Value-First Platform: AI Data Readiness - Apr 8, 2026

[00:00] [00:07] Speaker 1: Good afternoon, LinkedIn friends, Value First Nation. Welcome to another episode of Value First platform, AI data readiness here with Trisha and Aaron talking AI ROI today. How are you doing? Trisha: Doing good. Speaker 1: AI ROI, that just rolls off your tongue, doesn't it? Trisha: I know. Speaker 1: In my head I was like, how many acronyms should I just try to do as many acronyms as possible at this point? Trisha: Alphabet soup over here. Speaker 1: Yeah. I'll I'll plan better next time. Um, but yeah, well be honest, ROI not something uh I usually enjoy talking about. Um, but for the reasons that I think AI solves for, which is why I think this conversation is interesting. Um, we we talked a lot about observability last week. And I think that's what, that's the unlock, right? Like the transparency we can create now allows us to see and measure the entire story instead of just measuring the thing that has been observable because we only have SAS tools and we're not good at meetings and communication with each other. So we try to create ROI stories off of incomplete like information, context data, right? Um, so I am excited because we can solve something that I think has never really been solved before. Will it help people make decisions about AI? I'm not sure, but maybe we can help some people feel feel better about it. Um, but we were talking before we we went live, Trisha, you're doing an AI for for beginners course and the end is talking about ROI and that's the challenge is like if you're not, if your system is not observable, then like you're not going to measure it. Um, But I want to dig into whether or not like how do we help bridge that gap for people? I think. How do we help people think about it from like, it's not a bridge too far perspective. Um, and usually Aaron's pretty good at these kinds of conversations. Um, but I'm curious what you're seeing out there, why why you think, why you feel that way, Trisha, and then maybe Aaron some tips on how to have a conversation like that. Trisha: That's fantastic. I uh because I was the first to say after having talked with you guys last week, um putting in mind like what I was going to tell this team this week, it's a couple weeks from now, but like what I'm going to tell them about how to measure the ROI on their AI efforts. Um, the answer was going to be, they don't have anything connected. So and until they do, or until they do something and I'd love to be able to give them an idea of what that something could be as a starting point, they're not, they're like there's no point in them even trying. It's not a really great exercise. So this is just for a marketing team at this point. This is a marketing team that is overwhelmed with tasks and small company, they have no time, no time to do any of the strategic things that they want to do. Like every member is just painfully bogged down and overextended. And that was the reason that they opted into some AI learning because they haven't had a chance to learn anything on their own. And we're at this point just trying to automate some of the work they do. Number one, to build them some efficiencies so that they can create more content with less time. Um, do a better job of segmenting because they aren't doing any right now, really, none of any real strategic segmentation. Um, so we like what I'm hoping is that the the results of their efforts will improve because they're going to be more personalized in what they're saying. They're going to be creating content that they don't have now. Um, and they're going to be able to produce more than they're like producing now with less effort. But how do you go about Right. For sure. And they don't like they're not using any kind of time tracking software or anything. So how would you go about that? Aaron: I think as especially as it pertains to content creation, um you can do like a time to live or like time to first draft or um some kind of time boxed metric where um you're showing how how much more quickly they can create content. Um, and then like a total content volume over um a period of time compared to the previous period, I think would give you some good numbers. Um, as far as some of the other pieces, what are they using for AI? Is it like Claude or uh chat GPT or Trisha: They're going to be using Breeze and then free free chat GPT free Claude. Okay, got it. Um, one thing that I have been kind of thinking thinking about doing uh because I have something similar on my in my setup where I have like a session start hook in just an MCP for my operating system. So it has a session start and a session stop and when it does the session start it basically creates a um a data point. Mine goes to Superbase. So any kind of data database, you can put stuff in. Um, so when they start the conversation, um it should automatically kick off the session start, store a timestamp and the client that they're working on or the the thing that they're working on, um and then have a a session end where it marks that time and kind of saves it to your database. So now you have at least some conversation data where you can say like, this is how long I talked to the AI about this thing. Um, and we were working on client X. And so, um, now we can start to say like, well, did client X, you know, buy more business from you or or whatever that um that value is, um that you're adding. So if it's in the sales process or um, you know, this is a marketing team though. So I guess it would be like content specific. Uh oh, and then