Mastering Product Leadership, AI, and Ad Tech with Neha Bansal

Introduction to Neha Bansal and Her Journey

Productside | 00:02.51–00:41.73

Hi, everyone. I am Nicole Tisch, your host of our segment, Trailblazing Women in Product Management. Today, our guest is Neha Bansal, a product leader with years of experience in ads and analytics at companies such as Google and now at Metta. Additionally, she invests and mentors in startups in Silicon Valley. And she’s even recently written a best-selling children’s book.

I’m very excited to have her with us today sharing this amazing experience she’s got and I can’t wait to hear her insights. So welcome Neha.

Neha Bansal | 00:41.73–00:45.41

Thank you, thank you so much, Nicole, for having me here with you today.

Productside | 00:45.41–01:01.164

Very exciting. I love hearing the varied experience that women have when they find their journey into product management. And I love starting with that question that how did you end up in product management?

Neha Bansal | 01:01.164–02:35.416

Well, that’s a long story. I’ll try to keep it very brief. So I grew up in India and I studied econ and statistics for my undergrad. I fell in love with statistics, especially. And then I decided to go implement my skills, newly founded skills, in a consulting firm which focused on analytics to help financial services build their strategies. Did that for five years. Absolutely loved my experience in consulting.

However, I did miss the joy of watching my strategies come to fruition by going through execution and launch. And that made me feel a little dissatisfied with what I was doing at that point. So like a lot of people, I went to business school to decide what else is out there for me where I could create more value. And that’s when during business school, I ended up interning in Tel Aviv in Israel with

a beverage analytics company called Weisberger. I joined their data science team with the intention to help them build some recommendation engines. And as I was building those models, I was like, I need to talk to customers to really understand what they care about so that I can build the right inputs for the model. And I decided to go talk to customers. And that’s when I ran into Noam, who was the head of product at the startup. And she was like, you’re the first.

person who was thinking customer first and is not in the product team. So yeah, and then she took me under her wings and we ended up building the model and then shipping it as a product. I got that first-hand experience working with her and Noam is now a VP at Slack. So she, I think I got trained under some of the best people in the industry. Really grateful for her, my time with her. And that’s when I just realized that that’s kind of what I want to do for the rest of my life. So I came back to Columbia.

Productside | 02:35.416–02:50.926

Really?

Productside | 02:50.926–03:01.002

Yeah.

Neha Bansal | 03:01.002–03:08.355

Interestingly, back then, business schools did not have a structured product management curriculum like they have now. So I found myself spending more time in the engineering school and the technology classes at the law school, et cetera. And I think that was fantastic. So second year, spent a lot of time upskilling myself to be able to interview in product management positions. And therefore, after business school,

Productside | 03:08.355–03:29.312

Right.

Neha Bansal | 03:29.312–03:33.496

I joined Google as a product manager, and the rest is history.

Productside | 03:33.496–04:01.25

Fantastic. What a varied experience, especially focused on the customer. I love to hear that. We always encourage product managers to have their finger on the pulse of that customer. I love hearing that that’s the door through which you got into product management. And so now that’s been some of your history. What about today? What is really inspiring you in product management today?

Inspiration in Product Management Today

Neha Bansal | 04:01.25–05:06.008

gosh, yeah, I feel like that today, like this last two years just has shaped product management in so many ways. So I’d say that if I had to really choose, there are three things that is inspiring to me on a daily basis when it comes to being a product leader. First is contributing to the field. So there is immense satisfaction that I derive from

being at the intersection of customer needs and technology that’s evolving and helping match the two. So for example, as AI agents and operators are becoming part of user workflows, I am at the helm of the industry thinking about how do I take that technology and make it easier for my customers who happen to be advertisers? How do I make their workflows much simpler and help them?

generate more value from the advertising dollars they are spending. So I think that is extremely inspiring for me. Second part is team building and organization, building organizations. I think I’ve always been like someone who enjoys mentoring people and helping people grow. I think it comes naturally to me also because I am the oldest of amongst my four siblings.

Productside | 05:06.008–05:25.1

Pee-oo!

Productside | 05:25.1–05:25.656

Wow.

