Introduction:
In the latest episode of Applied Intelligence, we had the pleasure of hosting Jack Lindberg, the Director of Media, Insights, and Analytics at the Mars Agency. Jack’s unique journey from an opera singer to an AI and e-commerce expert gave us many insights. This blog post captures the essence of our enlightening discussion, exploring the intersection of AI and e-commerce and delving into the future of retail media.
Jack’s Transition: From Opera to E-commerce
One of the most intriguing aspects of our conversation was Jack’s transition from the world of professional opera to the dynamic realm of e-commerce. His shift, catalyzed by the COVID-19 pandemic, is a testament to the fluidity and interconnectedness of career paths in the modern world. Jack’s journey underlines a vital lesson: the skills and experiences we gather in one field can often pave the way to success in another seemingly unrelated area.
Decoding Amazon Marketing Cloud
A significant part of our discussion centered around Amazon Marketing Cloud (AMC) and its impact on e-commerce. Jack provided an in-depth understanding of data clean rooms and the unique position of Amazon in this space. His explanation of AMC as a “code-first interface” shed light on how brands can leverage this powerful tool to enhance their marketing strategies, underscoring the importance of adaptability and technical knowledge in today’s marketing landscape.
The Role of AI in Retail Media
Jack’s insights into AI’s transformative role in retail media were particularly striking. He discussed the early stages of AI in content generation and review summarization, highlighting its potential to revolutionize customer experiences. The conversation also touched upon the challenges and opportunities presented by AI, including privacy concerns and the need for customized solutions in various sectors.
Future Trends and Career Advice
Looking ahead, Jack anticipates significant developments in AI-driven search, predicting a shift towards more specialized AI models tailored to specific industries. His advice for those aspiring to enter the e-commerce analytics field is invaluable: master the basics like SQL and Excel and gradually build up to more complex tools and models.
Conclusion:
Jack Lindberg’s journey and insights offer a unique perspective on the evolving world of e-commerce and AI. His experience reminds us that the willingness to learn and adapt is crucial in navigating the ever-changing digital landscape. As we continue to witness the integration of AI in various facets of retail and marketing, Jack’s story serves as an inspiration and a guide for aspiring and established professionals in the field.
Hosted by: Imteaz Ahamed
Video Transcript
Imteaz (00:02.544)
and then we’ll get kicked off. Cool. Very cool. Yeah. Hi, everyone. Welcome to Applied Intelligence. This week, I have a really fun guest, Jack Lindbergh from the Mize Agency. So Jack is the Director of Media and Insights and Analytics at the Mize Agency and the product lead on Noctis, which is an AI copilot for Amazon Marketing Club. He is the 2023 Amazon Ads Partner, Macedo.
Jack Lindberg (00:04.744)
Cool, sounds good. Looking forward.
Imteaz (00:31.24)
and they’re full of professional opera singers as well. So there’s just so much to talk about with Jack and his very colorful background. I also met up with Jack a couple of years ago, well, he was at Pacview, which is another great e-commerce partner. So yeah, looking forward to this lovely conversation and welcome to the show.
Jack Lindberg (00:53.696)
Thank you so much for having me, excited to be here.
Imteaz (00:56.288)
So the question I ask on my guests on my podcast, and I love asking this question, which is if there was an autobiography of your life and it had five chapters, what would the chapter titles of each chapter be in your autobiography, Jack? Just so the audience gets to know who you are.
Jack Lindberg (01:16.24)
Yeah, that’s a really good question. I think for anyone my age to say that they have five chapters of an autobiography would be a little bit disingenuous. I’m 26, so going out in the world and saying, wow, I really lived five chapters worth of a life might be kind of hard. I think I would be at best at the beginning of chapter three. So I would sort of think about the chapters
Imteaz (01:34.456)
Thanks. Hahaha.
Imteaz (01:41.694)
Okay.
Jack Lindberg (01:47.96)
transitioning from being a professional musician to the world of e-commerce, sort of getting my bearings, my footing in the world of e-commerce. And then what I’m doing now, which is really like those initial opportunities of going out and trying to make waves and seeing what’s happening. I think my journey to e-comm, if there’s anything I’ve learned, is that everyone comes to this field because it’s so nascent.
from really random, totally off the beaten path, other fields, and listening to my story about coming from professional singing. If you talk to people in e-commerce, it isn’t actually that strange. Like I have one friend who was a professional portrait photographer before coming into e-comm. So I did my undergrad and my master’s degree in opera performance.
Imteaz (02:39.676)
Okay.
Jack Lindberg (02:47.568)
I studied at the Guildhall School in London in the UK. I was completing my masters right when COVID hit. I had my first professional season of swinging for the fences, being a freelance sopersinger in Europe, sort of rolling. And all of those gigs got canceled. I moved back home to New Jersey at my mom’s house. And three months in, she was like, I need you to get out of the house and get a job.
And the feeling was mutual. There’s only so much you want to live with your parents, as much as you love them. So with my two music degrees and no skills, I was sort of in a tough spot. And luckily, my godmother is a career coach, Debbie Walter. She teaches at Stanford Business School.
Imteaz (03:21.337)
Of course. Completely understand.
Jack Lindberg (03:42.604)
and runs her own consulting firm called Redbed Boon Consulting in Silicon Valley for professional career coaching. She’s my godmother. And I called her up and said, Aunt Debbie, I’m calling in all of my ex years of favors right now. Cash them all in. And I think I ended up trading her SAT tutoring for her daughter as a swap. But she talked me through.
Imteaz (03:56.31)
Cash in mid. Yeah.
Imteaz (04:03.592)
Okay, that’s a good trick.
Jack Lindberg (04:10.548)
basically figuring out what the next steps were. What did I like? What did I not like? What sort of soft skills could I play up in that first interview to say, yes, I don’t know how to use Excel because they don’t teach you that in music conservatory, but how do I get my foot in that door? And also how to pound the pavement and talk to people and learn about what’s out there. I didn’t know what.
analytics was or I didn’t know what a product manager was. And I was like, how was I supposed to know that was a career that I wanted if I didn’t know it existed? So I spent a lot of time just reaching out to random people and saying, hey, can I talk to you for 20 minutes about your job? No ask from you. Just tell me about what you do and what you like about it and what you don’t so I can learn about what’s out there to see what might fit.
Imteaz (05:01.337)
Can I ask you a question about that? How many people turned you down for that ask for 20 minutes of that time?
Jack Lindberg (05:08.032)
Oh, basically no one. I would say like 80% of people, 90% of people said yes. If I’ve learned anything, it’s that people love to talk about themselves. And if you go in and say, hey, I want to give you the opportunity just to tell me about yourself, and I’ll ask you questions, most people say yes. Because the key is like, you’re not asking them to do anything. You’re asking them to tell you about them.
Imteaz (05:14.816)
Okay.
Imteaz (05:19.88)
I’m sorry.
Imteaz (05:27.468)
Mm-hmm.