the segments I think could could be um just success rates, uh just kind of AB test, like this bucket got non personalized or non AI personalized um materials versus the AI um and then have those metrics to compare. Yeah, so I think a lot of it is not going to be strictly to a dollar amount. Um, until some of the like downstream effects happen where like you start to see leads coming in from that content or um, yeah, especially with a marketing team, I feel like it's you're going to have some leading indicators like the content velocity and things like that. And then your lagging indicators are going to be, you know, new business from AI generated materials. Trisha: I like those ideas. If they're using Breeze, are you still able to do like send a timestamp to anywhere? Speaker 1: Chris, can you plug in an MCP to Breeze? Speaker 1: You could, but let's not get crazy. Like how do we measure things before we had software everywhere? Trisha: Fair. Aaron: Fair. Um, yeah, I think especially with a marketing team, velocity, volume, um time spent on like you could identify like a bucket of your most valuable um things that marketers contribute to. So whether that's like new um like lead gen ideas or or whatever, like figure out what activities lead to the most value. Um, and then kind of track against that to where your your north star is like people doing more of this thing. So if you can take these other things off their plate and say, well, now they're spending 30% more time on those valuable tasks. Um, that can also be really impactful. Speaker 1: Right. And it's like very quickly the thing that we're not supposed to do or that we've been told, like you have to get to like an output understanding to make this make sense. But are you creating the right outputs and saving the time in the right spots on the right things? Where I was going with it, before we had software everywhere, we had to talk to each other. Right? Like and this is like software has almost created like if we're not measuring it with software, we're doing it wrong. So nobody puts qualitative measurement anywhere near where it should be. Like the first thing every team should do, if you're trying to get clear on your data and you've never asked your sales team what happened to win a deal, like that's where you start and then it tells you exactly what you need to measure. And so the how, it's like there's this pendulum swing where imagine you just didn't have software anymore. But you did have this amazing way to record every conversation you have about how you feel about the process and the impact that AI is having and the impact that teammates are having on you because now they can use AI. But we have that, right? Like and if we just let AI say at the end of the conversation, like are are people feeling like this is a valuable thing in their life right now based on the words that they're using and the tone that they have, right? Because I think um just like the trap has always been, like this is where you quickly go because every software we use is designed to help us do that already and now it's like AI is getting layered in. But can we actually do this, right? Like is it doing anything valuable? Like helping me send a thousand emails instead of 100 emails is um using AI a lot for sure. But it's probably not creating any value. And not to mention it it leaves no space for the immeasurable things. Like, are you having an amazing time? Like doing your job? Now, like most people that I've been AI native unlocked, the answer is like unequivocally yes. Sometimes they'll just tell you, I've never I've never had this much fun doing my job before. And from a leadership perspective, I want them to like it doesn't mean it's going to happen, but I want them to be forced to hear those things. it needs it's a huge part of the return you should appreciate and expect because on the other side, if you're using AI and it and you hate it, something has gone terribly wrong, right? And it's not AI's fault usually. So, I think if if like if you have 10 minutes to tell a team like what they can do, it is start talking to each other and recorded lines, like about how AI is, like what they want it to do. And I think to the Breeze expectation, like you could have a chat with Breeze and say, this is what I'm doing right now. And then at the end of the chat, say, add this note to the deal or to the to the contact record or to a lot of different places now, maybe in campaigns like if we're over in Marketing Studio, right? It's like that is underrated in almost every AI user. Like at the end of your session, just just what do we do? Like what happened here? Store it, right? Um, so that's where I would go. In terms of like what those conversations should be about, I think that's where a framework is makes it a lot easier for AI to help you like measure those conversations. So instead of, you know, all the usage metrics that we think about the time saved per task, um, it's more about it's so it's it's a lot of quality of life stuff. Like honestly. And it's like interesting because we talk about quality of life upgrades as if they're like, let's get to them when we can, they're nice to have. But that's such a huge part of AI, I think, and the value that it drives. So I'd love to go through these three dimensions because I think this is as close as we get to like the quantification of it without uh sacrificing the um the qualitative metrics at the same time. Uh, does that resonate at all, Trisha? Can you, can you work with that? Trisha: Can we see the, can we see what the three dimensions