Neha Bansal | 05:25.656–06:04.056

So it’s always been something I’ve always done that. So I think that’s why building teams and helping mentor other PMs into their careers, helping them grow, find their space in the industry, it’s always been very rewarding. And finally, continuous learning, as you can tell, and I know we’ll talk a lot about this today as well, that in this changing world, product management as a career requires you to continuously upskill yourself and figure out how can I

become better at my job, right? So that keeps me challenged and always on my toes. I absolutely love that about PMing.

Productside | 06:04.056–06:27.126

That’s so true. There’s always something evolving that you really need to continue your learning. I love how you are sharing that. Let’s talk more about the impact that you referenced. I would love to hear how you have seen your work impact maybe not just your customers, but the economy at large.

Neha Bansal | 06:27.126–06:30.254

I think it’s such a great question that rarely gets asked. let me think. So as you know, I am passionate about the ad space and the measurement space. And it’s evident through my work at some of the biggest ads companies, including Google and now at Meta. What keeps me excited about the products that I build, the work I do, is primarily two things. One is

Productside | 06:30.254–06:54.538

You

Impact of Product Management on Economy

Neha Bansal | 06:54.538–07:05.464

and creating jobs for millions of people out there through the ads business. And second is enabling businesses to grow. So let’s unpack that a little bit. When it comes to the first aspect, which is job creation, while at Google, my work supported creation of about $700-plus billion in economic activity for millions of American businesses, nonprofits, publishers.

Productside | 07:05.464–07:23.074

Hmm.

Neha Bansal | 07:23.074–07:50.83

creators and developers. now at Metta, Metta’s advertising technologies are reported to link to about $415 billion plus in annual economic activity that created 3.1 million jobs. And that’s just the power of advertising, And the second aspect is helping businesses grow. So more than 90 % of businesses that use the products me and my team builds are small businesses.

Productside | 07:50.83–07:52.222

Hmm.

Neha Bansal | 07:52.222–09:13.934

We, research that has been published by the National Bureau of Economic Research, has shown that for every dollar that’s spent on meta ads leads to about $3.31 in advertising revenue for our customers. and, you know, I just feel so proud of being the team that helps drive that number up into the North, right? And we do that by helping them find new customers, helping businesses serve existing customers better.

and also helping them free up their time from advertising so that they can do what they really love and care about. So that’s just, and this knowledge in the marketing industry has enabled me to not just create value for advertisers who are our customers, but also in my day-to-day work as well. So for example, through the learning site build as an operator in the industry,

allows me to also serve other startups as a mentor and invest in them so that they can find product market fit. This helps create jobs that these startups ultimately go and create in the economy. Yeah, I think those are just some of the ways that I feel very proud of what I do and it makes it exciting for me to wake up every morning and then say, okay, let’s go change the world.

Productside | 09:13.934–09:46.616

That’s great. I imagine, right? You have those statistics that not only help your clients, but certainly help you feel good about what you’re doing and seeing that impact. That’s really cool. Now, when you’re thinking about the ads industry and the user experience, you touched on this a little bit, but let’s talk some more about how you see that evolving as people move away from Google search.

to using ChatGBT. Where do you see that?

Evolving User Experience in Advertising

Neha Bansal | 09:46.616–10:00.941

Such a great question, Nicole. So I think what you’re really asking, let’s broaden that question a little bit. Read the question as, how does the advent of AI affect the ads business in general? So to give you a perspective, search ads is about 40 % of the total US ad spend. The next 30 % goes to display ads. And the remaining approximately 23 %

Productside | 10:00.941–10:16.246

Right, there we go.

Neha Bansal | 10:16.246–11:16.983

go to video ads, and the rest of the portion is like other smaller formats. So I’ll talk about how the top three formats are most affected right now and what I see as some of the trends and opportunities in this space. For search ads, we see reduced reliance on search engines, as you rightly mentioned, which decreases the effectiveness of search ads. And that’s a major source for a lot of companies, especially Google.

When it comes to display and video ads, people who are now using chatbots, they expect direct, helpful answers. So as people engage more and more with chatbots for keeping up with the news or interact with agents on individual websites for the information they’re looking, they do not want disruptive display ads or video ads. So they expect ads to kind of grow up and, you know,

Play along, like not be disruptive, be very, very helpful. So given this trend, I think that is one core mantra to keep in mind. And again, that’s not net new, which is focus on the user. Even more than ever before, ads need to be helpful and non-intrusive and therefore add value to the conversation rather than disrupting it. So with this,

Productside | 11:16.983–11:39.896

Ahem.

