Ciortea (33:29) And again, I I think that we all have done it in some ways beforehand. It’s just that now we need to leverage what we know and
have AI able to s help us scale it. Like I remember the social selling days, Chris, right? It’s like before
Chris Heuer (33:46) Mm.
Marius Ciortea (33:46) you’re making a sales call, go to LinkedIn, you know, learn about their past, where they have worked, look at their latest posts. is there something personal in the post that you can pick up on that you can use in your sales call to create a context of like understanding and and
something mutual, like maybe we are I came from the same university, we attended the same university, that was a big one, right? maybe we live in the same state. maybe we vacationed in the same spot, right?
And I think AIs can potentially do this much better and faster for us so that we don’t have to again to spend like twenty minutes on LinkedIn to go through all this stuff. We’re getting a nice report based on the things that we ask it to provide for us in a summary.
Chris Heuer (34:38) Yeah. And and then of course the other thing that we can do out of this is we can actually have it draft that initial reply to a query or an email and actually
Marius Ciortea (34:45) Yeah. Exactly.
Chris Heuer (34:47) act on it to the point of actually getting us prepared such that we can look at here’s what we suggest And of course, this is where we get into potential slop. Don’t just hit the easy button and let that send for you if you can. Look at it, look at that message, edit it, put it in your own voice, make sure that it’s not slop, that it isn’t bloated.
That it isn’t off using wrong phrasing and that it’s not wrong technically before you hit that send button. It’s really important that we do that.
I I want to point out that like none of this stuff is an information failure, right? We have all this data about the four people that are out there that is actually that one human being, right? The four people in the data that is the one human being sitting at home on the other side of the phone or on the other side of the screen. the data’s all there. It’s just a design and permission failure that we haven’t designed the system in such a way that we can do it. And
Look, there’s some good reason for some of that. I mentioned the idea of data migration, right? We have to keep a golden record. We have to make sure that somebody can’t screw up the data that we have. But that’s why the MCP server makes it so interesting and why the AI makes this more affordable. because the MCP server makes it accessible to us so that we can aggregate and process it, so that we can develop a better understanding, so that we can act with better intentionality and to produce better outcomes.
And of course, the AI is what makes it affordable in this case, because without these systems, without that AI being able to do that high-level processing really fast and scanning all of it and providing it back to us quickly, we’re not able to do it. And this is what I’ve long talked about as the AI dividend. That if we’re getting this time back where you don’t have to spend 30 minutes to prepare an answer to one person, wow, how many people can you actually respond to and build relationships with in that same 30 minutes?
You know, I don’t know what the number is for you, it’s gonna be different for everyone, but it’s a multiple, not a singular.
Marius Ciortea (36:43) Right, let me let me do this closing thing, just repeating on the on the points that we had earlier, and the exercise I think community managers should do. Go on the whiteboard and start with the aggregation piece. look at all the data that you can aggregate. Do not start with the tool. So don’t put in there like Salesforce CRM and data bricks or something like that.
start with like the type of data you want to aggregate and you think you have. Then underneath that you see where the data lives and if it’s possible for you to aggregate it or not. Once you kind of have made that first list, circle all the pieces that you do have access to. And then the pieces you don’t have access to put it to the side and think about long term how can I get access to it. So you have two buckets.
First bucket thinks I have access today. Second bucket thinks to have access tomorrow. Right. Start with the ones that you have access today. Go into the processing, right? Work with your LLM, how to process the data. And then s once you have that processing completed, make it actionable, right? So think about
Chris Heuer (37:56) Awesome.
Marius Ciortea (37:57) how do I make that stuff actionable? I think doing those three simple steps, especially the first one.
would make your community management so much better. and you will discover a lot of things that you may have not even thought about it just putting it on the whiteboard and and looking at it.
Chris Heuer (38:13) So one of the things that I’ve long been accused of accurately is that I often try to boil the ocean. And so looking at this whole thing, I’m imagining this future where we’re able to pull all those four people together and look at it as one human being instead of four digital representations from different records and channels and the rest of that stuff. But the reality is that that’s something we’re aiming to for a future and ideal state. I really love the way you break this down. You have access. What do you have access to?
What can you get from people who you know and trust who have
Marius Ciortea (38:44) Mm-hmm.
Chris Heuer (38:44) the ability to provide this access to? And then we can worry about the more complex stuff and the other systems over time. If you don’t have full access to the to the customer record or the customer success records, that’s okay. Let’s go with what you can get out of the CRM data and their order purchase history or wherever that might be. And eventually over time, as you start showing the value to more people, that’s when it’s going to start opening up and you’re going to earn the access to some of these higher tier systems.
And be able to do more with it.
Marius Ciortea (39:13) I hate to give like one more example because we’re talking this
Chris Heuer (39:15) Please.