Imteaz (05:34.904)
Right. And this is the thing, right? Like a lot of times in our lives, we’re too scared to ask for help. We’re too scared to like, really leverage our relationships and or just, you know, reach out to someone and say, Hey, I’m going through this thing. Can you just help me through it? And, you know, I’ve had a similar experience every time that I’ve kind of had a
career issue or personal thing that I’m trying to figure out anytime I’ve reached out to a friend or like somebody relatively in the know about that topic or area, everyone’s always said yes, here’s 20 minutes of my time or half an hour of my time to figure that out. So yeah, sorry, keep going.
Jack Lindberg (06:18.792)
No, definitely, yes. People are generally much nicer than we give them credit for. And people will say yes if you ask them for help, especially if it’s something that they’re uniquely positioned to help you with. So I think I had maybe 25 or 30 of those conversations of tell me about whatever you do. And I was like, well, that gave me a lot of lists of things that I really didn’t want to do. And e-commerce was…
Imteaz (06:29.516)
for sure.
Jack Lindberg (06:49.172)
sort of the best bet because if there’s anything I’ve seen it’s that prior experience in this field because it changes so much doesn’t help that much. That being a 15-year e-commerce professional if you’re not up to date on the cutting edge you’re behind right and that sort of shelf life of knowledge doesn’t maintain the same way it would for banking or something like that.
Imteaz (07:16.609)
Yep, and we work, and I’m sure you work across multiple clients now, but even in CPG, right? The way things have been done within CPG have not really changed that much, especially from a supply chain. Maybe the marketing side of the equation has changed with the advent of Facebook and Google. But
in terms of how consumers are purchasing products and brands, how they’re discovering products and brands, how they’re experiencing products and brands has shifted and is continuing to shift. And we’re going to get into Amazon Marketing Cloud even later, but the partnerships that Amazon is leading within this space right now is only going to further fuel that change. So I know I’m taking too much thunder from our conversation later, but yeah, certainly.
100% agree with you that, you know, having, not being able to unlearn what you’ve, like what you’ve learned in your career thus far, and then picking up new skills, you can’t do that ecom is going to be a struggle for you.
Jack Lindberg (08:25.812)
I definitely agree. And my first job out of grad school was literally as an intern at a small agency in New York, learning how to manage Amazon PPC ads. And I swear to God, I told my boss, my then boss in that interview, I don’t know anything about this, but I promise you I can learn. And I think that held true. And then I shifted over to working at Pacview.
HackVue is filled with some of the smartest people in the industry. A bunch of really dear friends are still there. But my first role there was in customer success, which basically was you’re the person people ask questions to about e-comm retail media. And you’re seeing a huge wide swath of accounts from all different types of verticals, all different levels of expertise on the client side, different agencies.
And it was sort of a great trial by fire to say, you don’t know anything about this, and you need to come in and figure out how you can be helpful in that half an hour or hour to the person on the other end of the phone and know your stuff cold to be a valuable resource during that window. And what I found was I was answering the same questions over and over and over again and thought, there must be a better way to do this.
And the way to do, you know, solve all of those same problems at scale is to move from talking to the customer to building things that talk to the customer. And I shifted over to the product team at Packview, where I did a lot of work on the Amazon DSP and Amazon Marketing Cloud products and helped out a little other places. And then most recently in April jumped over to the Mars agency to pick up Amazon Marketing Cloud efforts there.
and most recently have launched Noctis in collaboration with some of our friends at Analytic Index, which I’m really proud of. It’s an AI co-pilot. Basically, we’re using a large language model to retrieve all the information you’d never need about Amazon Marketing Cloud. So, ask it questions. It knows the answer about what does this field mean? Should I use table one or table two? And even more concretely,
Jack Lindberg (10:53.136)
ask it a question in English, and it gives you back the code you need to get the answer. It’s very cool.
Imteaz (10:57.707)
Super cool. And, you know, based on your LinkedIn posts as well, Jack, over the last few months, you know, even using…
chat GPT to write the SQL queries to like go over Amazon marketing cloud and just like the basic stuff, right? That was super helpful for me. And now seeing it in a product that’s fully scalable, usable across multiple platforms or multiple use cases within Amazon is super cool. So well done there.
Jack Lindberg (11:30.964)
Thank you. Yeah, I was something we’ll talk about later. But with the OpenAI releasing this OpenAI Agents feature, you can basically build your own custom LLM-powered tool. There was a lot of concern from some of the people in our company that didn’t really understand the product, as well as those folks in the weeds. Like, hey, can I have someone go build this via an OpenAI agent?
And the answer was, I tried it. And you can’t. It’s really hard. There are a lot of companies out there that are using AI in a way that our VC friends call a thin wrapper, where it’s basically a prompt and maybe a little bit of data manipulation on top of a LLM. Something with that level of sophistication, OpenAI basically killed overnight.
Imteaz (12:05.039)
Yep.
Imteaz (12:17.456)
Yeah.
Jack Lindberg (12:29.708)
And you need to be adding value on top of that to actually be competitive in the market and not be killed by the proliferation of these sort of agents.
Imteaz (12:34.656)
for sure. Yeah, I also think the level of base level of consumer or customer understanding of these products is massively increasing very quickly over time as well. So the snake oil salesman who’s trying to basically give you some user interface with a little bit of prompting.
that generates specific outputs for you, that’s not gonna fly. Maybe you made a few sales now, but yeah, that lunch is gonna get eaten very quickly by a lot of players. Very cool. Okay, let’s go into some detail and some clarification of what data clean rooms are, what Amazon Marketing Cloud is for the audience. It’s a relatively new technology.
Jack Lindberg (13:15.45)
I definitely agree.
Imteaz (13:33.38)
I think data clean rooms have been around for a bit of time. And obviously, Amazon Marketing Cloud has been around for a while, but certainly making a lot more noise over the last two years since unboxed, I want to say two years ago, and now the latest unbox conference where, you know, every, I want to say almost, it felt like 80% of the announcements were around what Amazon Marketing Cloud is. So if you didn’t take it seriously previously, you should take it seriously now.
Jack, what’s your definition of what a data cleanroom is?
Jack Lindberg (14:07.096)
Yeah, I’ve had to explain this to a lot of people, and I try to do it two ways, depending on the audience. So I’ll do both here just for the sake of the audience. The technical definition is basically there’s a scenario in which you have two data providers, company A, company B. And they know that there is a valuable output from the aggregation of their two data sets, but they’re not willing to share the raw source material with the other party. So.
A common example of this would be like a researcher and health data. My mom’s a public health professor. This is something that someone could do in her field. For example, I’d want to know how many people who live near within 15 miles of an oil well have cancer relative to people who don’t live near an oil well. And you could use that data from.
a partner in a data cleanroom to say, hey, I can get anonymous aggregations with some privacy controls to sift through that data. But I can’t ask what’s Joe Blow’s address. You can’t reveal that level of PII. What Amazon has done that’s kind of different is in most data cleanrooms, there are three parties at minimum.