Neha Bansal | 11:39.896–14:01.056

In the ads ecosystem, think of it if we think of this in two primary two sides of the coin one side being people who see the ads at viewers and the other side being the advertisers who are paying for these ads I Think there are some Some areas where we will see a lot more innovation. So on the ad viewer side

there will be a lot of innovation in terms of new formats as well as personalization. So for example, when it comes to new formats, we will see, and we’re already seeing lots of non-intrusive formats. For example, within the chatbots, I’m already working with companies that are investing in building sponsored answers to the responses in the screen. There is going to be interactive product demos.

which you will see as response to a question, or even AI-powered brand representatives within the video that you are watching. So instead of banner ads, you would see recommended products or services in a conversational way. Companies such as Rembrandt, Sizzle, they’re all investing and doing a lot of innovation in this space. The other big aspect is

when it comes to making ads out of long-form content. So let’s say we have this podcast interview that you’re recording. We could use AI to just create a small ad out of this to help product site promote its products. And that could happen with AI tools now instead of having to record an entire new asset and pull up a production team for doing that. So those are some just new formats that I see coming up in the industry already.

The second big aspect is personalization. So as there is more and more compute available for much cheaper pricing, the ability to personalize ads is not limited to just fundamental things like demographics or location anymore. We are able to churn through massive data to be able to identify niche interests, behavioral patterns, purchase history.

AI’s Influence on Advertising

Neha Bansal | 14:01.056–14:37.656

and even predict future actions that a customer is likely to take. Just to give an example, right now, when we do, let’s say, a display video ad, the ad itself is very static. Every person sees the same ad, more or less. But now, a travel company can change the ad live as I’m seeing the ad based on my own travel history as well. So if I’ve been to Spain before, no need to show you.

the pictures of Spain in that ad anymore. So a lot of personalization is happening in that ways. Another big massive example is localization. Previously, you would have to record different types of ads based on languages and countries. But now you can use the same video. It’s so cheap to be able to enable lip syncing in different languages. So that just makes ads very, very personalized and relevant.

Productside | 14:37.656–14:57.836

Thank you.

Neha Bansal | 14:57.836–15:02.06

therefore driving more conversions for the advertisers.

Productside | 15:02.06–15:08.8

That’s very compelling. think that will be of interest to. Sorry, yes, good.

Neha Bansal | 15:08.8–15:19.202

Okay, one thing I feel like there are times when your video is lagging for me. So is are you seeing me okay, because it’s if

The Future of Product Management with AI

Productside | 15:19.202–15:29.368

Yes. And I think this is part of the recording. Yeah, it’s it’s it’s buffering. That’s the word I was thinking. It’s buffering, but it’s going to be super smooth. So there won’t be any lag. And honestly, like right now, we see both of us on screen. They’re when you’re talking, they’re going to be focusing on like you’re getting that’s what our marketing team does. They pull you up to be the full screen and then

Neha Bansal | 15:29.368–15:32.973

Mmm.

Neha Bansal | 15:32.973–15:48.514

get good.

Productside | 15:48.514–15:52.024

when I’m asking a question, they’ll show me or show us both together again. So it won’t look like it’s a lag. It’s just that it’s buffering, yeah.

Neha Bansal | 15:52.024–15:56.684

Mm-hmm.

Neha Bansal | 15:56.684–16:01.9

Got it. Sounds good. And am I talking too fast, too slow?

Productside | 16:01.9–16:04.664

No, it’s perfect. Yeah, it’s great. It’s great. I hope I’m responding and I hope I’m not distracting. And I’m like, wow, like, my gosh. I’m like, that’s amazing.

Neha Bansal | 16:04.664–16:11.352

Thank

Neha Bansal | 16:11.352–16:18.796

And how am I on the script reading? you feel I’m natural?

Productside | 16:18.796–16:26.936

No, you’re fine. I mean, it’s it’s fine. You’re engaged. You’re engaged. And it’s like text notes and then engages. Right. Like so it I think that it makes sense. That’s good.