are before we dive into each one? Speaker 1: Sure. Trisha: Cool. So we got revenue. Speaker 1: Revenue influenced, influence protected, and capability built. Okay. Right? Uh, the hardest to measure and the most valuable. Let's go in order. Okay. So when a deal closes, the system knows which AI agents, and I would argue human humans as well, uh touch the engagement for how long and in what capacity. This connects AI activity to actual revenue, not through attribution models, but through conversational traceability. Like I think this is the like we should design every measurement framework right now to do this. Conversational traceability. And if you can get there, a lot of traditional dysfunction goes away. Trisha: I want to ask back up for a second. You said um something I really liked, which was like I think where you were going with it is if you can kind of train the people using AI at the end of a session with it to sort of like have ask a reflection question, how did this go? Tell me what you think about what we developed here and then store it. Um, for teams that are primarily using Breeze, is that is that possible? Or do you have to be using like a Claude or something like that that it can actually store those? Speaker 1: I mean, if Breeze is good at anything, it's adding notes to records. Uh, like for sure. Like the more unstructured you can get help from Breeze in, it starts to like work like really well. It's when we try to set all these boundaries and and create structure out of unstructure when it starts to be less good for me. Um, in my experience at least. So that's where it is just and I don't know if you can do Trisha: because marketing teams don't tend to touch people that much. They're like working on campaigns. Um, so they're working on assets a lot of the times. Speaker 1: Right. So like now the campaign, like campaign and marketing events are like real records in HubSpot now, right? So I've never suggested this before right now. But if you can put a note on those records, that's probably the place where you want to do it. Okay. Because I think the key is like, can we look back at after a week at the end of the campaign at the at the month, quarter, whatever and just have that information somewhere, so we have a chance like to consider it. If we don't, then all we have is the structured structured data, right? Um, so I think where there's a well, there's a way there and that's where that that might be the goal of that 10 minutes. Like if you're going to use Breeze to get help with copywriting or any of the multitude multitude of things you could use it for, just at the end, uh, you know, and there might be a way to template like I know there's a way to template prompts. I just don't with with custom assistance or Breeze assistant. I'm just not sure if you can do them at the end of the conversation. But you could use the custom, you could create a custom assistant that gets annoying towards the end of the conversation. Like, are you ready for me to make a note? Are you ready for me to make a note? Like, I mean, Claude does that to me all the time and I know he's like, must be using a lot of tokens this session. Like it's like 6:00 p.m. It's like, have a good night. Like, oh, we're not we're not done yet, man. Like good try. Um, yeah, I think that's a good habit for any any user using AI for anything. Like it's the quickest path to documentation that you can actually start to just make sense, just try to make sense of what's happening, I think in most cases. Um, so I'd say in that, on that note, like for revenue influence and revenue protected, if you don't have your revenue data together, this that might be tough. But that last bucket, capability build. Marketing doesn't need anybody else to like show some value in that in that bucket, right? Uh, but maybe we should assume or we can make an assumption in this case that the revenue data is good enough. Um curious uh Aaron for your take, like if these things on the screen, like let's going through what this looks like. Aaron: Mhm. Speaker 1: Um these things make sense. Aaron: For sure. And I I love the just adding a note. Like it's such a like elegant solution to this with Breeze because I was having the same thought of like how it would be it would be even nicer if it could be like an object that you could report on. But that's leaning more towards the quality uh quantitative data, um that we don't necessarily want to lean on as heavily. But um but yeah, I think knowing knowing what you're spending your AI efforts on, like um which types of tasks, like I think what I would probably do in this case with the marketing team would be break down the the daily tasks into buckets where you have some that are value creating, some that are just time sucks and you know, everything is kind of focused on this. And then um kind of use the distribution across those tasks as a baseline. And then as you add in AI, um not only are you going to see where the tasks where you're doing less things, um, but where you're adding new tasks, where that capability build thing comes in. So for a marketing team specifically, maybe they're doing a lot less um, you know, campaign personalization uh on like a per contact basis. Um, but now they have the ability to create these lead magnets that are like calculators and like little things that you can just plug in on your website and um giving people the ability to do more complex marketing briefs where now the downstream team is more equipped to implement everything. So maybe the time to implementation goes down. Um, I think really if