Data partner A, data partner B, and the intermediary, that cleanroom host. Amazon has decided that they’re gonna be two sides of that three-way relationship, and say they’re gonna be a data provider providing the Amazon ads data signals, and they’re using the AWS architecture to be the host. And then you as the advertiser can add in your own data as you see fit. The more fun way to describe it.
is thinking about my favorite episode of Brooklyn Nine-Nine, where there’s a great scene where Andy Samberg decides he’s going to buy a DJ turntable. And very much like data cleaners are like that, where the table itself is the architecture, and the different records you place on are the data sets. And you’re mixing them together to get some sort of aggregate output that.
Imteaz (16:16.197)
Okay.
Imteaz (16:25.68)
Okay.
Jack Lindberg (16:29.148)
is different, unique, and hopefully more helpful and better than the source input you have. In that episode, he ends up trying to mix two Klezmer records, and it goes horribly wrong. And I think that’s what most people’s experience, like using data cleanrooms, is if you put in garbage data where you don’t know how to get data out effectively, you’re not going to get anything useful.
from the cleaner of itself. It’s an enabling technology, not a spoon feeding of the right answers.
Imteaz (17:02.673)
Yeah, and I think the, let’s call it the advertising industry and the media industry, one of the, you know, paradigms has always been pushing ads and pushing more media and more creative, right? And now…
You kind of have to balance the mad men with the math men. And this is all math, right? Like in the sense that this is all enabling analytics technology that allows you to look at historical performance, allows you to look at future scenario planning of if you changed something within your media signals, and then finally it enables you to activate audiences on the fly and do like audience activation in real time.
It doesn’t necessarily help you with the art side of the equation, right? So if you don’t, if you have an agency partner that is very focused on the art side of the equation and doesn’t know how to balance the math side of the equation, you’re not necessarily going to win here. And similarly, if you’re too focused on the math and too focused on the numbers, uh, of your, you know, campaign performance, media spend, et cetera.
and you don’t look at the creative side of this equation as well at the same time, then this is not going to work either. So this is what I find is fascinating because you studied music in your past life, I studied visual arts in my past life and math, but for me this is a really cool amalgamation of two fields that from a marketing point of view.
Jack Lindberg (18:26.662)
Mm-hmm.
Imteaz (18:48.584)
is the synergy between the art of the possible versus how creatively can you find customer problems that exist and solve them through an enabling technology as well as add creative that actually really works within that scenario as well.
Jack Lindberg (19:06.5)
Right. Yeah, I think to add to your point, I think the data within a data cleanroom is really good at telling you the what and the how. Not so good at telling you the why.
Imteaz (19:16.617)
Yep.
Imteaz (19:19.868)
Exactly. And this is where you need humans, right? Like until we have an LL that can essentially be a brand marketer and or, you know, someone that really understands why people do something because of other causal signals. Yeah, you really do need humans to interpret all of these data and turn it into useful information, slice insights to actually do something with this stuff, right?
Jack Lindberg (19:47.556)
agree.
Imteaz (19:48.812)
So what really sets outside of the two parts of the three way relationship, what really sets Amazon marketing cloud apart from every other data clean room out there in the market. I know some of the retailers already have data clean rooms and have had data clean rooms for a while. But what really makes Amazon different?
Jack Lindberg (20:10.756)
Yeah, so I think what makes Amazon data clean rooms super different outside of the infrastructure piece is the way you interact with it. I’ve had a lot of conversations with other retail media partners or other retailers that are seeing the traction and press that Amazon Marketing Cloud has gotten and saying, hey, we need to step up our game with regard to clean rooms. And I think they are taking the opposite approach.
Amazon’s cleanroom is a code-first interface. You need to know how to code to see value from the platform, by and large. They’ve done some work to start rolling out, you know, no code or limited coding solutions that allow you to access some template types of data. But what we’re seeing from other cleanroom partners out in the space is something that I would describe as more greatest hits, where…
your access to the source tables to say, what’s here so I can make up any use case in the world is much more limited. It’s like, here are the 10 things we think you might want, and here’s a partner that’s going to show you a dashboard to access them, rather than saying, here’s our API go nuts, which is sort of Amazon’s approach. And I think those are really targeted at different types of people, like my colleagues who are expert media strategists and don’t know how to code.
would probably benefit much more from a platform that shows them the top 10 use cases, rather than someone like me who much prefers to look at the data and figure out for myself what I can do, rather than being spoon fed every use case.
Imteaz (21:55.795)
Of course, I think, yeah, Amazon’s.
a tech company first before it is anything else, right? So I think the quality of documentation that they give to highly technical people really enables people to kind of just come up with their own thing and then really customize solution to whatever they want. Whereas a lot of other partners will give you a stock standard, give the top things that you might be interested in. And if you work in a certain product category,
There’s very different to other categories. For example, if you’re in like pharmaceuticals and or healthcare, some of these lifetime values or some of the interactions that you have with the brand may only last for a couple of months and it might not be a sustained relationship with a brand. And like, for example, you know, the fashion category, there’s a lot of one-offs that you might buy from certain brands, but
Uh, if you’re in athlete, like athletic leisure, there might be more consistent purchase, right? So that customer journey is very different to, I don’t know, buying dishwashing tablets, for example, which you might be a lot more of a habitual purchase. Um, so clearly defining what you need and building custom queries and models to mine Amazon marketing cloud and then doing something with it, not just looking at what happened.
is super interesting and it’s very cool to see so many different use cases coming out of Amazon. So if I’m, let’s say if I’m a brand already doing over a hundred million dollars in sales annually on Amazon, what are the key steps that I need to take in order to kind of get started with AMC?
Jack Lindberg (23:45.664)
Sure. Yeah. So step one is checking to see if you’re eligible for Amazon Marketing Cloud. I know. If you’re doing over a hundred million in sales, I hope you are running upper funnel media through Amazon via Amazon DSP. That is a prerequisite from getting access to Amazon Marketing Cloud. If you manage to do a hundred million in sales and run no DSP, my hats off to you. I would probably still recommend it. Not knowing your business, you know, don’t take that as gospel truth, but
Imteaz (23:50.663)
Okay.
Imteaz (24:08.524)
Wow.
Jack Lindberg (24:16.584)
I would be surprised if you got that much, got that far without using any upper funnel media. Then you need to set up an instance, which is basically gathering the identifiers of your sponsored ads account and your DSP accounts and saying, hey, make a clean room that contains this data set to your Amazon rep. From there, there are really three options in terms of, or three combinations you could use to get further down into.
engaging with Amazon Marketing Cloud in a substantive way. So option one is you build out your own team. You hire some folks that know SQL and say, hey, your job is going to be to dig into this data set and figure out how to provide value to the company. That is, by and large, the least common approach. In many ways, I think it’s the best approach as someone who’s been a tool provider and an agency.
Imteaz (24:52.573)
Okay.
Jack Lindberg (25:15.473)
I’m gonna neg myself and say do it yourself. That being said, no.
