NetElixir’s Udayan Bose on Performance AI That Drives Revenue | eTail West 2026

Hello, welcome.

This is Omnitalk Retail.

I'm Anne Mazinga.

We are coming to you live from etail west 2026.

And I am so excited because the reason that we get to bring you all of this exceptional coverage here of the show is because of the gentleman standing next to me.

Next to me is Udayne Bose, the founder and CEO of Net Elixir.

Udayn, thank you so much for helping us bring all of our coverage to our audience at home.

Thank you so much for sort of allowing us to do this.

Well, I'm so excited.

Every year, Uday is the mayor, the unofficial mayor of Etail.

He has been to every detail that has no.

Since 2007.

Since.

So this is our 19th detail west and probably 30th overall detail.

Yes.

Okay.

Well, let's start by sharing with our audience what Net Elixir does and a little bit about what you oversee in your role there.

Okay.

Amazing.

So netalxin is an AI first digital marketing agency.

So we are trying to really redefine as to how the performance marketing agency world should look like in the future.

We have really built our entire suite of AI marketing solutions to work with our human experts who can really.

So the goal is human expertise plus AI should be able to drive some exceptional results.

So that's essentially what we are trying to do.

So, again, AI plus humans will always beat out humans without AI or only AI.

So that's our claim.

I think that makes it.

That makes a lot of sense and it breaks it down very simply for our audience.

So thank you.

You've been coming since 2007.

Udayn.

Yeah.

What's happening?

What do you think is going to be, like, top of mind for people at this year's conference?

And what have you been hearing from your clients?

What are they focused on?

I know just by saying 2007, and I know I'm dating myself a lot, but I've seen all the different phases.

2007, just for information, when we.

Our first show, iPhone had not been launched yet.

Can you believe it?

I mean, it seems like another world altogether.

No, I cannot.

So from there to where we are, where we are talking about finally AI delivering outcomes and measurable outcomes.

I think it's a huge, huge shift which we have gone through.

So essentially, I think the key buzz out here at the event has entirely been not only just AI, but how can really AI be utilized to deliver some tangible, quantifiable, measurable business outcomes.

So that's something which is so refreshing because the last three Tales we have been talking about.

I think initially AI was a buzzword, then last year AI used to be the next glitzy thing and it was more theoretical talk.

But this time I am really so excited to see a lot of people asking that show me an AI which delivers results and outcomes.

So that's I think the best part.

Odin, what are some of those outcomes that your clients have seen or that you've been hearing other retailers and brands talk about that they've seen?

And what's your advice for them?

If they haven't started on their, their AI journey yet, how would you recommend they begin that?

I think the first thing is just started now.

It's still definitely not late by any means as such.

Yeah.

And just to simplify all the AI in marketing part, and I would break it down very simplistically into two parts of AI.

One is the efficiency AI and the second part is performance AI.

Okay, so efficiency AI is all the gen AI tools and what you really build on top of that to solve problems.

Problems like maybe smarter reporting, campaign management, agenda creation and so on and so forth.

Right.

Like for example, at Net Elixir we have built a platform called LXR GPT which literally has reduced our overall team's campaign management time manual time by about 30% already.

We expect it to go to about 50%.

But the other part is performance AI whereby essentially you're talking about can we use machine learning and predictive analytics to drive quantifiable measurable business outcomes by identifying opportunities which in normal campaign management or normally we don't as visible as such.

And what does the ROI look like?

Like what, what can happen if you deploy the tactics that you're advising retailers do.

See, I think on the, on the journey or let us say the, the Alexa GPT front, we have seen already a 30% time saving.

Right.

That's a very quantifiable thing.

But that time has been reinvested in identifying new areas of opportunity.

I think it really has gotten diverted into more strategic areas and that's where our Alexa Insights, which is our tool, which essentially is more of an AI marketing experimentation platform.

That is the reason.

And we have been wearing these lab coats.

You're in a lab coat.

We are experimenters to identifying new opportunities where you can really find revenues.

And that has been really a game changer because we are able to identify revenue opportunities which normally in regular campaign management, on meta, Google, even email marketing, there was no way for us to do that.

So the ROI there has been close to 20x for the performance efficiency component on the gen AI or the Alexa GPT component has been just a saving of time and huge increase in efficiency boost.

Okay.

One of the things that you mentioned and why you're wearing this lab coat is about experimentation.

Y How should a retailer approach experimentation?

Because it can be costly and it can be, you know, you, you may fail, it's higher, a very high likelihood that you could fail.

