The AI Accountability Gap in B2B Marketing, with Holly Enneking

Holly Enneking: 92% of the people that we surveyed said that they were using AI more often to generate content than they had previously. Like, that feels right. But what I thought was really interesting was that the C suite level of the audience that we interviewed, they felt confident in the checks that they had in place. Only 44% of the marketers actually felt good about it. And so you've got this disparity between the people who are making the mandate and the ones who are doing the work of, like, actually, no. Like, I think I need more checkpoints in order to help ensure the quality as we're moving through this.

Lindsey Groper: Hi, and welcome back to SAS Half Full, the only show serving B2B marketers. I'm Lindsay Groper, EVP at Pan Blast, and I will be both your host and bartender. Today I had an awesome conversation with Holly Eneking, the VP of Marketing at, uh, Markup AI. And I know that you all are creating a lot of content using AI. We all are. It makes us more efficient. We can crank out more content and. But the issue becomes who's actually responsible for ensuring the accuracy of the final output. Well, as it turns out, everyone's responsible, meaning really no one's responsible. So Holly's going to dive into a recent report from Markup AI that addresses this accountability gap and talks about AI content governance and this idea of AI content guardian agents. So if you'd care to grab a drink and join me as I speak with Hollywood. Hey, Holly, welcome to Sass Half Full.

Holly Enneking: Thank you so much. Excited to be here.

Lindsey Groper: I'm excited to have you. You were actually sitting in the same market, which literally never happens unless I'm talking to a Pan Blast team member, which is like, once or twice a year on the show. So glad to be talking to another indie native.

Holly Enneking: Yes, Indy is the place to be. We got to get more. More folks around out here. It's a great place. Fully agree.

Lindsey Groper: If you're listening and have never been here, what are you waiting for? Although maybe not between like, December and February.

Holly Enneking: Start like April, May, June, Then we're really like, peak.

Lindsey Groper: Exactly. And I understand that we were able to get you a cocktail kit. What are you drinking today?

Holly Enneking: So I have a lovely cranberry margarita today, which I'm very excited about.

Lindsey Groper: For those of you that, you know, can't see, uh, she has it in a really pretty glass and has it all mixed up. I am not as fancy because I just grabbed what I had closest, uh, to me, which is a vodka seltzer. It is the Fresca version, which I love because I was a vodka Fresca person person before they decided to do it for me and put it in a can. Cheers.

Holly Enneking: Cheers.

Lindsey Groper: Well, Holly, before. Oh, that is really good. It's a great flavor that makes me feel like summer.

Holly Enneking: Oh, that's what we need.

Lindsey Groper: Before we dive into our topic, Holly, I do want to give our listeners a bit of background on who you are, how you got into the world of B2B SaaS AI marketing, and. And certainly also want to understand what does Markup AI do? Why does the company exist? So if you could hit us with a little bit of that would be

Holly Enneking: a great place to start. Yeah, happy to. So I have been in and around marketing my entire career. Now. I was saying 15 plus years. I think I'm getting close to having to say 20, which I don't super love. But, you know, here we are. Have always been on the brand and digital side of things, which is really sort of my passion. And I've gotten to work at a lot of different types and stages of companies, so in a larger, uh, scale SaaS company that was on a global scale, a hundred million revenue, a services business in the salesforce ecosystem, a small startup in the HR executive hiring space, and now at Markup. And I especially have loved being a member of the marketing community in Indianapolis in particular. I'm also a co founder of our indie marketers networking group here in Indy that we organize a lot of events every year. And I've just been really lucky in my career to have met a lot of really great, fascinating people that have just pushed me along the way in this career and have been lucky to land where I am now at Markup.

Lindsey Groper: Awesome. All right, and give us a little bit on Markup AI. Why do y' all exist? Who do you help?

