Productside Stories

AI Product Marketing: Selling AI Customers Don’t Trust

Featured Guest:

Zak Cherif | Senior Director of Product Marketing at KeyShot
10/27/2026

 

Summary

In this episode of Productside Stories, Rina Alexin talks with Zak Cherif, Senior Director of Product Marketing at KeyShot, who has spent more than a decade building go-to-market strategy at companies like Amazon, Microsoft, and DoorDash, including the launch of Microsoft’s first AI-powered sales companion.

KeyShot is the industry standard for 3D product visualization, and its customers render physical products before a sample exists. Many of them are enterprise, IP-sensitive companies that will not hand CAD files to a cloud model. So KeyShot’s first AI features run on a local model that users download once and run with no live internet connection, included in the KeyShot Studio subscription at no extra cost. Zak explains how privacy and “no extra cost” became the value props that earned hesitant designers’ trust, and how that early adoption gave the team the feedback it needed to plan further AI investment.

The conversation then turns to Zak’s day-to-day as KeyShot’s only product marketer: using AI as a sounding board and devil’s advocate, tracking competitors and voice of customer without the weekly Gartner calls he had at Microsoft, and pulling pricing and packaging questions (credits, tokens, and usage limits) into product development much earlier. They discuss hyper-segmentation, meaning messaging by industry and by how customers feel about AI, and where product managers and product marketers should overlap. Zak closes with advice for PMs who have no product marketer (be the representative of the customer) and explains why he pressure-tests his messaging in both ChatGPT and Claude instead of settling on one.

Key Takeaways

Marketing AI to customers who don’t want it.

  • KeyShot’s first AI features run on a local model that users download once, with no live internet connection. Privacy and no extra cost became the value props that got hesitant designers to try it.

Free AI is a feedback engine.

  • Including the local model in the KeyShot Studio subscription removed the adoption hurdle and gave the team the early signal it needed to plan further AI investment. Pricing and packaging questions about credits, tokens, and usage limits now enter the product development process early, instead of after the roadmap is set.

AI as a devil’s advocate, not a yes-man.

  • As KeyShot’s only product marketer, Zak uses AI for competitive intelligence and messaging variants. His favorite request: poke holes in this. He also explains why he switches between ChatGPT and Claude.

Hyper-segment your AI message.

  • KeyShot’s customers range from eager adopters of cloud AI to people who do not want to touch AI or cloud with a 10-foot pole. Messaging has to account for both industry and how customers feel about AI.

Product manager vs. product marketing manager.

  • Zak has worked where product owned the roadmap and where business leaders owned it. His advice for product manager vs. product marketing manager friction: learn who owns what, then work the overlap. With no product marketer on your team, take on the mandate of representing the customer.

Chapters

  • 00:00:00 – From Content Marketing to Product Marketing: Zak Cherif’s Path
  • 00:03:00 – What KeyShot Does: 3D Product Visualization You’ve Already Seen
  • 00:04:23 – How to Market AI Products to Customers Who Don’t Want AI
  • 00:07:40 – Why KeyShot Shipped a Local, Private AI Model
  • 00:11:23 – How AI Changed a Solo Product Marketer’s Day-to-Day
  • 00:12:15 – Use AI as a Devil’s Advocate, Not a Validator
  • 00:13:30 – AI for Competitive Intelligence (Without a Weekly Gartner Call)
  • 00:16:15 – Maintaining a Focus on Core Value Propositions
  • 00:18:05 – How AI Shapes Product Development and Planning
  • 00:19:58 – Free AI as a Feedback Engine: Pricing and Packaging AI Features
  • 00:23:20 – Hyper-Segmentation: Messaging by Industry and AI Sensitivity
  • 00:25:03 – Product Manager vs. Product Marketing Manager: Focus on the Overlap
  • 00:28:00 – No Product Marketer? Be the Representative of the Customer
  • 00:29:49 – ChatGPT, Claude, or Both? Don’t Settle on One Model
  • 00:32:05 – Closing Remarks and Where to Connect

Why Listen to This Episode?

In this episode, you’ll get:

  • An  built for skeptical, IP-sensitive buyers instead of another list of AI tools
  • Why KeyShot gave its first AI feature away inside the subscription, and what it got back in feedback
  • Zak’s habit of asking AI to tear your messaging apart instead of applauding it
  • How PMs with no product marketer can take on the job of representing the customer

Shipping AI features your customers aren’t sure about? Our AI Product Management course teaches PMs to build and launch AI features people use.