Imteaz (25:17.546)
Yeah
Imteaz (25:21.304)
But that’s not for everyone, right? Like, you know, you need to have a base level of understanding of data science, a base level understanding of Amazon DSP. And then you need to be comfortable with hiring your own data scientists, whether that’s, you know, internally hiring your own data scientists, but also, and, or looking at, you know, data science providers that can really learn and specialize in this stuff. This is some stuff like I’ve had to do personally.
in like to, to really maximize Amazon marketing card. But I know even within my own business, most people, most econ people, or even marketers would not be comfortable doing that because it’s just so different to any other form of data that we’ve had access to previously.
Jack Lindberg (26:11.152)
Yeah, I think also for most organizations, unless you have someone, frankly, like one of us, who is sort of fluent on either side, where I can talk to a data scientist, and I can talk to a marketer, and I can sort of figure out what the other means when they’re trying to talk to another, typically there’s a big communication gap in that sort of in-house system, where the marketing team is like, this is what data I need to, these are the questions I have. And the data science team doesn’t understand.
Imteaz (26:21.484)
Hehehehe
Jack Lindberg (26:40.144)
what data answers that question or what you would need to know to figure out whether that’s true. If you execute it well, it’s definitely the best way. It is a large organizational challenge for folks to execute well.
Imteaz (26:56.402)
And I think it comes down to data literacy on the marketers part as well. Like, you know, if you understand that the internet is a giant pivot table and each one of the tables has attributes and those attributes are basically columns. And if you can define those columns and explain to anyone in very simple English.
Jack Lindberg (27:00.832)
Mm-hmm.
Imteaz (27:18.788)
what each one of those columns actually means and does in relation to not just the table, but the rows as well. Then you fundamentally understood one, the internet, but two, Amazon Marketing Card is just a different level of pivot table. You’re just calling on all of this information at specific points in time versus the product table, the order table, the customer table, the ads table, et cetera.
Imteaz (27:48.664)
answered and you can really easily call out the parameters in which you want to call. A data scientist can help you figure out the technical side of that equation out, but the challenge is really articulating the problem and the parameters you actually need to solve that problem.
Jack Lindberg (28:06.388)
I agree. Typically, what I do when I’m talking to someone on the media side who’s trying to figure out how to use Amazon Marketing Cloud is they’ll ask me a question, and I’ll open up a literal whiteboard and draw them a table. And it said, is this what you want? And then I can figure out how to get what they need. But I have to show them what’s available to see if that helps them answer their question. So option two would be using a tool provider.
Imteaz (28:20.224)
That’s a great technique. Yeah.
Imteaz (28:26.136)
Mm-hmm. That’s a great one.
Jack Lindberg (28:37.62)
the tool providers in the space, by and large, are all doing some variation of the same thing, which is they’re saying there’s some number of things out of the box that you can ask, and that’s it, unless you pay us additional money to build you custom reports. Hope that doesn’t sound pejorative, but that’s what everyone is doing by and large.
Imteaz (29:01.252)
Yeah, I’ve seen so many dashboard solutions. Yeah. And I’m like, guys, this is a very generic way of trying to solve the problem or trying to find the opportunity within Amazon Marketing Cloud because it has no context of what my business channel actually is. So you’re trying to sell me a square peg for a round hole. And you’re telling me it’s gonna work when I have a very specific use case and a very specific
Jack Lindberg (29:05.106)
Yep.
Imteaz (29:30.376)
or a very different product category that is not very similar to most of the product categories that are sold on Amazon, right? Okay, so tool providers.
Jack Lindberg (29:39.304)
Yep. Tool provider’s number two. And I think that’s why I took a totally different approach at building Noctis, which was my goal is to enable people to do custom first. Like, you can ask it any question, and it’ll try to write you code that gets you the answer to that question. Not, we’re going to give you the 10 dashboards that we think are going to be most common. It’s to help you figure out your custom questions on your own. Because if all I did was give you the same stock dashboard,
Imteaz (29:58.225)
Very cool.
Jack Lindberg (30:08.88)
All I did to the industry was raise the floor. I didn’t provide anyone competitive advantage. I just said, now everyone has the same dad boards and we’re all starting from square one and square zero.
Imteaz (30:11.472)
Yep. So you know how you just talked about whiteboarding the solution to, or whiteboarding the question and then referencing the parameters to actually answering that question. How does someone like build up the skills to do that themselves? How do you improve that data fluency?
from a non-technical marketer point of view.
Jack Lindberg (30:43.7)
I think really the core skill is just understanding what data is in there. You could even take it further, like ask the person to draw you a chart. Say if you were going to present this data to a client, what would the chart looks like that tells your story? And then we back solve from the chart to the table and then the table to the code. And if you ask a marketer, what chart do you need on the slide?
They typically can get pretty close to what they need because what they’re used to is storytelling. And you just say, OK, what plot do you need to be able to tell in order to have this story work? And then you figure that out and backs off. I think part of that is also understanding what data is in there. I find with Amazon Marketing Cloud, what happens to most people is they get overwhelmed. And they’re like, oh, I’m going to do this.
They realize that there are all these questions they wish they could have answered, or they couldn’t conceive that they could ask because they didn’t know the data was available. So there’s some sort of exploratory work that I’ve done, which is just describing to people what is in there. And then they can sort of understand, oh, I can ask for this and this and this and this because I know what the available parameters are. It’s like, you know.
Jack Lindberg (32:08.148)
trying to play Scrabble when you’ve never seen a dictionary before? You know, it’s a little bit hard.
Imteaz (32:15.103)
For sure, for sure. Yeah, I’m a super nerd, so I love reading data dictionaries for specific databases and stuff, so I really get to the nuts and bolts of what is actually available. I think…
With Amazon marketing cloud, I think something similar is needed to kind of give, you know, marketers a world win crash course in terms of how to use this stuff outside of just looking at the dashboards, which is fine. But when you truly understand the audience signals, the media interactions, the path to purchase analysis, the number of exposures to conversion, like when you see the
level of granular detail that’s even just available from the basic stuff, it blows your mind, right? Like Google Analytics kind of has this stuff when you think of direct to consumer e-commerce and the interactions that are being recorded within GA. But given the size and scale of Amazon and how much sales of multiple brands are going through Amazon, this is just different level of scale.
Jack Lindberg (33:05.12)
Mm-hmm.
Jack Lindberg (33:29.972)
Sure. It is definitely exciting. To finish off the question, option three is you work with your agency. And as you mentioned, MTS, the goal is to have an agency that can do both the math and the art. Or at Mars, we call it art and algorithm, which is very pithy. That being said, if your agency, like I’ve seen with a lot of folks.
Imteaz (33:30.029)
which makes it super exciting.
Imteaz (33:36.181)
Okay.
Imteaz (33:50.927)
I love that. Yep.