Holly Enneking: Yeah, so Markup AI is incredible because we sit in this really interesting intersection of two realities that marketers are living in all the time, which is that AI has made it so, so easy to scale up content, but also that content is not always good or trustworthy. And this gap between that content generation and publication that most teams are trying to fill with manual review is just simply not scalable. And so what we do is try and close that gap. So with our Content Guardian agents, we sit inside the workflows that content teams are already in in order to scan, score, and rewrite their content against their brand voice, their terminology, their accuracy standards, your compliance rules, so that you feel confident about your content before anything goes live. And what's great is that it's not another tool that you have to like, copy and paste your content into in order to check it. It's built into the places where you're already working on your content, like Google Docs. So that means for marketers that you don't have to choose between moving fast and getting it right. You really are able to have the speed that you want with AI without having to sacrifice the quality of the content.

Lindsey Groper: I can relate certainly to this. And, and I feel like as an audience success marketers, they're all like, shaking their head as they're listening in their cars or wherever they are, like, yeah, huh. I feel that. And I was, uh, really interested to talk to you because you ran on some data. So I. While I love points of view and I love opinions, what I really like is when we have some new data to sort of back up, use cases, back up, what we're saying. Because one of the interesting things is you have this gap and your data will reveal all of this if everyone's using more AI. But also they admit they don't trust the output. But there's this accountability layer of, uh, okay, well, if everyone's responsible, then no one's responsible. So that's, I think, like, where markup AI really solves this need here. So talk to us about this new data report. Where did the idea come about to create some new data points for this market? Because I think you already had a lot. But just talk to me about where this originated.

Holly Enneking: Yeah, so where this came from was that we had been hearing a lot in the discovery work that we had done with marketers, that everyone has this mandate to adopt AI, uh, and to scale with it, but no one really has the rule book for, like, what do you do to trust the quality, to check the quality, to ensure that the things that you're doing are still at the level that you want them to be. And so we really wanted to help quantify that gap and especially between, like, the mandate at the executive level and then the implications for that at the actual, like, practitioner level for the marketers who are the ones who have to go take that, like, go use AI and put it into practice. And so we were really fascinated by some of the data that we saw. I mean, I think it's probably no surprise that one of the things that came out of our survey was that, uh, 92% of the people that we surveyed said that they were using AI to generate, like, using it more often to generate content than they had previously. Like, that feels right. But what I thought was really interesting was that the C suite level of the audience that we interviewed, they felt confident in the checks that they had in place, but then only 44% of the marketers actually felt good about it. And so you've got this disparity between the people who are making the mandate and the ones who are doing the work of like, actually no, like, I think I need more checkpoints in order to help ensure the quality as we're moving through this.

Lindsey Groper: Yeah, that's not super surprising because we're all using content or AI tools to create all different types of content from internal to external. Are you seeing any specific areas where it's like where you're seeing really big misses? Um, is that in more ad copy? Is that in blog content or is it just prevalent across.

Holly Enneking: I think it's really prevalent across the board. I think especially in addition to doing the survey, we have done a lot of more in depth conversations with marketers, about 30 people that we've spoken to. And it really feels like it's in the amplification of the content. So you've got your big core content piece and then you want to activate it across all of these channels. And so AI is a great place to help make that happen. But then suddenly when you have emails and blogs and social copy and you're activating your executive team to promote it and you're doing all of these things, suddenly you've used AI to generate it. But are the data points consistent across all of them? Are you using your voice correctly across all of them? Is what you're saying in the report accurately translated into what you're putting LinkedIn post? And that's where you really get into a bottleneck of just trying to make sure that everything is consistent and accurate and it can manifest in different ways across all of those things. Like, I think that's one of the biggest challenges is that the content creation is so prevalent and so deeply ingrained in so many marketing teams go to market strategy that it's just something you have to get right. And like falling down at one place just cascades into all of the others.

Lindsey Groper: Yeah, I saw that play out in real life. There was a, like a fairly recognizable brand that was promoting an event that they were hosting in London and the city name of London was spelled wrong. There was a missing O. You know, I wasn't like, oh, what dumbasses. I was like, like somebody needs to know. Because my assumption was that that was some core campaign and that they broke it all apart. The core campaign was approved. But then whatever tool they, you know, leverage to slice it and dice it. I actually, like, teed up a. Teed up a LinkedIn message, like, hey, I'm, um, just one of like. But I was like, it's not my job. Like, someone else will tell them, certainly. But it was killing me because my, you know, first reaction is like, how did this get through so many approvals? But it didn't. That's the thing is, by the time it morphed into whatever it is, it didn't. Are there any, like, oh, gosh, things that you have seen where AI just royally, like, effed up content?