Introduction & Guest Background

From Content Marketing to Product Marketing: Zak Cherif’s Path

Rina Alexin | 00:00:00 – 00:00:40

Hi everyone, and welcome to Productside Stories, the podcast where we dig into the very real and raw lessons learned from product leaders and thinkers all over the world. I’m your host, Rina Alexin, CEO of Productside, and today I’m talking with Zak Cherif, the Senior Director of Product Marketing at KeyShot, who spent the last decade-plus building go-to-market strategy at companies like Amazon, Microsoft, and DoorDash, including launching Microsoft’s first AI-powered sales companion. We’re going to be talking about how AI is changing the actual day-to-day work of go-to-market, where the line between product management and product marketing can get blurry, and how you market AI to customers who maybe aren’t sure they want it. Welcome, Zak.

Zak Cherif | 00:00:40

Thank you so much for having me.

Rina Alexin | 00:00:41 – 00:01:00

Zak, I’m excited to talk to you because I don’t always get to talk to a product marketer. In fact, I feel I talk to more product managers who wish they had product marketers in their corner. So you seem like a rare bird, and that is why I’m excited to also ask you: how did you end up in product marketing in the first place?

Zak Cherif | 00:01:00 – 00:03:00

Yeah, I like the “end up” part, because I think if you talk to most product marketers, they would say the same thing. It’s new enough that you kind of do end up here, and maybe that’s changed. But for me, my early career and background was in content marketing, where there were a lot of shared principles. You’re working with a lot of different stakeholders. You’re getting to know the product or the services that you’re selling very intimately. And as I grew within my role and got more exposure to different groups, I had my first interactions with product marketing teams.

One of the things that really stood out to me initially was going to events and seeing all these different channels where a company’s message was coming through. My world was so isolated to just digital, right? The blogs, the podcasts, the social media, whatever it may be. I was really dealing with the final downstream output of positioning and messaging and go-to-market. As I got to know more and more product marketers, I realized these are the people who are actually putting together the core message that I’m then building content around. And I found myself really fascinated with that. So I looked for any excuse that I could to get closer to the product marketing teams that I worked with and just learned as much as I could, consumed and consumed, and looked for opportunities where I could make that jump from just content to product marketing.

So for me it was a really organic path. But I still work so closely, obviously, with content marketing and all the groups that I was more officially a part of beforehand, and I have an appreciation for what they do and how difficult that work is. I love being able to work with so many different people on so many cool things. And I’m personally so technical and interested in how things work that, conversely, I’m always interested in working with product managers and getting to know more about their world. So it’s been a really cool experience.

What KeyShot Does: 3D Product Visualization You’ve Already Seen

Rina Alexin | 00:03:00 – 00:03:30

Well, in terms of what you just said, it’s interesting, because oftentimes product marketers do come from more of a marketing background, and sometimes they lack the technical aspects. But you clearly worked at companies that are tech-first and digital-first, and worked on products that were very tech-first and digital-first, so it makes sense that you would have the blend of both. But maybe tell us a little bit about KeyShot, because I think most of our listeners know DoorDash. What about KeyShot? What is your product, and who are your customers?

Zak Cherif | 00:03:30 – 00:04:23

Yeah, so KeyShot is the industry standard for 3D product visualization. I like to simplify it, and this is how I learned it myself: when you think of creative tools, you think of Adobe or Canva, maybe Figma for more web or digital experiences. KeyShot is where our users visualize physical products and make them look photorealistic or abstract or anywhere in between.

My favorite thing that I was told as I was trying to learn more about the business, because I had never heard of KeyShot before I started working here, was: you may never have heard of us, but you’ve absolutely interacted with a rendering or a visualization that was done in KeyShot. When you think of a physical product, you think of:

  • A product landing page on a website
  • Billboards and ads
  • Anywhere you’re seeing a physical product

KeyShot is the design tool where that’s coming to life.

AI Integration & Product Positioning

How to Market AI Products to Customers Who Don’t Want AI

Rina Alexin | 00:04:23

Okay. And then, with it being a design tool, in what shape or form is AI getting integrated into the product itself?