Neha Bansal | 16:26.936–16:32.366

OK, cool. So now when we can pick up, I’m going to continue to the second part of the answer. So when I was giving you that answer, I said there are two sides of the coin. One side was the people who view ads, and the second side is the advertiser and the platform side. So maybe we can start with, like you say, hey, that was so much happening on the ad viewer side.

Productside | 16:32.366–16:37.026

I’m sorry, say that again? You want to do what?

Productside | 16:37.026–16:51.544

Yes, yes.

Neha Bansal | 16:51.544–16:56.334

What about the advertisers? How are they using Gen.ai? If you want, can maybe just queue that in because otherwise the answer is too long, I think.

Productside | 16:56.334–17:10.816

Okay, okay. Right, right, right. So that was about just like making sure I understand. So that was focusing on the ad viewers. Let’s talk more about the advertisers. How is Gen.ai impacting them? Okay, we’re good. Okay, that was very compelling for how that impacts the ad viewers.

Neha Bansal | 17:10.816–17:21.974

Yes, perfect.

Productside | 17:21.974–17:28.804

Now let’s talk about how Gen.ai is impacting the advertisers.

Neha Bansal | 17:28.804–19:42.656

gosh, yes. So AI is being utilized for making advertisers’ lives better for quite a while now. And now it’s just become more and more supercharged in a lot of ways, right? So let me talk about a few areas. First is that creation itself.

Now advertisers are able to use AI to help creation of the ad itself in multiple parts of it. For example, generating images and videos, testing multiple different types of captions, different targeting, and different angles. So gone are the days when it took months for a brand to create a video for an ad.

They are just able to use multiple platforms, such as agents and synthesia, in video, et cetera, for creating video ads at a much rapid pace. Secondly, ads ranking. Ads ranking is a huge part of making sure that advertisers are able to get the relevant ranking for their ads and at the right bidding and costs for them. So now, thanks to the advent of AI, we have a lot more.

compute and processing power to be able to invest, provide more signals that informs ads ranking. Another interesting pattern I’ve been observing is how companies have been using AI to ensure that ads appear in the right place. So for example, if you are selling travel gear, you don’t want your ad to appear in probably, let’s say, a book.

selling website. So you probably would like your ad to be on some travel website, et cetera. So we’re able to use AI to be able to match the context with the ad much better and faster. Fraud detection is huge. So AI helps us detect and prevent ad fraud, ensuring that ads are reaching genuine viewers and the success accuracy rates are much higher.

Encouraging Women in Product Management

Neha Bansal | 19:42.656–21:48.782

Another big aspect that my team and I specifically focus on is performance optimization. So AI is at the core of building most of the ads reporting products. There are products such as dynamic creative optimization that allows advertisers to test different types of creatives and combinations real time and then use that creative combination for generating better conversions. And obviously last but not the least,

Performance prediction is also becoming much better now because of all that Ventilist technology, where advertisers are able to predict that for every dollar that they will spend, how many conversions can they expect? So building these forecasting models have definitely added a lot of value for advertisers in general. So I think just to kind of put this together, that’s how I’ve been seeing AI.

affect both ad viewers as well as advertisers. And some more fun trends that I see and I think is going to be valuable for listeners here is, one, AI-generated virtual influencers. So I think currently where brands spend so much money finding influencers that align with the brand personas, I think to plug that gap, AI-generated influencers is going to become huge over time.

Second is AI-powered shoppable video ads, especially on connected TV. Previously, if you’re watching, let’s say, Netflix or on Roku or Hulu, you would see an ad and you would just have to sit and wait for the ad to finish. But now, you’ll be able to watch a video and the ad placement is very much in the video itself.

so that it’s not disruptive. And you can actually click on it and go shop the product that you are seeing in the ad. Yeah, so those are some trends I would say. And of course, with all of this innovation and scale comes a lot of risks. So authenticity, detecting bias, ensuring the highest user privacy is all going to require a lot of expertise and innovation to make sure we continue to provide value in a safe and private way.

Productside | 21:48.782–22:08.342

people.

Productside | 22:08.342–22:31.884

Right, those are all important aspects to keep track of as well. And we’ll be hard pressed to find an area of business that AI is not going to impact. Right. So specifically regarding product management, how do you think product management will evolve as a career path now given AI?

Neha Bansal | 22:31.884–23:53.668

Yeah, that’s an excellent question as well, Nicole.

