IN CLEAR FOCUS: Satej Sirur, CEO of Rocketium, discusses how AI is transforming creative operations for brands like Amazon and Walmart. Satej explains how Rocketium combines creative automation, brand governance, and analytics to streamline content production while maintaining brand consistency. Satej also shares insights on using AI to improve creative workflows, how performance analytics reveal unexpected insights, and why autonomous AI agents may soon monitor campaign performance 24/7.
Episode Transcript
Adrian Tennant: Coming up in this episode of IN CLEAR FOCUS:
Satej Sirur: With our AI, we are able to very visually show you how your content is performing, what are the groups of images that are performing worse than the others, what is common in the best-performing ones, what are some suggested actions both in terms of changing the budget or starting, stopping an ad.
Adrian Tennant: You’re listening to IN CLEAR FOCUS, fresh perspectives on marketing and advertising produced weekly by Bigeye, a strategy-led, full-service creative agency growing brands for clients globally. Hello, I’m your host, Adrian Tennant, Chief Strategy Officer. Thank you for joining us. Today, brands are under unprecedented pressure to create, manage and optimize high volumes of creative content across multiple channels and platforms. With the emergence of artificial intelligence and automation tools, some marketing teams are rethinking their approach to creative operations. But a consistent challenge is maintaining brand consistency while scaling content production efficiently. Our guest today is at the forefront of solving these challenges. Satej Sirur is the co-founder and CEO of Rocketium, an AI-driven creative operations platform that helps enterprises streamline and automate their creative production processes. Prior to founding Rocketium in 2015, Satej held key roles at Microsoft and Amazon Web Services, and he was recently recognized as one of Rethink Retail‘s top AI leaders for his work in helping brands like Amazon, Colgat, and Walmart scale their creative automation and analytics. To discuss how AI and automation are transforming creative operations, I’m delighted that Satej is joining us today from Dallas, Texas. Satej, welcome to IN CLEAR FOCUS!
Satej Sirur: Thank you for having me, Adrian.
Adrian Tennant: Now, for listeners who may not be familiar with the term, could you explain what creative operations, or just ‘CreativeOps’ means, and why it’s become increasingly important?
Satej Sirur: Of course it didn’t. So the word like most other technical terms that come up in the industry comes up from a purpose that people are experiencing today in their daily lives. The work that they do between marketing, design and business teams to create content so that it shows up for each audience on the right channel. The behind the scenes work is almost like an assembly line. where different teams are involved, different steps in the process, multiple tools that are involved in the back and forth that happens between teams. That entire process is called Creative Ops. And the reason to focus on that today is that unlike the days of yore where you had one billboard, one magazine, or one television spot, where you had to spend a few months to get the messaging right. And then once you come up with that great single idea, it goes and sits there for several more months. So you could take all your time to do it. But today audiences are on many more channels. They care about many more things. Digital gives you a way to tailor the message for each person. The message also very quickly becomes stale once people have seen the content a few times. So even if it was a great idea and it worked, you have to periodically keep making some tweaks to it. Not to mention the myriad sales events and various other events that keep happening for each company, which necessitate another set of messaging, another set of content. So that process is now come into clear focus, if I could say, which is the reason for that is just so much more content is getting pushed out. Our teams are being stretched. And so people are wondering, isn’t there a better way to do this versus going to Don Draper and getting one great idea.
Adrian Tennant: Satej, what inspired you to co-found Rocketium?
Satej Sirur: So Rocketium, luckily for us, we have given it a broad enough name. It indicates the elemental form of a rocket, so it could really be applied anywhere. Rocketium originally started off as a gaming company. The idea was people could come in and create content to share their ideas. ideas inspiration anything that they want other people to know in the form of a game and then over the years as we spoke with different people who came to us with different needs we realize there was a much larger problem to solve as it happens what we are doing today was the result of a couple of fortuitous events. There was a lunch with a journalist who ended up writing about us. Somebody from Amazon saw this article and then reached out and then shared some internal challenges that they were facing. We realized the underlying platform could be used to build a product on our own. And what started off as a product for Amazon’s advertising division, we realized it was a much broader problem. And like I said, every company has been doing this in some way or the other. And we were able to realize that this was a broader problem and many more companies need this.
Adrian Tennant: Well, I mentioned a few in the intro, but what are some of the best-known brands or organizations your platform serves today?
Satej Sirur: So one is, I just wanted to give you a sense of the kinds of companies we work with. So today we see this as this virtuous triangle in a way of marketplaces or ad platforms where the content eventually shows up. Then the brands who have that messaging that wants to go out and then agencies who work with both the marketplaces as well as the brands. So on three kinds of companies in this virtuous triangle, so to speak, are our customers. Amazon, whom I mentioned. In brands, you have companies like Colgate and Kellogg’s. In terms of agencies, we have Publicis as a large partner of ours and several more.