Zak Cherif | 00:04:30 – 00:06:35

Yeah, there are a lot of different opportunities, especially with how quickly AI is growing and being adopted, and you see it everywhere that you read or consume things today. AI is a little bit of a tricky one when it comes to product design, and I think it varies person to person, company to company. But for us, one of the biggest things that our customers are concerned about, especially as you get into those higher enterprise-level, higher-revenue, more IP-sensitive companies, is that actual 3D data: the CAD files, the intellectual property of what their actual products are. Depending on the level of sensitivity around that, or even willingness to share some of that information or data with a third party or AI-based functionality, it’s a very different story back and forth.

Today, where we’re really laying the groundwork within KeyShot is trying to build and launch complementary AI features. We don’t want to necessarily automate every aspect of product design. That’s where our core user base, industrial designers, engineers, you name it, that’s where their creativity shines and where they actually stay true to the accuracy and the fidelity of their company’s products. So there’s a little bit of risk with AI, because you’re getting more and more confident in what the outputs are, but there’s an inherent risk in:

  • Sharing your data
  • Something, especially from a design element, getting altered in that process

A pretty common use case for us is that a lot of companies will use KeyShot to actually render a physical product before they get a physical sample. So the introduction of AI adding any complexity, any risk of issues or inaccuracies, muddies that to some extent. A lot of where we’re starting right now is more digital outputs:

  • Backgrounds
  • Changing materials
  • Using AI prompts, really getting into the actual design workflow versus the final output

And then we’re investing time and doing a lot of research to figure out how we continue to launch and complement that design workflow end to end with AI baked in, in a way that our customers feel the most comfortable with. That’s the most important thing to us.

Rina Alexin | 00:06:35 – 00:07:40

So I want to definitely come back to the comfort level, but the reason I asked you about KeyShot is that I can imagine, when we talk about how AI is changing our world, it changes our processes internally, but we’re also shaped by what our industry or our customers need. The market requirements are changing, customer behavior is changing, and in different ways by industry. So your own point, in terms of where you can insert AI into the product and where you might not, really shapes what you can even do. Just as you were talking, AI is very probabilistic, right? It can insert a lot of error. I’ve even heard this from product managers and designers, where the designer will see that there is an unclickable button in some kind of prototype that a product manager might not, but you can’t really put that in front of a customer. And that’s kind of where I want to start. Given your industry, given some of the constraints that you just mentioned, how do you even position AI in your product to that market? What do you really need to be sensitive about?

Why KeyShot Shipped a Local, Private AI Model

Zak Cherif | 00:07:40 – 00:09:45

I think it’s kind of an ongoing learning, to be honest with you. One of the big decisions that we made, in terms of what’s available today that’s AI-based functionality in KeyShot, is based on a lot of feedback we got from customers in early conversations, and a lot of the current feedback: how can we get access to automation workflows that are AI-powered without the security concerns being a larger factor?

The first set of AI functionality that we launched within our core product, KeyShot Studio, is actually a local-based engine, meaning that you download a model one time. It’s a certain file size, and that AI model lives within your on-prem desktop software. There’s no live internet connection like you would have with a cloud model, or ChatGPT, or even Google Nano Banana when it comes specifically to 3D design. And I think that has established a lot more trust with our customer base: I can at least explore with the fidelity that my designs aren’t getting passed back and forth to the internet, and that I’m not running into security concerns I might run into with other cloud-based AI functionality.

That early feedback and that early signal is informing not only how we’re building, but how we’re talking about it. From a product marketing standpoint, I’m always curious about value props and what stands out most to our customers, and that translates to the order that we’re showing things on a page or the language I’m putting together for a sales team on how to actually talk about a product. For us, the resounding use case value prop that came up over and over again was: “I appreciate that this AI is locally based. I’m a little bit more comfortable trying this out. Maybe I’m getting a little bit less pressure from my IT department about whether or not I can even use it.” That still exists, of course. But if we had gone a different route and really just matched the market deployment, where everything is cloud-based, I don’t think we’d even have enough early feedback or signal to help inform additional AI investment.

Rina Alexin | 00:09:45

So when it comes to having a localized model, how do you get feedback from what the customers are actually doing with it?

Zak Cherif | 00:09:50 – 00:10:40

A lot of it’s word of mouth:

  • Reaching out, trying to get a customer on the phone, email
  • We have a very active community for KeyShot that we host, and a dedicated team that’s managing that
  • Our product team does a great job of soliciting a lot of that feedback, and our customer marketing team as well
  • I’m always curious, so I’ll come over the top in the forum, or I’m scouring Reddit threads, or reaching out to as many people as I can just to hear, “Have you tried it? What was your experience? What worked great? What was disappointing? Where can we improve?”