Jack Lindberg (34:00.404)
If your agency doesn’t take the call from analytics, because they’re very expensive, I’ve seen frequently, luckily not at Mars, but working at other places or with other partners, hearing horror stories about people who said, hey, I can’t get my agency’s analytics team in the room because we don’t have an extra $100,000 around for them to do something. I’m like, you know, there’s…
Analytics is now table stakes. It’s not an add-on. It’s not a nice to have. It’s a must have. And watching agency partners do a disservice to their brands by scoping it independently as an extra line item sends me up the wall.
Imteaz (34:33.749)
Yeah. Yeah, that doesn’t make sense to me. Which is a good lead into the next question, which is, you know, what are the hesitations that you’re seeing from clients slash
Jack Lindberg (34:50.697)
Yep, I agree.
Imteaz (34:58.064)
sellers on Amazon in reference to using Amazon Marketing Cloud.
Jack Lindberg (35:03.632)
I think some people think it’s too hard, and they just sort of give up before they’ve given it an honest try, which if I was a non-technical person and you showed me the AMC native console, I think that’s a fair reaction to say, I’m not gonna be able to figure this out. And I think the really big piece, which I harp on a lot is,
Imteaz (35:21.773)
Yeah.
Jack Lindberg (35:32.66)
The worst thing that can happen to me as an analytics professional is I do a really beautiful analysis. I put together a great story, and then no one changes their behavior. And I think that’s what happens with AMC a lot, is folks say, hey, isn’t it wonderful that your path to purchase goes through these campaigns in this order, and this is your Beth conversion path? And isn’t that awesome to know? And then no one does anything differently. Which I was like, what was the point?
of this analysis if we didn’t use it in a way that drives business impact. I think most brands don’t know how to take it to the next level and say, what’s the action step? And their agency partners and their tool providers don’t help them as much as they need.
Imteaz (36:20.16)
Yeah, so you know, this is literally just another tool and just another way of you know, you seeing your data. But if you fundamentally don’t take action from it and or change your behavior or change what you’re doing, definition of insanity is expecting a different result if you do the same stuff, right? So yeah, so again, we’re only going to get more as AI develops as
Jack Lindberg (36:28.608)
Thanks for watching!
Jack Lindberg (36:38.925)
Yep
Imteaz (36:47.996)
access to analytics develops as we get faster and faster computers the velocity of which this stuff will keep coming out is only going to increase but fundamentally if we don’t make little tiny changes or big changes and then measure the impact of that change relative to what we were doing before what’s the point so yeah okay where do you see data cleaner in
you know, Amazon’s announcements at unboxed in terms of connecting everything to Amazon marketing cloud, AWS, AWS’s partnerships now with Meta, Snapchat, TikTok. I think it was a couple more, as well as plugging in all of the video partners as well, like Free V, Prime Video, and Close Night Football. As this ecosystem just keeps getting bigger, where do you see?
the technology heading.
Jack Lindberg (37:49.216)
Yeah, I am probably going to give you an unorthodox take here. I think what I’ve heard a lot of people say is that there’s going to be some sort of aggregation of clean rooms, a clean room of clean rooms, or how I like to think about it, like the one ring from Lord of the Rings, the one clean room to rule them all. That’s not going to happen. Data privacy laws.
Imteaz (37:56.162)
Let’s go.
Imteaz (38:13.053)
Yep.
Jack Lindberg (38:18.236)
and especially like GDPR and whatnot, don’t allow you to use data for anything that the customer didn’t consent to. And building a massive clean room of clean rooms is definitely one of those things they did not sign up for. So this idea that you’re gonna be able to have like universal identity graphs that are, have 90, 100% match rates across all of your partners. And you’re gonna know track one consumer across the entire journey flawlessly and seamlessly. That’s not gonna happen.
Imteaz (38:48.401)
Nope.
Jack Lindberg (38:49.728)
If anything, the opposite is going to happen. Even now on Amazon, there are… The way Amazon identifies who you are is based on your status being logged in on amazon.com. And they use that email to see who you are. And all of your campaigns that are targeted via audience track you around the web and show you ads based on your browsing behavior and what you’ve looked at.
That being said, there are a large number of ads, like offsite contextual, so on and so forth, that target based on the content of the page, not who the person is. And therefore, when you look at the Amazon Marketing Cloud records, you see I have an impression and I don’t know who the person is. And if you say, what’s the total number of prescience when I have a known user versus when I don’t have a known user? The.
no user number or percentage is going to drop over time. And where that’s going to lead is basically conversion modeling to say, can I take a wild guess or a scientifically informed guess, depending on who you ask, on can I figure out maybe without cookies who that person is, or what type of person that is, or what sort of things could I say about them? Like,
If I show them an ad that’s contextually targeted to dog food, can I make a guess that they’re in the in-market for dog food audience? Probably. And as Amazon Marketing Cloud and all data clean rooms get further down this path of third-party cookie deprecation, the ability to track users seamlessly across the web is going to become more and more broken, more and more fragmented, and you’ll need more intense modeling to figure out. In the instances where I got…
I know I served an ad and I don’t know to whom. How do I figure out what happened?
Imteaz (40:43.358)
Yep.
No, the cookie deprecation piece, the following consumers around multiple parts of the internet when they’re logged in or incognito and or using privacy filters, et cetera, is only going to become more difficult.
Imteaz (41:15.056)
AWS clean rooms and being and letting AWS basically be the clean room provider of choice for multiple publishers and now social media networks as well is kind of Amazon’s insurance policy against all of this stuff. Right. I think this is a very intelligent play to at least in the Western markets have a strong defense.
against cookie defecation and at least marry up so much of the media sphere, the publisher sphere as well as the, obviously the e-commerce sphere as well and try and keep people as closed loop as they can within Amazon’s infrastructure. Now how Amazon navigates the privacy part of this and the consent matching across all of this stuff.
Imteaz (42:12.884)
I think their PR lobbies and a lot of their industry influences have a bit of work to do here, but I think this is the strongest play to marry all of this stuff together so far within ads.
Jack Lindberg (42:32.084)
Totally agree. Thankfully, there are lawyers that get paid a lot more than I do and are smarter than me that will figure this stuff out.
Imteaz (42:39.987)
Super cool. Let’s transition into analytics and AI. How have you seen AI transform the landscape of retail media and analytics so far?
Jack Lindberg (42:53.396)
I think what we’re seeing with AI is really just like step one of where AI can go. I think when we like conceptualize AI, the sort of seismic change it is, is at least on par with the advent of mobile. If not like the advent of the internet.
Imteaz (43:15.608)
Mm-hmm.
I’m going to go ahead and turn it off.
Jack Lindberg (43:22.548)
like I’m overselling it, but I think we’re still very much at the early stages of what that’s going to look like, right? For media, I think we’ve seen a lot of folks start with AI-generated content. Like Amazon released a feature that allows you to basically prompt an image generation LLM that says, take my product image and put it on my kitchen counter surrounded by a Thanksgiving meal for my toaster oven, whatever it is.