you start with how your people are spending their day as a baseline, I I think the the ROI becomes kind of an emergent property out of that where like you're starting to see shifts that um, you know, you otherwise wouldn't focus on if they weren't kind of broken down into these like, what is valuable, what is not valuable kind of buckets. Trisha: Mhm. Yeah, and I see that with a lot of companies, especially when the teams get overwhelmed is just like I I used to recommend to different people on my team. I'm like, take a work like take a work vacation because you have these critical, important things and you can never get them done. So put your whole calendar as if you're out of office this entire week, don't take a single meeting. You're working but you're not here and get the thing done. Um, like there's nothing like being able to just create some time and space for people to like think creatively. So there's a lot of value in just build it like clawing back some time for them. For sure. Um, and this company also they're going to be getting content remix, which means that they're going to be able to produce a lot more like they've never really been able to produce video at scale. There's probably an opportunity for them to do a lot more video now. For sure. Which we know should probably do a lot for them because video is just a lot more engaging than text on a page. Right. When executed well. When executed well. Yeah. Speaker 1: Yeah. How do we prevent just AI slop land? Um, like because it that's the the risk of like focusing on I don't like the conversation about comes and outputs anymore, especially in a marketing context. Like when you got to start showing up every day or way more often doing brand new things than you used to do before, like getting to the consistency of doing something every day or multiple times a week that you weren't doing before, that's a win. Like getting teams need to see that as a win. But doing it in a way where you can observe what's happening at the same time so that you can know, oh, this one didn't hit because we did a terrible job like with it, not because video is not the right thing to be doing. Right? Like if you do five terrible jobs in a row because you don't know what you're doing with video, like that's a capability, you know, problem. Again, not a video specific problem. So, um, I'm curious Aaron, how you would help a team like think about this like from that perspective. There's a bunch of things I'm looking at on the screen right now, what you need before you measure. It's like I'm looking at them. I don't know shit. Like none of the teams have this stuff. Aaron: Right. Right. Speaker 1: Like baseline honesty is on here. I was like, uh oh. Trisha: And you were just talking about slop and I have another client who's like, yeah, we're not going to do that thing anymore because we've tried doing it a bunch of times this year and it it doesn't do any it doesn't move the needle for us. And I'm like, but the reason it's not moving the needle for you is because you spend five minutes thinking about this every time you do it. You're not you're not putting it like it's slop. Aaron: Right. Like sometimes the play is not more content. It just you know, as a general rule, sometimes it's better content. Like one of the companies that I worked with uh and we built like a an agentic content engine for them. Like I wasn't trying and and like I was very explicit about this with the team. like, I am not trying to take work away from the people who are writing. I'm trying to get you a first draft that is worth your time to humanize it and and add your own like you're still a writer. You still do content. That's still incredibly necessary. What's not necessary is hours of research and doing an outline and figuring out what the head header structure should be. Um, all of that can be offloaded and you can focus on you know, more of that like qualitative uh measures of the content itself. Um, and you know, you may want to monitor things like um, maybe team trust in outputs and customer trust in outputs. like customer sentiment around the content that you're putting out. Um, so that, you know, if you're putting out AI slop and not getting any engagement, that says more about like you said, like that's more about how you're using the tool than the tool itself. Speaker 1: Chris is thinking hard. Trisha: I know. I'm looking at Chris. at Chris. Speaker 1: Yeah. Um uh yeah, so let's think about like what they would need to do. Let's say they take you up on that offer of like just start cracking notes with with Breeze. Like we have to help them and there also has to be like a safety. This is where like the culture of the organization starts to come into play. Like, do they feel safe putting a note on a record that other people like might see and might draw in questions? Like um So culture aside, like we've got to create visibility. Um, and whether it's a so we we talked about like as you're doing the task as you're in the session. But if that doesn't work and you've already got some kind of review schedule that you're end of week Friday. Like what did you do? Like what like was the project? Like go to the campaign record. This is where just like the architecture of your data system can just help you where you can just show up on the campaign record or in Marketing Studio and immediately have the context you need to start thinking about it. And then you can have a custom assistant that's just like asking you questions about how how the thing went, right? So just opening up that conversation um so that we