Holly Enneking: Oh, my goodness, the things that AI can hallucinate sometimes is just wild to me. I was doing in a previous role before coming into markup, was working with a healthcare, uh, system and doing, like, creating some content around. I think it was like cervical cancer and like, raising awareness for screenings and things like that. And so I was trying to do some research and trying to find some studies to help back up some of the claims. And it was just hallucinating reports that never existed. And like, if I hadn't taken the step of being like, okay, show me these sources and let me, like, click through and like, it made up data and then made up the sources that looked completely legitimate, like, its ability to sort of, like, look good enough, but, like, the second you start to dig deeper, really fall apart. It can be so easy to be duped by something like that, especially if you're under pressure to move really quickly and just get things out the door, like the opportunity is there to just legitimately miss something. It's just, just from the space of, like, all marketers trying to do more with less. Like that, uh, reality is right there in front of all of us.

Lindsey Groper: Yeah, we did internal survey and it showed that one in three people take the first AI answer or AI overview at face value and move on. Because they're just so tired of verifying. They're like, I don't even know where to go. So I'm just going to use this, which is terrifying. All marketers are being asked to do exactly what you just said, which is use AI to do more and more and more and more. An obvious solution would be to use markup. But if you're a marketer who does not have a solution like this in house, what are some of the immediate, uh, guardrails that they need to be doing, like, uh, putting into place like this quarter?

Holly Enneking: I mean, the most immediate things that any marketing can be doing is just get your sort of compliance house in order, what are the controls that you want to have in place in order to make sure that you've got that checklist, even if it's a manual review to start, like that maturity scale of how you're using AI is a great place to sort of benchmark yourself against, but you got to start somewhere. And the basics of what are the rules of play for your tone of voice and your brand? What are the words you're using and not using? What are the things that have to be true for you in order for something to go live? Who are the people internally that you need to get their sign off? If you're in a regulated industry, like who are the legal teams, who are the compliance officers that you need? Who are the subject matter experts? If you're in, uh, you know, a really knowledge based space that needs some of that additional context to be brought in, like figure out what those things are, that baseline of what has to be true, and then making sure you follow it. And don't short sell that piece of the timeline just because you've sped up that content creation. At the end of the day, we still have to put out content that like, we feel good about as marketers, that uh, we feel good putting our brand name against, like making sure that you're really taking that piece seriously. And yes, the first 80% may be sped up, but that 20% is going to be the difference between, you know, good content and great content. And the great content is what's going to drive engagement. And so just having some clarity internally and getting people enabled around it is the best thing you can do.

Lindsey Groper: That's really good actionable advice. What did you find in your research for this report? Specifically where does ownership lie currently?

Holly Enneking: Yeah, that was one of the really interesting things we saw in this report, is that there is not a lot of clarity around who owns it. Some marketers feel like they own it. Sometimes it's the IT teams that feel like they own it. It can be really disparate and in some cases what it feels like is that no one owns it. More often than not, the internal perception is that it should be marketing owned. But in, you know, teams where there's a big engineering IT presence, like there can be tension there between who really is owning it and enforcing it. And so that's really a question that, you know, if someone were to walk away and think about, what does this mean for me? I think a great question you can ask is like, who does own it? Am I responsible for owning it how do I have a conversation either within my marketing team or a level up within my team around what that looks like and how do I have good partners internally for helping enforce it? Because without having those conversations of clarity, you're going to continue to either be responsible for something you didn't know you were responsible for, which is never a good feeling, or, or there's an opportunity for you to take on ownership of that and be able to lead the charge in a way that can feel really empowering and maybe not something that the marketing team would have access to if you didn't speak up and ask the question of like, who should be owning this and how do we drive it forward?