And we’ll see AI content continue to improve. Like Amazon’s also released like AI review summaries to allow the consumer to say, what’s the gist of all of these reviews in aggregate? I think what we’re seeing right now is our use cases where AI is very good, like out of the box, which is simple image generation, summarization,
where it’s not so good out of the box is complex generation or anything that requires fact checking or technical knowledge. It’s a little bit less good. With analytics, what I’ve seen that’s really quite interesting is how far you can get using, for example, like ChatGPD to, especially its code interpreter, to push.
the envelope of what someone who isn’t as data-fluent can get out of their data set with some help from an LLM. If you ask it the right questions, it’ll generate the code that gives you the response you need, which creates what I would describe as a plethora of citizen data scientists. It doesn’t scale. It’s not a model that you can build into an enterprise-wide application, but it does give people the tools to get much further on their own.
Out of the box.
Imteaz (45:24.758)
Super cool. Sorry, go ahead.
Jack Lindberg (45:26.036)
Um, one thing I would add to this is what’s going to come around the corner and that most people aren’t prepared for is prompts driven or text driven LLM based search. Our conception of the digital shelf, which is like a series of tiles of product images, um, is potentially fundamentally broken by someone going into chat GPT and saying,
Imteaz (45:47.383)
Yeah.
Jack Lindberg (45:55.752)
What’s the best orange juice for me to buy for breakfast? And the LLM delivering you a single response of this product rather than showing you the entire plethora of options.
Imteaz (46:09.653)
Or even just your four options rather than one, right? Like all 20 on a product listing page on Amazon, right? Like how do you influence that?
Jack Lindberg (46:19.035)
Yeah.
Imteaz (46:23.244)
is going to be super interesting. I think, again, Amazon’s play with Alexa, and I made sure I turned off my Alexa before we had this conversation, because every time we talk about Alexa, she goes off. But even with Alexa and more Alexa screens being available, I think it’s a perfect use case of a chat interface leveraging search and being more conversational.
when it comes to purchasing decisions for consumer as well. So, and that’s going to be in marketing hub too, by the way. Super cool. So as a product manager, how did you leverage AI to enhance product operations and decision making processes within the work that you do?
Jack Lindberg (47:14.052)
I think AI is a really great sounding board. It’s very good at… So I’ve been trying to adopt, to back up on stuff, some sort of Amazonian processes about the PRD, this sort of writing the press release before you ship it, before you write any code. And I found that sort of writing helps to codify your own thinking.
Imteaz (47:35.288)
Mm-hmm.
Jack Lindberg (47:40.932)
Even better is when you feed that into an LLM. It’s just going to take your thoughts and sort of flesh it out and give it back to you. And basically, if I handed someone a bullet point list of my ideas and it were to write the press release for me, would that press release make sense? And if the answer is no, then my thoughts aren’t very clear. And LLMs are great amplifiers of what you feed it. And they sort of fill in the gaps.
with what an average person might fill in the gaps with. And if they fill in the gaps incorrectly, that’s probably because you gave it like incomplete or imperfect lines of thought. And I found that super helpful.
Imteaz (48:25.594)
That’s really interesting, Jack, because the beauty of being a good communicator is being able to communicate things with a lot of simplicity. And an LLM kind of gives you that, right? Simplicity because it’s operating for the masses.
As long as the input that you are giving it is unique enough that it stokes something new, brand new and creative, then the language of the LLM kind of helps you get to the simplest way of communicating that. That’s really cool.
Jack Lindberg (49:06.256)
And if you put in something too generic, it fills in the blanks with something super generic. And you’re like, this sounds like an undifferentiated idea, and I don’t know what this means. Ha ha ha.
Imteaz (49:18.75)
Super cool. Okay, so what are some of the biggest challenges you faced in terms of integrating AI into analytics and how did you address it?
Jack Lindberg (49:28.924)
Yeah, I think some of the biggest concerns that we’re going to have with enterprise applications of AI, specifically third-party large language models, is going to be data privacy.
Jack Lindberg (49:48.276)
The big challenge is how do you enable folks to use these technologies without divulging their data to the owner of the LLM or some other system that enables the LLM to learn from your data and potentially regurgitate your results to someone else? So we sort of approach this in two ways. One is thinking about how much value can I provide using an LLM?
that knows about the problem generically, but not about your specific data. So the way we set up Noctis was it knows about Amazon Marketing Cloud. It knows nothing about your data. It doesn’t even have access to your data. So you can ask it, how do I see my top 10 products that drive new to brand sales? And it’ll help you with that. It just doesn’t know what the answer is, because it doesn’t have access to your data, which sort of firewalls it from saying,
I’m Pepsi, what are the top 10 best selling products for Coke? And having the response come back with a real answer. The other way you could solve this that we’ve been working on too is what I would describe as containerization. Maybe that’s not the right technical term. But the idea that you, on a per client basis or a per account basis, you’ve basically built a self-contained LLM that
Imteaz (50:59.739)
So, I’m going to go ahead and start the presentation. presentation of the
Jack Lindberg (51:14.056)
whatever the generic data set is, plus this advertiser’s unique data set and nothing else. So you don’t have that risk of advertiser A learning about advertiser B. We’ve gotten guidance from a lot of clients that they don’t want their data in LLMs at all. So if you’re trying to leverage this technology, I would definitely recommend thinking about what can you do with AI that doesn’t involve
ingesting customer data, that’s still valuable. And what customer data means depends on who you’re talking to. It could be the existing bullet points on your PDP. Could be considered customer data. So even for the simplest tasks of generating PDP content, you need to make sure you’re within the rules of engagement you set up within your brand or with your clients if you’re an agency or a tool provider.
Imteaz (52:05.86)
Right. That’s really interesting. I mean, from my point of view, all of that stuff is public.
Right? Like it’s published in the open web. There are a million scraping providers scraping Amazon on a daily basis. So like, you know, for someone to pull an Excel spreadsheet out of top 20 ASINs in category X with all of the bullet points links to the A plus B plus images and the content associated with all of that. So relatively basic and easy exercise to do.
Jack Lindberg (52:25.684)
Hehehe
Imteaz (52:49.768)
Um, but I, you know, I hear your point in terms of, uh, being very careful in terms of, I don’t think it’s just customer information. It’s more like sensitive commercial information being ingested into an LLM and then, you know, being aggregated and published elsewhere, but it’s more so. No, this stuff is here. It’s, it’s a race to find the use cases and the value. What risks are you willing to take?
Jack Lindberg (53:03.601)
Sure, yeah.
Imteaz (53:19.376)
in order to find that value. That’s the call that most corporations are trying to make today is do we go softly and you know be a follower rather than a leader within this space because the risks are too high or do we lean in very heavily to you know to find the use cases and to find the value ahead of everyone else.
Jack Lindberg (53:43.184)
Yeah, if I were to…
start a new company with my clone. I’d want to start a company that basically does LLM and AI implementation for enterprise. That all you do is you go in and say, hey, this LLM stuff is scary. Here’s how we do it with all your privacy concerns addressed. And you can make a lot of money that way.