Here are some trends that I see when it comes to

comes at the intersection of product management and GNI. So first is that the cost of reasoning and intelligence is going down. So if you think that I have, or if anyone thinks that, I have been to these fancy Ivy League colleges and I’m like the most intelligent person theoretically, that just cannot make the cut anymore. You have to be able to figure out that

What is the value add that you can provide beyond these reasoning models that are already out there? So just to give you an example, there was a research that showed that a chat GPT, and it’s just chat GPT, and compared with the doctors who were using chat GPT, chat GPT did much better at predicting the symptoms or the disease a person had. But the person just said, hey, I have ABC symptoms. Do I have cold or flu or something else?

Right. So there’s going to be the bar is much higher when it comes to intellect and reasoning. Now, second is.

Skills for Future Product Managers

Neha Bansal | 23:53.668–25:29.454

Now that many can build or fine tune an LLM at a much lower cost, which is what we saw with DeepSeek a few weeks ago, the real differentiator is deeply understanding what customers want. And then being passionate about removing that friction from their existing experience and then embedding AI seamlessly into their daily lives and meeting them where they are.

Customer focus is just going to become more and more big. Third, would say is that it’s something I love what Andrew Ng said recently. He’s the famous researcher and founder of deeplearning.ai. He said that writing software, especially prototypes, has become much cheaper. So this will lead to an increase in demand for people who can decide what to build.

An analogy that he shared was, imagine back in the day when there are two complementary goods, like cars and gasoline. So when the cost of cars went down, demand for gasoline went up significantly. So it’s the same concept. So that’s the cost of engineering is going down. The demand for product management, AI product management, is going to go up significantly.

of people who can decide what to build. Yeah. So I think the future for AI product management is pretty bright. I’m definitely on that side of the school. But I think there will be skills that will remain indispensable, such as people management, influencing, creative thinking, prioritization that folks need to become better at and build on top of existing tools.

Productside | 25:29.454–25:49.902

Great. Great.

Productside | 25:49.902–26:14.986

Great, yeah, let’s tap into that a little bit more, because I want to focus on regarding this evolution, how can we encourage and even celebrate women who are interested in really pursuing product management? Let’s talk more specifically about those skills that they can invest in building now to be prepared for a career in product management.

Neha Bansal | 26:14.986–26:24.11

Absolutely. let’s think of it in terms of the deliverables that are primarily expected from product managers. And then we’ll talk about how the path to that deliverable is shifting and what are some tools people, including women and men, can leverage to be able to build those skills. So if this part is clear that

Productside | 26:24.11–26:44.542

Thank you.

Neha Bansal | 26:44.542–27:39.758

user understanding and identifying pain points, active listening, people management is still going to be core. I don’t think machines can do better at that than humans, at least for some time. With that part clear, let’s talk about the deliverables. So some standard deliverables that PMs traditionally are expected to focus on is a vision, strategy, roadmap, and the PRDs that we write, like the product spec stocks.

I think first is the vision. My personal take is that gone is the time when you would think of a two to three year ahead vision because just the pace at which the industry and the market is changing, two to three years are just too far ahead. So while you can do that exercise, you need to know what is happening over next month, next quarter, and…

Productside | 27:39.758–27:43.138

Thank

Neha Bansal | 27:43.138–30:09.73

Teams need to be very agile as they plan for it. There is very limited room for people who are less adaptable to this change in changing and evolving space. So second is strategy. This is my favorite part of being a product manager. And I think we should spend some time thinking about how would building strategy, the crafting strategy change? Because even today, there is so much content out there.

around how do you build good strategies that can stand the test of time, market changing, et cetera. So my take is that the need for strategy is still going to be there very strongly. But PMs who leverage AI tools to make the, to craft their strategy and evolve it continuously are going to succeed. So what are some of the skills that PMs need?

First is writing. So while all the LLM platforms and ChargPT, et cetera, have shaped the way writing, job of writing, I do think that writing very clear strategies is not something ChargPT or Gemini or any of these tools are very good at. They’re probably OK at giving you generic guidance, but they don’t serve the business need as such.