But I’d say it’s very tried and true: just trying to meet people where they’re at. I think the first question is, do you know that we have this, and if you’ve used it, what’s your experience been? And then building from there.

Rina Alexin | 00:10:40 – 00:11:00

Yeah, it reminds me of how different cloud software, like SaaS companies, were compared to on-prem. But certainly there are industries that really want to be protective of their IP, that have serious ramifications if something were to come out. I can imagine a design, an original design: you can only have one original design. Right?

Zak Cherif | 00:11:00

Right.

Rina Alexin | 00:11:01 – 00:11:15

So it’s interesting to see this, I will call it an innovation for on-prem, to enable technology use while still maintaining this kind of “it’s okay, you have your own black box.” Right?

Zak Cherif | 00:11:15

100%.

Rina Alexin | 00:11:16

So we talked a little bit about your industry, and certainly how you’ve had to maneuver in order to gain trust with your customers with regards to AI use.

AI in Day-to-Day Product Marketing

How AI Changed a Solo Product Marketer’s Day-to-Day

Rina Alexin | 00:11:23

But let’s talk about you yourself. How has AI really changed your day-to-day?

Zak Cherif | 00:11:25 – 00:12:15

Immensely, in every form and fashion I can think of. The speed that I can work at, even having a dedicated companion in some senses, like an assistant that can talk through product marketing with me. I’m the only product marketer at KeyShot, and I’ve been in similar team formats at other companies. I’ve worked at other companies where we have a product marketer per product. There are other organizations that have product marketers per vertical, and they’re all focused on the same product. So there are so many form factors that product marketing can exist within. But as a solo PMM, especially:

  • Being able to refine my prompts and run things like competitive intel
  • Getting different variants for messaging that I’m testing
  • Getting honest, critical feedback about messaging that I’m preferring

Use AI as a Devil’s Advocate, Not a Validator

Zak Cherif | 00:12:15 – 00:12:55

Even something as simple as:

  • “I think this sounds good. How does this read to you?”
  • “Can you play devil’s advocate for me?”

I think that’s by far my biggest use case: poke as many holes in what I’m saying here, what makes sense, what’s confusing, what’s not. Of course, I’m always doing that with my teammates as well, but more often than not, I’m finding an angle I didn’t consider, or somewhere that I’m underselling something or overselling something. For me, my usage has grown day over day, week over week. It’s been incredibly helpful to have a sounding board, but also a central place that’s constantly learning and growing from how I’m interacting with it, and then making the work easier over time, which is incredible.

“Can you play devil’s advocate for me? I think that’s by far my biggest use case.”

Rina Alexin | 00:12:55 – 00:13:30

The more I talk to people, the more happy I become hearing when a person is using AI to poke holes in their thinking, because I think too many people use it to validate their thinking, and it will do that in a heartbeat, right? It’s kind of built that way. So do not use it to, I don’t know, argue with your spouse and show, “Ha ha!” Because it will agree with you. But if you can do it in the way that you’re describing, I think that’s probably the best way to stress-test your thinking.

AI for Competitive Intelligence (Without a Weekly Gartner Call)

Rina Alexin | 00:13:30 – 00:13:45

So you mentioned competitive intelligence. That’s something that, even if you had a team of other people to lean on, I feel like competitive intelligence is sometimes the can that gets kicked down the road the most. So how are you using AI today to maybe have more frequent competitive intelligence?

Zak Cherif | 00:13:45 – 00:16:15

Similar to the way that product marketing teams are composed, I’ve had different access to resources depending on where I’m at at different times. In my time at Microsoft, I met with Gartner every week, every other week. A lot of the motions I got comfortable with, and access to information that was really helpful and informative to a product roadmap or a go-to-market, you get used to that and you love it, and then you lose it when you leave. So for me, I’ve found Claude has been extremely helpful in terms of:

  • Pulling market intelligence
  • Helping me track different launches across competitors

One of the unique use cases for KeyShot, more so than anywhere else I’ve worked before, is that our competitive landscape is very complex. There’s a lot of, I think the word is, coopetition, right? We work with a lot of the people that others might consider our competitors, and then we compete in a very specific way. Our main focus is 3D product visualization, and for a lot of other tools in the market, that’s one piece of their offering. The CAD solution may be their bread and butter, and they also offer visualization and rendering as a post-production output of the models that you’re building out. Or vice versa: you can do some rendering and then pull some things into Adobe or whatever else it may be.