Imteaz (54:04.12)
Yeah, and one, of course, and one easy approach is to, I say easy, it’s easier to say, I’m sure the technical people will tell me I’m crazy. But one, and one other way is basically masking and pseudo-anomizing your data, right? So if you’re uploading customer data, you can hash all of that out or make up names and then de-identify, keep all of the purchasing and transactional history, you know, consistent.
but pseudonymize that data so no one else can actually make any meaningful information out of it. But at least from running a model point of view, you can see the impact of, you know, your media activity, for example, against your user base. So there are lots of approaches to it. And it just comes down to how open are you to the risk? Yeah.
Jack Lindberg (54:57.724)
Yeah, for sure. I think.
Jack Lindberg (55:02.216)
The LLM and AI approach is going to be one of those things where two guys in a garage can build a $500 million company. And if you’re an existing CPG and two guys in a garage can build a bigger company than you have, you should be wary of not adopting this technology.
Imteaz (55:13.765)
Yep.
Imteaz (55:19.765)
Yup.
Imteaz (55:23.444)
Yeah, for sure. For sure. With your unique educational background in music and opera, how’s this influenced your approach to analytics and AI in this business context? And more specifically, how would you recommend someone with a background in e-commerce get started in specializing in the analytics space?
Jack Lindberg (55:45.872)
Yeah, I think what people think about product managers and analytics professionals and what they think about musicians. If you’ve ever seen those alignment charts of what I do versus what other people think I do, product managers and professional musicians have very similar charts. I think what most people think product managers do
Imteaz (56:04.608)
Thank you.
Imteaz (56:12.126)
Okay.
Jack Lindberg (56:15.992)
is they yell at software engineers and release really cool features. And then they take all the credit, and they feel very popular, and they wear a black turtleneck and stand in front of a giant projector. And I think most people think professional musicians perform. And they get up there, and they sing it on the huge stage, and they have adoring fans. What they actually do as a professional musician is you spend 90% of your time practicing alone.
Imteaz (56:37.607)
Yeah.
Jack Lindberg (56:45.744)
in a room by yourself. And that process of practicing is one that requires ruthless, ruthless precision to say, you’re the only person in this room that can give you feedback. How do you develop the skills of active listening and introspection and diagnosis to say, something went wrong.
Imteaz (56:48.104)
No.
Jack Lindberg (57:15.228)
What are the possible core things that I know are the inputs that make this thing happen? How do I diagnose this problem? How do I solve it on its own? And then realizing that music is an art form where it could never, ever be good enough, or it can never be perfect. It can only be good enough. And how do you basically self-diagnose in a way that you actually unsupervised build that muscle of self-improvement?
Jack Lindberg (57:47.509)
It also causes people to have what I would describe as either extremely strong or extremely weak egos. That process of like ruthless self-analysis breaks people. There’s only so much negative self-talk. You can say, oh, that was bad, oh, that was bad, before you’ve crushed yourself. And the folks who are doing it most successfully…
by and large have figured out how to compartmentalize in a way that says, when my singing isn’t going perfectly, it’s not because I’m a bad person. Um, and I know that sounds silly, but it’s very easy in our professional lives to conflate our professional success with our personal worth. And as a musician, you get really used to that. Like either being the thing that breaks your career.
or the thing that drives your success, because you’re the person who’s most able to go into the practice room and say, I’m going to build this better than it was yesterday. And it doesn’t matter if all my fans and my teachers and my mentors thought it was amazing. I know there’s incremental gain I can get. And being a product manager is very similar, right? You are the person who’s saying, the bar can be set externally. Like maybe I get…
I went from a zero to $100 million in ARR in a short time frame. And you need to be the person to tell yourself, self-critically, those accolades don’t mean anything. I’m the person who decides what good for me is, and I need to go out there and make that happen and feel that drive without the external measures of success being the driving factor. If you wanted to go and make money or if you wanted to go and get famous.
Imteaz (59:23.049)
Yep.
Jack Lindberg (59:38.728)
being a professional musician or a product manager are not the easiest ways. I think also what professional musicians do that I think we all need to take into business context is that level of practice and preparedness is industry standard. If you show up like I’ve seen happen,
If you show up to the first rehearsal, like what’s called first day of school, you show up to the first rehearsal of a gig and you’re not prepared, you will get fired. Like on the spot. And because if you’re singing, for example, some relatively common opera like Don Giovanni and you’re singing a role, you get fired because you don’t know your music.
There are 100 other people that are a phone call away that could fly in tomorrow and do it because they know the music. So the level of professionalism that you need to have just to get in the room is really high. I think we all sort of, we get busy in our business lives and we fly by the seat of our pants and we say, oh, I’ll just go to this meeting and sort of figure it out as I go.
Imteaz (01:00:46.258)
is very high.
Imteaz (01:01:01.589)
doesn’t work.
Jack Lindberg (01:01:02.088)
That’s something that we all need to do better at. As a professional musician, you get fired on the first day. You’ve met my old boss when I was at Pac-V, Riyad Iddu. And our first meeting together, when I was like the first one on one we had, he was like, hey, the number one rule we have as a team was no winging it. And as a musician, he and I are both musicians. He plays bass. I sing.
Imteaz (01:01:09.994)
Thank you.
Imteaz (01:01:25.793)
Mm-hmm.
Jack Lindberg (01:01:29.64)
We were like, OK, we would never show up to a gig and not know our music. So you would never show up in professional context and if not have done your homework. I just see this happen too much and this is something that I wish people did a little bit better job at.
Imteaz (01:01:46.188)
I love watching documentaries on superstar athletes and just superstars in general. One of my favorite ones is, I think it was in Jiro loves sushi, where he goes and sees this Japanese soba noodle chef who’s like the top soba noodle chef in Japan. And the guy has been at the top of his game for 20 plus years.
And when they’re interviewing him, he wakes up in the morning and every day he has a routine in terms of how he cuts his soba noodle. And he says, one day I will perfect my craft. And he’s been doing this exact same thing for 20 years straight. And when you see that level of diligence, discipline, and practicing your craft on a daily basis to make sure that you’re still at the tip of the spear.
It’s only commendable, right? Like, you want to operate with people who are at that level consistently all the time.
Jack Lindberg (01:02:56.616)
Yeah, it’s definitely something that seems non-concrete, but it’s definitely helped me transition from music to e-commerce and product. Because
As someone who’s relatively new to the industry, there’s no way I can know everything. I can be the most prepared. And that’s just something that you can just, that’s like a grind rather than a like innate skill. I think we like need to collectively reflect on like what things that are our own successes are due to innate talent versus effort. I think you’ll find that the majority of the time things are due to effort, not from innate talent.
Imteaz (01:03:21.46)
Of course.
Jack Lindberg (01:03:43.664)
Ahem.
Imteaz (01:03:43.712)
No, of course. I think it’s a balance between the two. And obviously having talent is a gift, but no amount of gifts can excuse hard work or make up for hard work.