Adrian Tennant: Great. The Rocketium platform combines creative automation, brand governance, and creative analytics. So could you walk us through a typical creative production workflow?
Satej Sirur: Yeah, and because Rocketium did not start by saying here is an existing tool and here are the gaps in it, but looked at an entire end-to-end process and figured out what are the different bottlenecks within it. Rocketium today, like you pointed out, has expanded to have many different areas within it. So the typical process that happens within Rocketium is combines all of these three parts. So it starts with somebody setting up the standard assets and designs in their design system, so that it can be used to make many more versions of content. Then a team, typically a business team, marketing or category teams and so on, would raise a request on Rocketium, filling in whatever information they need, which channels, what sort of copy, which are the assets to use, whatever. Then either a creative manager or designers themselves pick up these tasks and then go into our design product where they make all of the different content versions. And before it goes out, because the request brief is already within Rocket EM, they check against that, make sure everything that is needed is there. There is also a layer of automated checks. which looks for brand elements, platform elements, any sort of best practices and checks against that. So if you are violating any of those, you cannot even share this content, you cannot export this. So that is one level of checks, but then approval workflows are built in. So you can assign it to the next person, move it to the next status in your custom workflow, they get notified, they come in, whether they are designers or not, they can look at the content depending on the permissions. They can make small edits themselves and now the content is ready to be downloaded. This is another place where Rocketium helps by resizing the files to the appropriate platform guidelines, making sure files are named according to your format and so on. These are small things that waste a lot of time for teams. And then once the campaigns are running on each of these platforms, Performance teams are also able to go in and look at the performance of it, diagnose any problems before they become chronic issues, and then repeat that process all over because some content needs changes or you want more campaigns to go out.
Adrian Tennant: As regular listeners know, we love case studies on IN CLEAR FOCUS. So, Satej, could you share an example of how one of your clients has transformed its creative operations?
Satej Sirur: Yeah, so I’ll start maybe with our first customer who has now expanded to multiple geographies, multiple categories and business units, all of whom work independently. There are about a dozen teams at Amazon who are Rocketium customers. But if I take one of those, which is one of their largest categories, in their case, what happens is the process more or less sounds like what I described before, but The people involved in that are the category teams. They have a need for content because it has to go live sometimes on their website, sometimes for paid marketing campaigns. They would raise this request and the rest of the process is the way I described. But because Amazon has such an ops-oriented culture, such a data-driven and analytical culture, they go a few steps beyond what I described before. And I want to go a little bit deeper into that. One is when they raise requests, they have to follow a certain format because the design has to be made in multiple sizes. It has to be made for multiple content versions. So it has to be given to designers in a certain format. They had an internal format within the category team. The design team had a different format. So they worked with our team to use AI and transform from their internal format to the format that is required by designers. And this cut down several hours from their process. So this is where, whether you do it on Rocket AM, whether you do it outside, it starts with having a certain kind of culture and mindset to say that, can I go into each of my processes, figure out where things are taking time. Same way towards the end of the process, when you have to do review, not only are automated reviews happening, but the team said, Hey, we put in so many comments. Can we see what sort of feedback we are even giving to the designers? So again, they worked with us using AI, we analyzed what the comments are about, and we realized a large percentage of comments are related to the logos. And so they were able to change their process again to ensure that an entire class of feedback just completely disappeared. And so that’s a process that they have and want to highlight the fact that because this is not trying to replace one single tool, but a process, there are many steps in the middle that bring in inefficiency or rework or any sort of repetition. And AI and automation and data can be used to improve those.
Adrian Tennant: Well, you mentioned AI. So how does incorporating Artificial Intelligence change how marketing teams and agencies approach creative production?
Satej Sirur: I believe that using AI as a silver bullet and this giant hammer that is going to come in, bash into the existing process and make it this beautiful, homogeneous and always working kind of a process. I do not think that will work. This is not just because of those 15 fingers and the kind of cliche places where stuff breaks. But this tech is used more as a scalpel. It is better than using it as a sledgehammer. So the way we see teams using it and where we have found them getting value is by examining their process and not only the big broad steps, like I said, a review or approval or any of that, but a smaller process. we are checking for certain kinds of things or we are renaming files. So if you break down your process into many such steps, you will find that almost all of those individual steps can nicely be improved using AI. And when it comes to analytics as well, a very similar sort of a process exists because one part is getting data from different sources. One is extracting information from that about the creative choices you’ve made, seeing patterns within that, seeing how stuff has changed over time. So AI can be used for any number of these things. Some places it will do well, some places it will not do well. AI also is good at more probabilistic stuff. It is not really good for very deterministic stuff. What I mean is, if you say that, okay, look at this and tell me the patterns or look at this and give me a rough sense whether something is okay, versus saying that, can you precisely calculate 17 times 855? It is not going to be good at the latter, right? You should use it in the right way, not use it as one single solution, but a series of point tools. That is where it would really work well.