For us it’s interesting, because depending on who you talk to and what industry they work in and the products they work on, their view of KeyShot and the use cases that it’s meant for is vastly different person to person. So having AI technology that can compile all of that and also help make sense of it all is very helpful. I think this group might be a competitor, but there’s a lot of research online that can be pulled in where maybe they don’t view us as a competitor, or maybe they’re starting to, or we’re seeing launches coming out that could be viewed as something competitive. It’s extremely helpful for sourcing a lot of that, but also for getting customer research and feedback, voice of customer: how are people talking about other tools or competitors in the space, and how can we use that to better inform our messaging and our go-to-market efforts? It’s a very wide range, I’d say. Even just helping ground ourselves with an outside perspective: if I ask a product manager who our main competitors are versus a seller today, I’m getting two very different answers. They both can be right, but how do we account for all of that?

Maintaining a Focus on Core Value Propositions

Rina Alexin | 00:16:15 – 00:16:45

So if I understood you correctly, essentially, because there are a number of different tools that somebody you’re selling to might need in order to make a finished product. Some of those tools have something that could compete with you, some of them don’t, but the ones that don’t, they might want to build it, right? It might actually inform their product roadmap. Are you doing the same? Are you also informing your own product roadmap based on that?

Zak Cherif | 00:16:45 – 00:18:05

Yeah, in a way, for sure. I think maybe unintentionally we created this niche for ourselves where we’re the best at visualization. So even competitors who have that aspect within their functionality, it won’t be the reason why they invest heavily for that to become even on par with what we do. But now the question is, if they were to go that way, what are we building that will continue to set us apart and maintain a leader position? Or do we expand our horizon and say there’s another world we live in where we build our own CAD functionality, where you’re pulling in your 3D data and your models and actually refining that? To us, that’s not as important as making sure that the visualization aspect is always best in class. Everybody’s going to be different, but for us it forces that question to the top and makes us think through where we invest. Is it continuing to enhance this experience to make sure customers are getting value when they’re using us as part of their process, or do we expand our foothold across the entire process? That’s something that’s always been in the back of our minds. But access to more information, consistently, in a consumable way, like we get from this AI functionality, makes those conversations easier to have more frequently.

How AI Shapes Product Development and Planning

Rina Alexin | 00:18:05 – 00:18:25

Have you seen any changes in your product development process? We should probably talk about what your release schedule is, but are you seeing a better quality of decisions being made on what to include in a release? Is the speed of that going up, or maybe it’s just the quality of it?

Zak Cherif | 00:18:25 – 00:19:58

Definitely. I think the communication has probably been the biggest unlock for me:

  • More frequent communication
  • Faster documentation

I find this funny, but the design around how I’m getting this information has clearly improved a lot. Some of that might be AI, some of that might be creativity on the product management side specifically, but absolutely it’s changed, and I think for the better.

There’s a kind of parallel path of how we’re utilizing AI within the product planning workflows, and then also thinking about our AI investment, and how more access and more usage of AI tools is helping inform us on how we need to position and prepare these for our customers. The prompt is king; the chat is king today. So as we think about launching AI-specific features, how do we introduce them in a way where we’re not requiring customers to learn a new behavior or change the way that they interact with other AI technology today? From a pricing and packaging standpoint, there’s a lot of conversation about:

  • How you price something like this
  • How you make it easy for people to understand what a credit or a token is worth
  • Whether you give them different usage limits, whatever that may be

All of those things are now becoming really important conversations in the product development process, where before I think they would be more likely to happen once we had a completed roadmap and were figuring out how to take these things to market. So fundamentally, I think AI has had a huge impact across the board, from planning our roadmap to what we’re prioritizing within it.

Monetization and Product Strategy

Free AI as a Feedback Engine: Pricing and Packaging AI Features

Rina Alexin | 00:19:58 – 00:20:35

Can we talk briefly about that pricing and monetization? At the start you mentioned that you have a customer base that’s very sensitive to AI, which makes sense given their requirements around IP. I’m just curious, how do you even decide? You’re going to give them a localized model. How do you monetize that? Because that value is given to them once, and I suppose you personally don’t have the costs; they have to pay the costs of it being a localized model from an energy standpoint, I guess. But how do you even monetize that kind of feature that you’ve invested in?