Lindsey Groper: Yeah. An old colleague of mine, Chris Lucas, if you're listening, he would always say if there are two systems, there are no systems. It's so true. It definitely applies in this instance. Do you feel like there will be an emerging roles that have AI content governance as part of the job description?

Holly Enneking: I think so. I think especially as the role like AI is really changing the way that marketing teams work. I know that for myself, even as I've used AI, is moving away from thinking myself purely as content creator and really trying to put on an editor hat like, how am I bringing my expertise as far as like quality and taste and our point of view and perspective into what the content includes. And some of that becomes like, what are my perspective on the tools that we're using and how they're being applied across the board? One of the things that we saw in our survey was that almost 70% of marketers were using either multiple or unapproved LLMs. And so there's just a lot of disparity between the tools that our people are using and especially for scaling and enterprise sized businesses, that security risk really becomes important. And we definitely see the CIO becoming a role that either is going to be a partner with marketing or have a stronger say in what are the types of tools that are being used, how are they being enforced. This is actually an interesting stat that Gartner had recently published around Guardian agents that they expect that, you know, 40% of CIOs are going to have a stronger role in having some oversight over what the AI tools are because there becomes so many implications around data privacy and PII and exposure of data and risk that potentially opens the business up to. And so I think, you know, that introduces a need for, within the marketing organization better skills and maybe more, you know, specificity in how you're using AI and what that skill Set and what that implementation looks like, but then also outside of it, across the organization, having more ownership probably in that CIO or CTO role is probably not far off on the horizon.

Lindsey Groper: Yeah, and I want to dig into that the shadow AI a little bit more. I mean, I suspect that most orgs would report that there are, there's a lot of use of shadow AI. How can marketers get control over shadow AI without stifling this like, spirit of experimentation? Right, because it's like you want folks to be trying new tools and doing new things, but also have to balance out the associated risks.

Holly Enneking: What I think about when it comes to this is doing my work out loud and encouraging my team to do their work out loud. Like, what are you doing? What are you trying? Having a spirit or a culture internally of open experimentation and knowledge sharing of what's working and not working creates an environment when there is some safety around the testing that you're doing, but also creates visibility into what are the tools that are using, how are they being used. If you don't create a space where people feel confident talking about what testing and what they're trying and what's working and not working, then it creates this culture of like, oh, I have to hide the tools that I'm using or I know I shouldn't be doing this and so I'm going to like, do it on the down low. Like, if you as a leader or a member of your team can enforce this sort of culture of knowledge sharing that really goes a long way. One of the things that we've seen in Markup that has worked really well is we have this incredible engineering team and they were able to build us, uh, sort of an internal tool that we use that's of part pulling in all of the different models so we can pick and choose between ChatGPT and Claude and Gemini and all of those sort of in this one space that is built to protect our personal data, our customer data, our employee data, uh, and still give us access to all these tools with all of the freedom that we would have outside and sort of bring it in house. And putting some investment of both resources and budget into that at the leadership level really creates a space where like, people feel like they have the tools they need and they don't have to go outside for them. I think that's really where companies have the opportunity to really be leaders in this is just willing to be open and have those conversations and then you have the visibility into what is and isn't working and where might there be risk that someone wouldn't be aware of.

Lindsey Groper: I really like that, of doing your work out loud and being the model for your team. Um, on what that looks like, I want to switch gears and talk about another area that the report addresses, which is associated risks. There's regulatory risks, there's copyright and infringement, there's overall just brand inconsistency. Are there certain areas that you feel marketers tend to overlook when it comes to the risks associated with AI output?

Holly Enneking: Yeah, I think it really depends on the business. But I think no matter the business, there is always risk that happens. If you're in a regulated industry, obviously that risk means opening yourselves up for lawsuits. Fines like that can be a pretty extensive list of and like really immediate hard problems that you're up against even if you know you're not in a regulated industry. Though I know for me and for some of the marketers that we did more of the in depth interviews with, they really recognized that even just doing something like you had mentioned the typo in you know, a content piece, uh, that has uh, flowed out to a bunch of places, something like that really impacts your brand trust. Even those small things make a big difference. And that can be the difference between someone taking your brand seriously or not and having credibility and losing credibility. And so while it may not be the same outcome as, you know, some sort of fine that you're getting that can have a ripple effect in your long term revenue. And so as marketers, our responsibility, we all know that like we have to be putting our best foot forward. And so like that can be really challenging for marketers and I think we all feel that. But I think too especially you know, the risk isn't just for the regulated businesses alone. Like it really is across the board for everyone. Like nothing is more important, especially for B2B brands than um, you know, the trust and credibility that we build over time with our audiences. And small things can add up and make a big difference in whether or not you maintain that trust.