Adrian Tennant: Rocketium works across every step of the campaign lifecycle, from planning and execution to optimization. How does the platform maintain brand consistency while still enabling personalization at scale?
Satej Sirur: Yeah, so when you have to do personalization at scale, what is happening is for each audience on each channel, based on the kind of message that works for them for each market, you have to make different versions. So one part of that is starting with a base design or a few key visual concepts and then making versions of it. So that’s the first place where brand compliance and governance and standardization comes in, that you are able to set up these designs Have the approved ones show up and older ones be hidden. With role-based access control, you can limit who can make new designs, which ones can be used, and who can see what. So that’s the first level of ensuring brand consistency. The second is that as all of these versions are being made, if you had to do it by hand, Simple stuff, like even if somebody is not a designer, they have done this enough number of times where in a presentation that you’re making, you would put a page number somewhere, you would put a title somewhere. Now, if you were copy pasting that and resizing and making all of these changes one by one, there are going to be a few pixels of differences as center alignment and all of those kinds of things. So because of tech, what is happening is these elements are being standardized with algorithms and software. You are automatically able to resize the text to fit within a certain box, make an image become big or small to fit within the container and things like that, which prevent you from having to manually do that. So that’s the other place where consistency comes in, which is especially important when you have so many different varied lengths of text and imagery and languages and fonts and so on. The other thing that happens after that is the automated checks that I talked about. So you can look for things like what words are we using, colors, overlap of different elements, certain kinds of tone of voice. So even some of the probabilistic stuff that I mentioned can be looked for. And then finally, There is a level of human checks because you can assign it to the next person, they can come in very easily see it. That might not seem like a lot, but in many of our companies for this review process, they used to have to export all of the files, put it in a Google Drive or a WeTransfer, put it in a presentation or a PDF, send that across somebody’s annotating that it was a big mess. Because the making and reviewing is happening in the same place, the entire process is cut down by a lot, right? So, honors for those places. One thing that we are exploring for the future is this whole concept of agents. With that, what we’ll be able to do is, these autonomous agents could be running and checking your different campaigns where content is running. So, whether you’ve made it on Rocketium or not, whether you’ve done the approval on Rocketium or not, we can see what is running live and check that against your guidelines and warn you if something is not on brand, right? So that gives one final net of safety. But yeah, those are interesting things that do come in the roadmap.
Adrian Tennant: Let’s take a short break. We’ll be right back after this message.
 | Alan Barker: Hello, I’m Alan Barker, the author of “The Complete Copywriter: The Definitive Guide to Marketing with Words,” published by Kogan Page. I’ll show you how to exercise your creativity, generate powerful ideas, maintain reader attention, and bring your copy to life. You’ll also learn how to develop a coherent content strategy, how to survive as a copywriter, and how to nurture a satisfying career. Whether you’re a professional writer already, a brand manager, or someone who creates content as part of another kind of job, this book will help you to develop the skills to craft compelling, customer-focused copy. As a listener to IN CLEAR FOCUS, you can save 25 percent on “The Complete Copywriter” when you order directly from Kogan Page. Just enter the exclusive promo code BIGEYE25 at the checkout. Shipping is always complimentary for customers in the US and UK. I hope my book helps you to become a more versatile, effective, and confident copywriter. Thank you! |
Adrian Tennant: Welcome back. I’m talking about AI-driven creative operations with Satej Sirur, co-founder and CEO of Rocketium. I’ve had a demonstration of the platform and one of the interesting aspects is that an administrator can set permission levels for different types of contributors. That means that non-designers have the opportunity to go in and potentially update headlines and copy, but they can’t change any of the brand design elements – is that right?
Satej Sirur: That’s exactly right, Adrian. And that is another thing that helps with consistency and compliance, not in the sense that you are able to prevent these people from doing something, but you are removing designers as the key piece in every one of these small changes. So designers are more free to do creative work or to review other things, but non-designers can come in, make small changes, and then you have the automated checks and so on, at least one level of review, and then designers are doing the rest. But that really frees up a lot of time for both designers as well as marketers.
Adrian Tennant: Got it. Satej, for brands that work across different international markets, how does the platform support localization?