Zak Cherif | 00:20:35 – 00:22:35

Interestingly enough, for the localized model, we built it in a way where it’s included within your KeyShot Studio subscription, so there’s no additional cost. We don’t run into the same usage issues because it’s not cloud-based, or the idea of who’s going to cover the cost of it. It’s all part of that downloadable model. Part of that was that we want to introduce AI functionality, like I said, in a way that’s not dramatically changing customer behavior or adding another roadblock or hurdle to engagement or adoption. We can learn a lot more from it the more people have access to it and can actually interact and engage.

That’s one of my favorite parts as well: the value props that landed really well with customers were:

  • First, the local and private security of the local-based model
  • Second, there’s no additional cost, which is not the case for any other competitor today, which is great

Within that, obviously, it’s not going to be updated as frequently as a web-based model. So there are some drawbacks. It’s not that we’re the latest and greatest and keeping up with the demand. But if the trade-off is, “This is free within my existing subscription, and I don’t have to worry about my designs getting leaked or sent out to the cloud,” why wouldn’t I give this a go? That’s also been a big part of those feedback conversations.

“If the trade-off is, ‘This is free within my existing subscription, and I don’t have to worry about my designs getting leaked or sent out to the cloud,’ why wouldn’t I give this a go?”

In terms of the longer-term pricing strategy, as we think about more AI functionality that’s coming, that’s one of the big hurdles we’re working through today. We’re learning a lot from competitors and from other AI leaders who are changing their outputs on a day-to-day basis. You’re getting a new model in cloud or ChatGPT what feels like every week, every month at least. You’re getting models that are updating on top of one another, with different uses for each of them. We haven’t added that level of complexity in what’s live today, but as we think about where we’re going, it’s definitely something we’re paying a lot of attention to: how the market is going about it.

Rina Alexin | 00:22:35 – 00:23:20

Yeah, I can imagine that is a tough set of conversations strategically. We haven’t talked about customer behavior, but on the retail part, consumers are getting used to having AI at their fingertips for free. There’s Gemini, there’s ChatGPT, there are all the free versions of these tools. On the business side, willingness to pay is, I think, correlated with the sensitivity around making sure my IP is safe. But given what you are delivering to your consumer base, which is, I think, closer to a free model, how do you then get them to pay for other features that cost your company actual tokens and usage? How do you get them on a usage-based model when they’re not used to it with your company?

Hyper-Segmentation: Messaging by Industry and AI Sensitivity

Zak Cherif | 00:23:20 – 00:24:45

I think when I think about what’s coming, and I won’t get into too much detail, there will be more of a setup in line with what you expect from the market today. We have a huge appetite for cloud-based AI functionality. We also, on the other side of the spectrum, like we’ve talked about, have customers who do not want to touch AI, or even cloud for that matter, with a 10-foot pole. For us it’s going to be a lot of accounting for those different audiences and making sure that we understand how we message this and where the appetite is going to be. Traditionally we’ve built pretty all-for-one, and even within KeyShot, people use it much differently depending on what they’re designing:

  • Retail products, whether it’s apparel or footwear or anything like that
  • Furniture
  • Manufacturing or consumer electronics, where we also have a huge foothold

In a nice way, we’ve accounted for different messages for different audiences from a base functionality standpoint, and now we have to apply that same level of thinking when it comes to AI: not just industry by industry, but across that sensitivity spectrum of “Am I even open to using something like this, or is it completely out of the question? We don’t want anything to do with it.”

Rina Alexin | 00:24:45 – 00:25:03

So are you saying that with AI you might be enabled to have not just personalization, but almost like hyper-segmentation of product, as in you can have a very different level of product for all of the different types of segments you just mentioned? That’s very interesting, and I wouldn’t be surprised if many companies go there.

Collaboration and Advice for Product Teams

Product Manager vs. Product Marketing Manager: Focus on the Overlap

Rina Alexin | 00:25:03 – 00:25:21

But that definitely requires a very strong collaboration between you and your product management function. So maybe talk to me about that. Given that the future may require really strong understanding of your various markets that you will serve, how do you stay very close with product management without stepping on each other’s toes? Because there’s certainly still a lot of overlap between you two.

Zak Cherif | 00:25:21 – 00:27:10

100%. I think that’s the common theme that connects all product marketers and all product managers. The big first step that I want to take anytime I get into an organization is to understand the lay of the land in terms of who owns what. I’ve worked in organizations, Microsoft, DoorDash, KeyShot, I’d say as well, where engineering and product own the roadmap, they own the vision, they own the product strategy. That’s great. And there my role is more: how do I represent the customer within that? How do I play devil’s advocate, like I ask my AI to do, and make sure that we’re not building what we think people want, but what our customers are asking for?