So when I hire for people, I’m looking for folks who have very clear frameworks up their sleeves for drafting a clear strategy in less than four to five pages. So for example, and then they should be able to use AI tools to fill in the various pillars of the strategy. So for example, thinking about how do I do the research, they should be able to use AI for doing their deep research about the topic.

including understanding what’s the competitive landscape like, what was done in the past, what we could do better. So that’s important. Then being able to leverage internal dashboards and process massive amounts of user data using the analytical tools that are available now, I think that’s going to be huge. Because then a product manager who is able to do that does not need to depend on a data scientist or a product analyst.

Neha Bansal | 30:09.73–30:21.934

to help pull those numbers. That’s where I think I’m seeing a huge shift. So PMs are expected to operate as a data scientist as a huge part of their jobs now than before. And I would then also say that, yeah, so I think on the strategy part, those are some of the things. And then finally, obviously, you can use chat GPT and stuff to make what you have written a little bit more crisp and punchy.

Productside | 30:21.934–30:39.096

Mm. Mm.

Neha Bansal | 30:39.096–32:57.384

But I think the thinking has to all come from you still as a PM. So having those tools up your sleeve is going to be helpful. With that, let’s talk about roadmapping. So I just want to say people need to stop making two to three-year roadmaps. Market is just pointing at a far greater rate. Roadmaps need to be maybe a month or a quarter at the max. And they need to be re-evaluated every time you hit a metric goal. Or if you’ve gone through 50 % of the roadmap,

you need to think of, what needs to be different here, and especially looking ahead. So I think that’s definitely important. And I think one huge thing that helps here is prompt engineering. So this is where you learn how to give the right set of prompts to be able to use these LLM tools to work as your assistant.

I strongly advise PMs to build their own repository of prompts. It’s almost like your personal back in the day when we would have templates for everything. But now it’s about having a repository of prompts, knowing across which engine does which prompt work better so that you don’t rely on just one LLM model. And you’re building your knowledge by feeding in similar prompts into different LLM models and then having the best output to serve your needs.

And then within the roadmap part, want to, I do want to say that prioritization is a very critical skill. And this is going to continue to be an area that humans and PMs can excel in. So just to give an example, I was speaking with a friend and she talked about how even research in the AI space is bottlenecked by human. And it’s a management problem because let’s say someone who’s managing a team of researchers.

That is the human who has to decide how to allocate the compute power across different AI research projects. So you’re thinking about ROI on that investment. And that is something that humans are just so much better at doing because they are able to process information from so many different ways and make that decision. At least, I think, for the next year or so. And then the final deliverable. Let’s talk about PRDs.

Productside | 32:57.384–33:02.262

Yeah.

Neha Bansal | 33:02.262–34:27.18

I think this is the part that’s going to really get automated. writing PRDs, LLM tools can write like 80 % of it now. And the remaining 20 % is where you can make a difference based on your understanding of the user problem and having a lot of domain expertise. So what does it mean in terms of skills? I actually encourage engineers on my team to write specs to the extent possible because that allows

PMs on my team and myself to scale a lot more. And then we just help plug in the remaining 20 % with the expertise that’s needed. So this allows organizations to operate much more efficiently and therefore have a much more profitable engineering to PM ratios. we are obviously hiring. The hiring, as I can see in the industry, is a lot more of PMs who have domain expertise.

because we know they’ve built something like this before. They have failed at this probably. And then they have a lot of learnings to offer so that we can do it faster the next time. So those are some trends that I think that are going to stay. So the advice here for PMs, and especially women PMs looking for changes, build a domain expertise and see how you can continue to add value there versus being generalists.

Productside | 34:27.18–34:59.786

Right, right. And that agility, right, you were talking about that agility, adaptability, and still the passion for the customer has got to come through from product managers for sure. That’s not going to go anywhere. So I definitely still believe that. So that’s fantastic. This has been such a rich educational podcast today with you, Neha. Thank you so much for

being with us and for sharing your expertise and your journey and passion. I really appreciate it.

Neha Bansal | 34:59.786–35:04.204

Absolutely. Thank you so much for having me, Nicole. Wonderful conversation today.

Productside | 35:04.204–35:04.204

Yes, absolutely. And thank you to our listeners for tuning in. If you enjoyed today’s conversation, don’t forget to subscribe and stay tuned for more empowering stories from the world of product management. Until next time, I’m Nicole Tisch and this is Productside Stories.