Satej Sirur: So localization in the logistical aspects, what I said of having templates that are available, assets that are available, raising requests, having different roles to allow people to do different things. All of those are table stakes and they certainly help in this local market enablement because people are in different geographies. You might not work in the same time zone. So some bit of automated access and freedom within this sort of framework is very, very crucial. Then there are some things within the product that help in localization. A, from day one, the product is international. So no matter what language you use, what right to left or the UTF and some of those technical things that you have, that is supported pretty much from day one. Where we go one step further in that is as you bring in content from different languages, and you have mapped your brand fonts to individual languages, which you can do in Rocketium. We detect the language that is used in your copy and automatically set the font to that language’s approved brand font. So that’s as you’re making, when you asked about personalization, as you’re making content that sort of scale, having to go to every text element and set the font. These are small things. Only practitioners will realize the challenge in that. It’s not just about copy pasting a design and copy pasting the copy within that. There is also changing the font. So that happens automatically using AI as well as the more old school machine learning services of Google, Amazon, Microsoft. There is built in translation. So an entire design, whether it’s a static or an animated GIF or a video can be translated to another language. And because the review workflow is there, it can look for certain kinds of words, and then another person, a translator or an expert in that market could come in and make those changes. So localization, we support in all of these different ways. Perfect.
Adrian Tennant: We’ve talked about creative automation, which enables teams to create more content faster while potentially reducing costs. We’ve also discussed brand governance and the ability to ensure compliance while decentralizing content creation. So let’s talk about the third component, creative analytics. Satej, how are you seeing brands using insights from the platform to optimize their campaigns?
Satej Sirur: The first part where Rocketium adds value is the more logistical part of it, the operations part of it, which is, how do they even look at data? Today, they have to go to different platforms, download the data into an Excel, put it into another Excel. All of them have different formats. They have different terminology. Their campaigns are called different things. So that is the first part that we help with. We connect to each of those platforms, and on an hourly, daily basis, this data is automatically synced. So you bring in all of the data from each of these platforms. Second thing that we do is we add information that is not already in your ad name or campaigning. Most companies rely on naming their ads a certain way, which is test one underscore audience one underscore design style one and so on. Not only can we extract that, but we can also analyze your design and figure out from it different creative elements that you might not even consider are experiments that you’ve tried. So we can look at models, messaging, background, layouts, colors, and many more things and extract that information. And then third, we are able to allow you to make different kinds of widgets and reports using all of this data so that not only are you able to see how are my spends trending, what are the best and worst performing content, how are different creative ideas performing, but you can dive deeper and say that for a given business unit in this region for a given time period, when I try a 50% off versus a smaller discount percentage off, how does it vary? And the really cool thing about it is that some of these decisions you might not have been able to analyze before because it was just within the copy and you were just maybe trying different layouts. But now using this, we’ve been able to help our customers find out that a 50% discount offer in one of our customers case actually performs worse than a 30% discount. So not only was their ad performing worse, when it performed okay, they gave out more discount than they should have. Now that is you’re wasting money on both ends of the spectrum, right? So those are the kind of things that people can discover. The other thing that we also do is you have so much data. People who are able to like some savant, if you’re able to process all of this and really see the patterns and go deeper into each of these aspects, then it’s easy for you to use the reports and dashboards. But for most other people where this sort of analysis could take time or it requires specialized skills. With our AI, we are also able to very visually show you how your content is performing, what are the groups of images that are performing worse than the others, what is common between them, what is common in the best performing ones, what are some suggested actions, both in terms of changing the budget or starting, stopping an ad, or here are the creative ideas that you should be trying. So some of that place, AI really allows us to go a few steps further than what teams are able to do today.
Adrian Tennant: I think it’s very interesting, this ability to monitor and optimize credit performance within the same platform just makes a ton of sense.
Satej Sirur: That’s right. And this is where even the agents would come into picture. This is again, something on our roadmap where for every creative strategy that you have, these agents could be monitoring all of your ads and looking at are there dips, are there rises, are there some things that are more promising? Is there an answer you will know anyway in seven days? Can we tell you seven days ahead of time that this is looking promising, you should spend more here or this is dipping, maybe you should do something else, right? Your audience is fatiguing and whatever the strategy might be. So agents allow us to 24-7 autonomously work and look at something. So the strategy is yours. The agent is just doing that on a regular 24/7 basis.
Adrian Tennant: Well, you mentioned the product roadmap. So what plans do you have for the platform?