Marius Ciortea (39:15) time, but it’s just like one thing that changed the quality of suggested responses in the community was simply telling the LLM, here’s where the documentation lives.
Chris Heuer (39:27) Mm.
Marius Ciortea (39:28) That’s super simple. Every company probably has a documentation area, docs.something or other, right? And I just told the AI, here’s where the here’s where the documentation is. So every time now a question comes in, the AI goes through, looks through the documentation if there’s maybe an answer in the documentation. And often there is. And it tells me then, hey, that problem of the customer is solved in this article in this documentation. I suggest replying with, have you checked out the documentation page with this link here? Right?
And then I go and first look at as a human, like, is that a good answer? Yes or no? Maybe I change it, Then I press reply. And now we we have something I could have never done before. You think I can remember our company’s documentations inside out, especially
Chris Heuer (40:12) No
Marius Ciortea (40:13) different version? Nobody can. AI did it for me. So this is something super, super small and simple of of like providing where knowledge already exists and you just need to weave it into your processes. Aggregation.
Here’s the website processing, AI knows about it, contextual question comes in, here’s the answer.
Chris Heuer (40:35) Yep. And of course, this is why we also suggested, and we’ll continue to talk about building your own personal intelligence system as a community manager for yourself, of all the things you’ve written, all the things you’ve read, all these different processes and responses you’ve written, and having your own record so that the AI can be richer in serving you in this way. But again, it’s all about serving us to be more human, not to give up our agency or to be less human at the end of the day. And keeping that mindset is really important. But we’re out of time.
where I really love where where we ended with this, Marius, is moving into more of this practical guidance. And so maybe that’s what we do next week as we get more onto some of the specific tools and processes that an individual uses and you can share some more of, you know, what you’ve been doing and and what other people in the community are doing. And I’m gonna try to get a couple of interviews with a couple of people and see if we can talk to some other people about what they’re doing.
Marius Ciortea (41:25) And same you. I mean, you’ve done
a a ton of work with Claude Like you tell me all those things that you’re doing in Claude and new features that exist. So I’m curious to hear what are some of the features you use in Claude most often. so people to know I I’m more on the ChatGPT side simply because my company is kind enough to have a or supply me with an enterprise version of ChatGPT. So I I start tend to specialize in chat chat GPT as the tool of choice.
so I would love to hear on the Claude side what are the some of the advancements that made you more productive in your day to day job.
yeah, that would be a great conversation for the next time. It’s like what what AI should you use? Should you be using all of them? should you be using by case by case or should you just like
take the time and really be mastering one of the tools versus being master of none.
Chris Heuer (42:17) Well, that that is fantastic. But I’ll tell you one thing right now. If they only give you Claude, use Claude. If they only give you GPT, use GPT. Don’t worry about the others.
Marius Ciortea (42:28) let’s
hold on this one. I think this is the conversation for next time. So this is the teaser, everyone. but we probably have to call the quits
Chris Heuer (42:31) Awesome. Awesome. Awesome.
Marius Ciortea (42:36) here.
Chris Heuer (42:37) Yes. Well, again, thank you, Marius, for joining me and and really excited. Thank you everyone else for joining us here today. Please, if you are interested in being guests on the show, you can go and reach out to us through the website, through the contact form. if you are interested in having us talk about something, please send those suggestions in just in the comments or any of the social channels, spending a lot of time in LinkedIn really more than anything else out there.
And just really so much thankful for everyone joining us.
Marius Ciortea (43:03) And any tips and tricks?
Like if you guys are are doing some of this aggregation processing, context taking action, let us know. I’d be interesting to know how you guys leveraging that framework.
Chris Heuer (43:15) And again, I just have to take a moment to say thanks to Higher Logic for making this all possible. It’s a lot of work each week to get here and spend this 30 to 45 minutes with you. And just grateful to be able to have the time to research it, to work on these ideas and develop them further for the benefit of all of community management. And that’s really the the unique thing here. You know, Higher Logic is a sponsor, they’re not controlling the content. I’m running this in terms of putting together the prep materials and working with Marius as a collaborator in the open.
And again, that’s one of the great things that Marius really understood that while working for the generalities of community management itself, that is the rising tide that will lift all boats of all community management platforms, not just his. And hopefully we’ll have some of the other community management leaders from different platforms on the show in future episodes. And if you’re listening to this and hear that, please make sure you reach out and tell us you’re interested, because we’d love to talk to you as well and get your take on some of this and continue to build the corpus of knowledge.
that makes community managers get back to the work that they love to build those relationships, to be the human relationships and to give the busy work to their assistant to the AI. So thank you very much, Marius. This is Chris Heuer for CommunityAssistant.ai. Marius, any closing thoughts?
Marius Ciortea (44:32) No, thank you so much. see you next time.
Chris Heuer (44:35) So take care. Have a great day.
Marius Ciortea (44:36) All right.
Made with Claude, not by it · CC BY-NC 4.0