Lindsey Groper: When you talk about getting your compliance house in order, which is one of the first things that you said, I want to dive into that a little bit more because if, if as your statistics show that most marketers are sitting there like uh, yeah, we, we really don't have our compliance house in order as it, especially as it relates to the AI content, who do you need to get in the room? And I know this is going to vary by like size of org because uh, if I'm a market leader, I'm like agreed. Who do I Need to get in a room together. Are there templates that I can use? How do you suggest doing that? Or, uh, what's like, the first thing that you should do in order to make that happen?

Holly Enneking: Yeah, I think the first thing that I would do if I was sitting down and like, what was it that I need in order to feel confident? It probably starts with a list of names of, like, these are the people that I would want to give me their blessing on this content. And that can be someone who's a subject matter expert, that can be someone who's on your, like, legal and compliance team in order to say, like, yes, we have done what we need to do. Like I mentioned when I was working with this health care system, we had a medical editor who was on our team. Every piece of content that we wrote went through this medical editor, and she was incredible at helping us make sure that we were writing at the right reading level, that we weren't saying things that we shouldn't be saying, that we weren't using language that was potentially setting us up for risk. Like, her review was incredibly valuable to us. And so that was just a part of our process. Anything that we went, went through her. And so if you can know internally who is that person? Maybe two people, like, what's your task force in order to get content out the door? And there's probably some thought around, like, what is the rubric of what requires review and doesn't. It may not be worthwhile to have your, you know, team reviewing every single social post or every email that's going out the door. But those big content pieces, yes, you probably do want those reviewed. And so having internal rules around what needs to be true in order for it to be reviewed, who needs to sign off of it, off on it, the last thing you want is to ship it and then you get feedback two days later from the legal person that you didn't wait on, that you've got to make a change. So I think those are a great place to start.

Lindsey Groper: Okay, that's really helpful. Thank you. Didn't want to leave, uh, our, our listeners hanging of like. I agree, Holly, but I don't know what to do for us. What other stats are important do you think to highlight from your report?

Holly Enneking: Yeah, I think a couple of the other things that were interesting was one that stood out to me was that, you know, 45% of the marketers that we surveyed did believe that AI could be responsible for checking itself. But on the flip side of that, 77% of our respondents said that they were still relying on manual reviews and spot checks in order to check their content. So there's definitely some tension between feeling like, oh yeah, AI's got it, but also I'm still going to double check it and make sure that things are where I think they need to be, you know. And the flip side of that 45% is that like 55% already know that they need something more in order to help them make sure that their content is ready to go. So that number is only going to change, especially as we see the outputs continue to be challenging. Yes, AI can write a lot faster, but like I said, that quality that, you know, reliability of the output is not always there and manual reviews just are simply not enough.

Lindsey Groper: It's a really interesting stat. What do you think is driving that over claiming AI maturity?