Satej Sirur: So one more philosophically, I wanted to say that what is our guiding principle when we think about what we build. So there is a great Jeff Bezos quote when people ask him about the future and to predict it. And he says, he focuses on what will not change. Something that will not change is something you could focus on because that will stay constant for the next five, 10, 15, 20 years. So in our context for teams, what will not change is they will always want better performing campaigns. they will always want to reduce the cost. They will always want their teams to be happy and productive and efficient. On the audience’s side, they will always want messaging that is more relevant. They’ll want messaging that’s more timely on the channels where they are. So if you focus on that, our roadmap becomes very clear. So some of the things I mentioned around agents, the power of that is that the AI that today is doing a great to amazing job when you ask it in a prompt, Now you could say why do I even want to come in every day and ask you and then you give me a report and then give me the answers and then I go and take the action. Can I not ask you to tell me when there is something I even need to look at. So offloading work like this to an agent so that it is doing stuff on its own. breaking down the process into individual steps can we say that okay which of these steps where agents can come in and help you make better campaigns or lower your cost make your team happier and so on so agents is one big area where we are focusing the never-ending in quest to help people be more creative. So the design product that we have built today is already fairly capable, but there are many more creative things that we want to do. So whether it is going deeper in video, more stuff on the vector graphic side, those are things we’ll be working on. And collaboration and finding patterns in feedback and who’s available to work, who’s overloaded, some of those things to help your team work better. So that’s the other broad theme of process for us.
Adrian Tennant: What advice would you give to marketing teams and agency leaders who are looking to test scaling their creative ops with AI tools?
Satej Sirur: One is don’t think of AI as this single solution, a single prompt where you write and you would get an answer. That is not how you should be using it. Because A, it is not ready. Two, it creates a black box where you do not have control over it. Even if it’s the right answer, you can’t even about why did I even get that answer? So you really want to break this process down into individual steps and bring in AI over there. The other thing that you want to do is more culturally, you want to bring in some aspects that are going to allow you to win. So just questioning the defaults. Today, you work a certain way because you’ve always worked a certain way. It doesn’t mean tomorrow you need to work the same way. It doesn’t mean just go around breaking stuff everywhere in your process, but really question what parts of the process do you not enjoy? What parts of the process are really sucking up time and focus on bringing in AI over there? Similarly, bringing in a culture of continuous improvement and analytics to measure what is happening. I gave a couple of tactical examples of looking at what sort of feedback are we getting, what sort of requests are we raising and so on. But even in driving better performance, there are much better, more analytical things that we should be doing. A lot of companies today are not looking at creative analytics because they look at creative as a black box and they just say that we are going to be playing around with budgets and bids and audience targeting, but creative ends up being an afterthought. So being able to use AI and automation to scale out the content should allow you to be more thoughtful in planning your campaigns as well. So I would say broadly, these would be the three areas that I would recommend companies look at.
Adrian Tennant: Great advice, thank you. Satej, for listeners who want to learn more about the Rocketium platform or connect with you, what’s the best way to do so?
Satej Sirur: So we’ve tried to put everything about our beliefs and what the product does, what value companies get on our website. So I would recommend visiting our website, seeing for yourself if this resonates with what you are doing, and then we have a simple form, fill that out. But most of the team is focused on customer success. So if you feel so inclined, you could reach out to any of us on LinkedIn and we’ll be happy to connect you with the right team.
Adrian Tennant: Perfect. Satej, thank you very much for being our guest this week on IN CLEAR FOCUS.
Satej Sirur: Thank you for having me, Adrian.
Adrian Tennant: Thanks again to my guest this week, Satej Sirur, co-founder and CEO of Rocketium. As always, you’ll find a complete transcript of our conversation with timestamps and links to the resources we discussed on the IN CLEAR FOCUS page at Bigeyeagency.com. Just select ‘Insights’ from the menu. Thank you for listening to IN CLEAR FOCUS, produced by Bigeye. I’ve been your host, Adrian Tennant. Until next week, goodbye.
TIMESTAMPS
00:00: Introduction to AI in Creative Operations
00:21: The Importance of Creative Operations
02:07: Defining Creative Ops
03:54: Inspiration Behind Rocketium
05:06: Notable Clients of Rocketium
05:40: Overview of the Creative Production Workflow
08:04: Case Study: Amazon’s Creative Transformation
10:26: The Role of AI in Creative Production
12:04: Maintaining Brand Consistency with Personalization
16:06: User Permissions and Brand Compliance
17:15: Supporting Localization in Global Markets
19:00: Creative Analytics for Campaign Optimization
22:22: Monitoring and Optimizing Creative Performance
23:07: Future Roadmap for Rocketium
25:08: Advice for Scaling Creative Ops with AI
26:49: Connecting with Rocketium