Jack Lindberg (01:03:56.252)
Right. Yeah, so the second question you had, which is about how do you get into e-commerce and learn about analytics, it’s a lot of drinking from a fire hose, I must say. I think the places I would start would be, you can get by and do a lot, just like learning Excel and SQL and some dashboarding tool.
Um, like if you are a data analyst, that’s probably the majority of the skill set you’ll need to get moving. Um, if you want to get further down the road and start building models and Python or R or using some more advanced technologies, that’s definitely open to you. I think the first thing I would do is just like, go learn the basics of SQL and thinking about the basics of SQL.
I found has been, because I had to learn it from scratch too, was a really helpful way for me to conceptualize how data works and how it’s structured and how to access it. And thinking about, like, when I describe data to people, I accidentally frequently describe it as SQL in my brain. Ha ha ha. Because it’s just the way I’ve learned now to logic through, like,
Imteaz (01:05:18.74)
Hehehehe
Jack Lindberg (01:05:25.204)
How do I get the data I need from this thing? I think that will be the best place to start. Be kind to yourself. It’s not going to happen all at once. You’ll make mistakes. It’s not going to be perfect the first time. But just going out there and trying will be very useful. And if you have SQL skills, even intermediate SQL skills, you are much more valuable as an employee to any company because you can.
get your own data and figure out what it means without any help. So I would recommend that to anyone in any field if you have any data that you potentially could touch. SQL is a good skill to have.
Imteaz (01:06:07.32)
So based on your experiences, what emerging trends in AI do you foresee having the most impact, specifically on retail media and analytics in the near future?
Jack Lindberg (01:06:19.316)
I think the I think LLM search is going to be really powerful. I think we will really have to see whether the giants of traditional digital media being searched and social are going to basically self disrupt with these technologies. Like, is Google going to throw out?
AdWords to say our search is actually completely LLM driven. And my hunch is there are strong forces saying no, but the universe is saying yes. I think we’ll have to see, like, will the incumbents figure out how to self-disrupt before they are disrupted themselves? Because what we’re seeing in the industry is building an LLM is so expensive.
Imteaz (01:07:02.52)
Thank you.
Jack Lindberg (01:07:17.6)
favors incumbents to some degree. But the incumbents need to be willing to basically blow up the bridge that got them to where they are.
Imteaz (01:07:29.088)
or work out a financial model that still works for them, right? Like, it’s a balance between the two.
Jack Lindberg (01:07:31.762)
Right, right.
Yeah, I think that’s going to be really, really interesting to see how that unfolds. Basically.
Jack Lindberg (01:07:45.04)
I am hesitant to say it’ll happen, but I feel like if I were to place my bets on what’s going to happen, there will be only a few sort of incumbent technology providers that are operating towards this AGI sort of approach, and then a wide variety of vertical specific LLMs or AI tools that are saying,
I only know about XYZ type of data, and that’s what I’m going to focus on. Something to help lawyers write things for. For marketing, something that helps you write PDP copy is something you can get really specific and train a model to do better than any other model, but then it’s not used for everything else. I think we’re going to find those sorts of specialist models start to pop up that pick a very narrow sliver of what you can do within an e-commerce business and say, I’m just going to build that.
like sort of what I’ve built that says, I’m just going to write code for Amazon Marketing Cloud. Just pretty niche and say, I’m just going to nail that one. And I’m not going to worry about writing poetry or PDP content.
Imteaz (01:08:55.105)
So, on a personal front check, what habits and productivity hacks make your life easier?
Jack Lindberg (01:09:03.432)
Um, I spend a lot of time listening to podcasts. Every time I’m commuting or walking anywhere or in the gym or in the supermarket, I’m listening to a podcast. Uh, I find I don’t like sit still enough to read. Um, I’m just like listening to a podcast is how I’ve, I kept trying to learn new things and stay up to date. Um, definitely go listen to the other.
podcasts that MTS has been on or has published himself. They’re a good listen. Oh, I can shamelessly plug you.
Imteaz (01:09:38.316)
Thanks for the talk, man.
Imteaz (01:09:45.908)
Which other ones do you recommend?
Jack Lindberg (01:09:48.724)
The podcasts I really like, I like 20VC with Harry Stebbings. I like Lenny’s podcast, which is about product management. The Y Combinator team has a really good podcast. The A16Z team at Andreessen Harwis has a very good podcast. And then…
I also listen to a few political podcasts, like the All In podcast, which talks about politics and tech. And then I listen to Breaking Points pretty frequently as well.
Imteaz (01:10:20.086)
UK.
Imteaz (01:10:24.256)
Yep.
Imteaz (01:10:27.904)
Very cool. I have a book recommend- I know you don’t read to, you know, just find this book on Audible. But the book recommendation is The Hard Thing About Hard Things. It’s a book by Ben Horowitz from A to Z. And there’s a specific two, three pages in that book called Good Product Manager, Bad Product Manager. If you just Google Good Product Manager, Bad Product Manager, you’ll find it.
you’ll find the excerpt online on the A16 website. But it basically goes through with the good behaviors of a good product manager versus the bad behaviors of a bad product manager. And Ben Horowitz actually produced this document and gave it to, I think it was 16 odd product managers that he had reporting to him, half of which, whom he thought were ridiculous and people who was gonna fire.
And out of those eight odd people, I’m butchering the numbers, but out of the eight people, six of them read that document and understood what he was actually looking for as a manager and they improved dramatically. Two people he wanted to get rid of anyway and they didn’t get it and they didn’t change. But I think communicating what you want in a team is a very good thing that a lot of leaders don’t do very well.
So yeah, definitely I would recommend you have a look at that book and in particular that section. It’s a great read.
Jack Lindberg (01:12:04.8)
Thank you. I’ll definitely take a read.
Imteaz (01:12:09.389)
To close out, Jack, how can people reach out to you if they want to learn more?
Jack Lindberg (01:12:14.748)
Yeah, so best way to reach out to me is on LinkedIn. I’m posting there a couple of times a week about Amazon Marketing Cloud or Noctis, and updating people about what I’m thinking. You can also track down the Mars Agency at themarsagency.com or take a look at Noctis at analytics, analyticindex.com slash Noctis if you’d like to learn more and sign up for a trial and give it a whirl.
Imteaz (01:12:44.952)
Super cool. All right, thank you for coming on the show, Jack. Pleasure to have you any time, and all the best with all of the endeavors that you are currently on. Would love to have you on the show, let’s say in a year’s time, to hear developments on Noctis, developments on you, and then take it from there. Thank you so much.
Jack Lindberg (01:13:05.214)
That will be awesome. Thank you so much for having me. It’s been a blast.
Imteaz (01:13:09.068)
super cool. Thank you.
Other Links
Jack Lindberg Social:
Linkedin
Instagram
Watch the full Podcast
Subscribe to my newsletter
Listen to the full Podcast
My Social