Holly Enneking: I think some of it comes from, and maybe it's just like when they think about certain types of content that they're checking or that they're drafting and that they trust versus don't trust. I think some of it too is that we're all just trying to figure out like how do we meet this moment of that like top down pressure of saying you have to adopt AI and thinking that it's maybe more capable of doing what we think it can than it would be. And I mean, one of the things that we really heard in conversations and something that stood out in the report that we did was that people went to AI originally to generate content not because of speed or cost. They really wanted AI to improve content quality. And that just really hasn't been the case in what we've seen. And so I think some of that is coming from that adjustment of oh, I thought it could do these things, but maybe that's not exactly what is happening. And so learning to sort of readjust our expectations and our processes around it as we're getting more familiar with what is and isn't possible. I think the only other thing that was really interesting from the survey was that of the, you know, people that we spoke to, both across marketers and executives, 96% said that some sort of Guardian agents would be valuable for them. And this idea of Guardian agents is sort of a new category. It's something that Gartner has been speaking about a lot more. And so we sort of fall into like a content guardian agent sort of space within that. But the idea of tools to help AI to check the AI really like what are some things that I can do that are really tactical to help me make sure that all the things that need to be true are. And catch those. Not just the standard use cases, but also the edge cases where if I'm just relying on manual review, that's not going to happen. And I think even though there is, you know, a lot of trust in that, 45% who say that they rely on it, 92% who are using it a lot more. There is a very big, almost 100% audience that says they need other tools in order to help them make sure that their content is right.

Lindsey Groper: Yeah, it's the wild west out there now. People are purposely writing typos and LinkedIn responses so that they know it's human. And I'm like, oh my gosh. Uh, it's just killing me. I'm like the, the little red squiggly line that indicates a typo. It's like, now that's proving that.

Holly Enneking: Yeah, that's some sort of like, like it's wild.

Lindsey Groper: And AI checkers are rejecting content that are literally written by humans. I don't hope you saw it was going around on, on social, but some guy uploaded his like thesis paper from like 1965.

Holly Enneking: He said it was all AI generated.

Lindsey Groper: It's just like, you know, because AI is just feeding itself. You know, everyone is scrambling right now. Uh, if you're listening, like, no one's got it figured out. I love the idea of a content guardian agent, like a specific use case. Do you currently have an agent that is available? Is that something that's coming? Where's Markup with that?

Holly Enneking: Yeah, so we have some things coming soon which we're incredibly excited about. So if people are interested in getting in and trying these agents as soon as they're available, you can actually sign up for priority access right now at Markup AI, just a simple name email and you can get in in the list in order to be notified as soon as our agents are live. And you'll be able to use them directly in Google Docs, which is awesome. I know of a lot of the marketers that we spoke to and I myself, I'm in Google Docs 99% of my day. So having something that's right there that I can be using to check and you can pick and choose the agents you're using and they check all sorts of different things. So we're excited to get this out and get it in the hands of marketers and get some feedback about how it's working for them. So if you're interested, especially in helping drive the future of what these tools might look like this is a great opportunity to get in early.

Lindsey Groper: Awesome. And assume web is the place to get the full report as well.

Holly Enneking: Yes. Yeah. If you go to the website, you can get our AI Trust Gap report. And it's, you know, free, easy to download, no gate or anything. You can just come in and download the PDF.

Lindsey Groper: Well, this has been awesome, Holly. Thank you so much. As I end every episode, I ask my guests if they have a signature or a favorite toast to send us out.

Holly Enneking: This is not appropriate for my margarita, but I come from a German American family. We go to a big German festival every year. So Prost would be the pro. The Cheers pro.

Lindsey Groper: I think I've maybe had pros one time in six years. I've ran this podcast. Not very often. Yes. Uh, which German festival is it?

Holly Enneking: Um, it's called Freudenfest. It's in Oldenburg, Indiana, Southeast Indiana. Every third week of July, every single year. Has been going on for about 70 years now. It's a staple. My whole family goes every year. I've got the traditional German Durandal. A lot of my family does as well. It's my favorite weekend of the year. I love it. So if you're ever in Oldburg, Indiana, need some fried chicken? That's the place.

Lindsey Groper: I actually have a friend who was, uh, like, born and raised in Oldenburg, so I have heard all about this for many years and she does the same thing. Yep. All the full gear, all the set, and I've not been but heard it say grand old time.

Holly Enneking: Oh, it's so fun.

Lindsey Groper: I will drink to that. Thank you. Thanks again to Holly for joining me on Sass Half Full. Love that. I got to connect with another indie local, and there were some really actionable takeaways for this particular episode. Things that you should be doing now and tomorrow, not next quarter or next year. I hope you enjoyed the episode, really appreciate the listen. And until next time, bottoms.