can start to get, you know, visibility. Uh and then taking it to the um a lot of this real like I think my my challenge right now is all of this is based on the observability like conversation that we're having where the system is doing this for you and we're trying to like translate real time. Because none of it works without the humans doing their part, right? Um, but you know who is using AI, how often and what tools. Um, but you cannot connect any of it to business outcomes. Uh the note situation, right? There's still a verbal bridge that's going to have to happen like during some meeting. And usually this is when like slides get created or the spreadsheet gets created or even the HubSpot dashboard in this case. but this is what we got done this month and you know, you put numbers on the slide but then when you're presenting, you're having like the real conversation about what the slide means. While sometimes like saying, don't worry about what's on the slide, right? Um so yeah, I think I need some help walking through these uh steps in a in a meaningful way. Yeah, I feel like some of this stuff is just so tool dependent. Like each different AI that you're using has a different mechanism of getting visibility, a different mechanism of being able to track the things that you're actually trying to do. So I feel like it's very easy to get boxed in with a certain solution and then kind of hit a wall and just stop there. Um, so yeah, I think once you get even just a baseline of who's using it, how are they using it? Um and for for what for what outcome? Like what are you trying to achieve in this conversation? Um, even if it's self-reported, honestly. Like I know that you Trisha, you said they're not tracking time or anything. But, you know, if you if your team is tracking time, you know, having them put in their time tracking, I used AI for XYZ. Um, can also kind of help and it doesn't rely on the tool. It does rely on humans remembering to track their time, which I personally am terrible at. So I'm not saying this is going to work uh across the board. But um but yeah, I think um kind of to your point earlier, Chris, about having more conversations around what the team is doing can also be helpful and having those kind of, you know, working sessions where it's like, you know, what did you do with AI with this week? What did you do? And like that cross pollination of ideas kind of um can both give you insights uh into this kind of data that we're trying to collect. Um, and at the same time kind of spur uh new uses and more uses. So uh I think that's that's a good, a good place to start if you have nothing else. Trisha: Mhm. I think also you could if there's to your point, if there's if there's a thing that's important for this team, it's like, well, how much of my week day is being spent on something strategic versus just tactical? Like I'm just checking off a list of shit I got to do before I can get to the real stuff. Um, just bringing their awareness to, you know, how much time this week did you spend on the stuff that's your highest priority or highest goal. And can you tell over time whether you're finding more time? Are you able to find more time to work on that? Like that's the kind of thing where everybody agrees that's where you should be investing your time and any any way to allow that team to have more of that time and less of the junk time, if everybody already agrees it's a win and then you can see that you are increasing that time, everybody's going to be sold on that. Speaker 1: Yeah. I do think there's still some some gaps there in terms of like especially as you're learning to use AI, you don't go faster with almost anything that you do because you're having so much back and forth than if you would have just like done it yourself the first time. You know? Yeah. And so that's where and again, culture needs to allow for this. But it does work when you let it go to this um this space of like when I can when people ask me why I do so much content and how I can fit in the the uh, you know, the clients, it's because I enjoy, you know, what I'm doing. And now I doubly enjoy it because all the stuff I used to sacrifice to do this isn't getting sacrificed, like the documentation and the project management and all these things that I would be stressed about the entire day. So I might be doing something I want to be doing and getting stuff done, but I'm stressed all day. Now, and sometimes AI offers it up for me, but like what's the energy level at the end of the day? How do you feel about the day? Right? The day the day you just had, right? If it was the I think there's a big correlation between like the more conversational your day was, the the like whether it was with AI or humans, the more likely you are to have uh like a healthier kind of energy, let's say. And like some of this is in the framing and like such a point of view perspective where how you frame it matters like so much because this stuff here like baseline honesty and conversational infrastructure. Again, you take the software out, right? This is a daily stand up, right? And this is that meeting that you can't get sales to show up for like a 30 minute training session. But they will tell you for an hour and a half how awful the last thing marketing did was or the the the last show, they will rant for a long time because it's important to them, not because other reasons. But because usually people care and they want to get better, but this ends up being the only outlet where they can just share what's on their mind and like help themselves release