I’ve also been on the other side of the spectrum, most specifically at Amazon, where the business leaders or the marketing leaders, whoever it may be, outside of product, are defining that roadmap end to end. For me that was a huge experience booster: understanding the complexity of building requirements, scoping them, prioritizing them against one another, doing that twice a year, and then working very closely with product management, TPMs, whoever it may be, to validate that we can do what we’re actually setting out to do. Thankfully, I’ve seen it from both sides.

I’m interacting with my product team almost daily, if not multiple times a day, and I think having that connection is super important: having a working relationship before anything else. And maybe just advice for any role: really understand how people like to work, and then adapt to that. It’s been a lot less about ownership for me personally, and what’s my area versus what’s yours, and more trying to find that Venn diagram. There’s actually a lot more in the middle where we overlap, and then for our separate ownership areas, what are we responsible for getting across the line?

Rina Alexin | 00:27:10 – 00:28:00

I like that you mentioned that, because I talk about the need to have a RACI or DACI, whatever, to establish very clear lines around our roles and responsibilities. But you said something I think a little bit more profound, which is also recognizing that you are working with various humans. So not only is product management and product marketing perhaps different within every organization, as you just said, but the people you’re working with, their psychology, all of that, you have to understand. Sometimes I do encourage, especially, the typical MBTI personality tests, which can often be very powerful, especially once you find out, “Oh, I’m an extrovert and this person’s an introvert, and we have very different styles of how we process information” that could actually be in conflict. That’s why I think what you just said is quite important.

No Product Marketer? Be the Representative of the Customer

Rina Alexin | 00:28:00 – 00:28:15

But still, and I mentioned at the start, many of the product managers I talk to, even at large organizations, do not have access to a product marketing resource. So what advice do you have for them? Can they use AI to help? Where should they start?

Zak Cherif | 00:28:15 – 00:29:49

Absolutely. For me it goes back to the best definition I’ve heard of product marketing, and where I try to emulate that as much as I can: product marketers should be the representative of the customer within the company. And I think that’s a mandate anybody can take on. Quite honestly, I would say that to a salesperson or a CSM or a CEO, whoever it may be. If you’re leading with customer centricity at its core, you’re always going to be going down the right path. You’re going to make decisions and learn a lot. At the end of the day, if that’s your core tenet, that’s the most important place to start, in my opinion.

“Product marketers should be the representative of the customer within the company. And I think that’s a mandate anybody can take on.”

My advice to a product manager without a product marketing counterpart would be to start there and think very critically:

  • Question the decisions that you’re making
  • Validate the strategies that you’re building
  • Make sure that it’s not what I think this should be versus what my users are going to think it should be

And then, of course, AI is a great tool to bounce those ideas back off of:

  • Here’s something we’re building
  • Here are the problems that customers face today
  • These are the solutions we’re hoping to get from it
  • How can I build out a cohesive narrative that really gets that across, and where are my gaps?
  • Where are the holes that I need to fill in?

I think there’s a lot of access to resources thanks to AI today to help bridge that, with or without a product marketing counterpart. But selfishly, if anyone ever needs any product marketing help, they can always reach out to me as well.

ChatGPT, Claude, or Both? Don’t Settle on One Model

Rina Alexin | 00:29:49 – 00:30:05

I will have to ask you just one quick follow-up, and I don’t know if you have a good answer, because you might have an LLM or an AI that’s sanctioned by KeyShot. Of the more popular LLMs out there, is there one that you think is better at messaging?

Zak Cherif | 00:30:05 – 00:31:55

It’s funny you say that, because I’ve actually seen it go back and forth. I started very much as a ChatGPT guy, and I found, as I got deeper and deeper, I’d say even over the course of this year, I felt very back and forth. In January, for example, we had our revenue kickoff coming up, and I was bouncing some ideas about different taglines. We were introducing a lot of new messaging, and I built a lot of that myself and then used AI to pressure-test what I had in place. When I got down to specific things like a headline for a web page, or whatever it may be, I had my idea of it, and then the pressure tests I was getting back from ChatGPT were very fluffy, right? It was like, “Your idea is the best. Don’t worry about anything else,” which I love to hear, but I know that’s not true. That’s the whole reason I’m here. And that actually led me to start using Claude for the first time, where I felt like I got a lot more critical feedback.