like of the stress of the anxiety of the day. We can capture all those moments now and like start to take action on them instead of, you know, find ourselves having that same conversation every week. So the baseline honesty is there. It just doesn't usually show up in the systems because we're not putting it there or we can't put it in put it there. Um, like I'll never forget the first time I brought a sales team onto Sales Navigator. And I had to explain to a 30-year sales vet, who was amazing at his job, like one of the best people persons I've met in my life. And he asked a question, so when I connect with somebody on LinkedIn uh and they they connect and say hello, what do I say? And I'm like, the same thing you'd say if you just knocked on their door and they opened the door, right? But because it's a completely different like environment and space, it's like, oh, I don't know what to put in this box because it's not a conversation in my mind that I'm used to. So that's where we can let the conversations happen wherever they need to happen now and then bring it into the system. Now, even in an unstructured way to where whenever we do bring it up again about how people are feeling. Again, you need the framing, you need to decide what your company cares about, what your team cares about for AI to make sense of it in a lot of cases, which is where unified customer views and value path and you know, these kinds of things help with that dramatically. Um, but I think like for teams that aren't using it at all, you have to find those moments of what are you doing already? that we just move it over here or we just activate the AI button and now all of a sudden this you don't have to change your behavior very much, you just have to put it in a certain spot, right? And I didn't think about this, but maybe I'm sure that goes into some of the like the will to make a campaign record like a real real object and associate to all the things, right? A marketing event record is now, you know, a real object where you can you can do all this stuff. Now that marketing has an infrastructure where they can do things like this inside of the system, right? What other um questions, concerns come to mind as you think about how to help either the marketing team that you're training or leaders, you know, understand how to have this ROI conversation. Trisha: I don't know how I'm going to share that question. I the takeaways that were helpful from earlier in our discussion that I felt I could really do something with is things like if they have a goal to be able to create more of a specific type of content, they're not creating it all, most companies have some sort of metrics on how much they're creating. Like they have a content calendar. They're they can like figure out how many new blog posts, social posts, whatever they're creating. So if there's a a new type of thing they're trying to do or they're trying to do more of a specific kind of thing or they're trying to do a better job of a specific kind of a thing. You know, if like by segmenting it more instead of just a blast, whatever, they don't have to like cook the whole ocean. They can go pick out the things that they think are the biggest value drivers and just put some intention around figuring out time spent and outputs and just kind of like begin to track that over time and see if things shift. So in practical terms, that like that was one takeaway from our conversation today that I think does make it achievable for them. Um, you absolutely can add notes from Breeze onto campaign records. I personally have like never used campaigns on HubSpot. I have never found them very useful. But maybe this is the reason to begin using them. And then I also looked on like landing pages and website page and there's no note thing there. However, you can add a note to a task which can be associated to the landing pages. So I still don't know what to do with that, but I that's very interesting. Yeah. No, the campaigns thing is only like a week and a half or two weeks old as far as like it becoming a real object in the CRM. Um and I think over time Breeze is going to get better and better, better and better at seeing like even comments. Like so the landing pages, website pages, a lot of the marketing area has these commentable places to where you can tag each other and have a conversation. Like it always gets to like, can we create visibility over the conversations that we're having first. And then, again, work with AI to start figuring out what reporting might look like based on the conversations that are being had, right? Um, so uh last thing I want to show like so efficiency, like this conversation let's see these two because it don't like if you're going to talk to Breeze, I mean, I would argue like just talk and maybe you do some of this stuff in the beginning and it'll show you start to show you like why it's it's probably not driving the value, right? But these two things over here. So like from a marketing perspective, requirements quality, eliminated a category of build failures. The inability of teams to communicate with each other to like get the output that you want, right? Like getting to like versions 11, 12, 13 of like a one page PDF, right? That's not okay. But we've got brand guidelines and we've got subject matter experts that need things spelled a certain way, no matter how marketing says it looks. Like got to have all those zeros in there, it's significant digits. Sorry, we don't just get to cut at them off because it doesn't look good. Um, this is the kind of thing where AI uh in both