“It was like, ‘Your idea is the best. Don’t worry about anything else,’ which I love to hear, but I know that’s not true.”

Fast-forward five or six months, I had the same experience within Claude. I went back to ChatGPT and found that was a little bit more advantageous for me. So it might be a little overkill, but when it comes to pressure-testing, I find myself trying it in both and seeing what comes back. I’ve dabbled a little bit with Perplexity and Gemini and whatnot, and I’ll use those here and there, but I think ChatGPT and Claude have been my main two. For me, it’s about pressure-testing them against one another and figuring out what makes the most sense. But who knows? A week from now, a month from now, I might give you a different answer.

Rina Alexin | 00:31:55 – 00:32:05

I mean, the models do change quite a bit, but I was curious to hear your experience. And I think that’s the right call: don’t get settled with just one, because you might get a suboptimal output as a result.

Closing Remarks and Where to Connect

Rina Alexin | 00:32:05 – 00:32:15

All right, Zak. Well, I so enjoyed talking to you today. How can our listeners follow you or get in touch after they listen to this episode?

Zak Cherif | 00:32:15 – 00:32:30

Thank you again for having me. I’d say LinkedIn is the best way. You can find me at Zak Cherif on LinkedIn, and anywhere else you find me, feel free to reach out, but LinkedIn is the place I’ll be most active.

Rina Alexin | 00:32:30 – 00:32:50

Wonderful. Thanks again, Zak. And thanks, everyone, for listening to this episode of Productside Stories. If you like today’s conversation, please don’t keep it to yourself. Share it with a friend and make sure to subscribe so you don’t miss a future episode. I’m Rina Alexin, and from all of us at Productside: let’s do product better together.

Frequently Asked Questions

What is AI product marketing?
AI product marketing is positioning, pricing and launching AI-powered features so customers understand what they get and trust it. In this episode, KeyShot’s Zak Cherif describes it as fitting the AI to the customer’s real concerns, such as IP security and design accuracy, rather than leading with the technology.

How do you market AI to customers who don’t want AI?
Start with the risk they care about, then build the feature around that concern. KeyShot found that enterprise and IP-sensitive customers worried about sharing CAD files with cloud AI. So it shipped a local model that runs on the user’s own machine, and it messages privacy and no extra cost first.

Why did KeyShot launch a local AI model instead of a cloud model?
Many KeyShot customers won’t send confidential 3D design data to third-party or cloud AI. The local model is downloaded once and then runs with no live internet connection, so designs never leave the machine and IT objects less. The trade-off is that it’s updated less often than cloud models.

Should AI features be free or usage-based?
It depends on what you need right now. KeyShot included its first local AI feature in the KeyShot Studio subscription at no extra cost to drive adoption and get early feedback. Zak expects future cloud AI features to follow the market’s usual credit or usage models.

How can product marketers use AI without getting “yes-man” answers?
Ask the AI to argue against you: poke holes in your messaging, play devil’s advocate, and stress-test your thinking instead of confirming it. Rina Alexin points out that AI tends to agree with you, so a critical prompt matters more than a polished output.

Which AI model is best for marketing messaging: ChatGPT or Claude?
Zak’s advice is not to settle on one. He found ChatGPT’s feedback too flattering at one point, switched to Claude for sharper critique, later went back to ChatGPT, and now pressure-tests messaging in both, while also trying Perplexity and Gemini.

What is the difference between a product manager and a product marketing manager?
It varies by company, so first find out who owns the roadmap. At Microsoft, DoorDash and KeyShot, product and engineering owned vision and strategy and the PMM represented the customer. At Amazon, business and marketing leaders defined the roadmap end to end.

What should PMs do if their company has no product marketer?
Take on the product marketing mandate yourself. Zak’s best definition of product marketing is “the representative of the customer within the company.” Anyone can play that role by questioning decisions, testing strategies, and using AI to build a coherent narrative and find gaps.

How is AI changing product development and go-to-market?
Communication between teams is more frequent and documentation moves faster. Pricing and packaging decisions about credits, tokens and usage limits now happen early in development, not after the roadmap is set. Zak’s rule is “the prompt is king”: add AI features without forcing users into new behaviors.

What is hyper-segmentation in AI marketing?
Hyper-segmentation means writing messaging for both a customer’s industry and how they feel about AI. KeyShot’s customers range from eager adopters of cloud AI to people who won’t touch AI or cloud with a 10-foot pole, so one AI message can’t land with all of them.