So how are you advising that retailers take the first step into experimenting?

Yeah, I think the first part essentially is really identifying what your business outcomes or problem says.

Right.

And then identifying a very simple question.

I am really trying to define a campaign, let us say a campaign on Google Meta and so on in a certain way.

Right.

Within the campaign management.

Yeah.

What may I be missing out, which the regular campaign management is not?

So I think just first is acknowledging that you may be missing out on a bigger opportunity is the first step.

The second part is identifying hypothesis.

For example, we had a tremendously successful run with a company called CarParts.com.

yes.

They sell obviously car parts.

They had about 40 million SKUs that they sell normally.

40 million SKUs out of which about 3 million SKUs did not sell for the last two years.

Now there is an inventory carrying cost which is quite ginormous.

Absolutely.

Yeah.

So their challenge to us was, let us see how many of these SKUs, dormant SKUs, are you able to reactivate?

So we set up those campaigns on Google Ads using, and then really fed them Alexa insights powered audiences, the high quality audiences or the high value audiences.

And in just four months we were able to reactivate 1100 SKUs that they had not sold in 24 months.

That's just insane.

Right?

It generated in just in four months a billion dollar extra revenue for them.

In a span of 12 months, the same new customers, we got about 30% new customers out of that.

They sort of went ahead and spent about two and a half million dollars, the ones just by reactivating these dormant skills which otherwise regular campaign management was unable to do for the last two years.

And yeah, you can't find those without tools like AI supporting your business, giving your data analysts the information that they need to make the next step in the next decision.

All right, Udane, let's close by talking about something that you mentioned, performance AI.

Now what is your advice for retailers as they're trying to figure out how to apply this in their business today?

Absolutely.

So I think there has been very clear message being sent by all the channels like Google and Meta and even Amazon that they will be implementing AI to take off all the manual work or campaign management work or the work that marketers have really been used to doing for the last many years now.

Right.

So understandably those.

And they essentially package it almost like as a black box.

Now the only way to crack the black box because the moment you tamper with their algorithm, let us say Performance Max or Advantage plus, it really sort of the performance gets impacted.

Right.

So essentially what we are seeing is the only way that you can really tap into the power beyond just what Google and Meta sort of really provides to you is through the experimentation part.

So Alexa Insights, what we do is pretty straightforward.

We use the first party data, we really run our algorithms, the machine learning and predictive analytics algorithms to really create cohorts like who's likely to buy from you in the next two months, which of the specific customers are likely to churn in the next 12 months.

And then we are able to inject these audience into Google and Meta and get them to find lookalikes.

Yeah.

So you're slowly expanding the quality of customer that you are able to acquire through Google and Meta through smarter experimentation.

Right.

So the plan should be to not only personalize your, you know, your experiences for your consumers, but now you're able to personalize who you're targeting using these tools.

Absolutely.

This performance AI that you're targeting.

Absolutely, absolutely.

I think look beyond because there are opportunities have really exploded and AI really helps us to tap into these opportunities.

My big message would that be that utilize these AI tools, AI will not be only Geni.

Like I am a huge advocate for machine learning.

Right.

And deep learning.

So those are the old school AIs, but the impact that they can have working with your first party data is just ginormous.

And I mean we are seeing transformations.

I have shared with you the car part story.

Yeah.

I mean those guys were so happy.

Think of the impact on the business.

Right.

Well, again, is there, I have to ask too, is there anything that you would recommend that people do not invest in right now when it comes to AI tools?

Or like, have you seen, have you seen things that retailers or brands that you've worked with have tried that didn't work that you would say?

I would avoid that listener.

So I would think one of the trends, I think some of you may already sort of know about it, that there are a lot of applications that mushroomed last year actually at the same event, and this year I'm seeing some of them as well, which essentially slapping a different skin or an application layer on top of the core LLMs.

Now, if you really look at it, if you have been following the news, Claude has been systematically removing all those application layers.

So don't jump in just because that's the next glitzy gem thing, but jump in because it really is meaningful, different for your business and can really help you scale and grow.

So that would be my request because I feel at the end of it, the LLMs, whether it is ChatGPT or Claude, will actually build those applications very quickly or someone will and suddenly that tool that you had really invested in goes for a toss because it doesn't really add any extra value and you are ending paying up for it, right?

Yeah.

Okay.

Uday, thank you so much.

If people want to learn more about how to smartly experiment with AI, how to see their business transform with the digital agency of the future, how do they get in touch with you and Netalix?

Come over at booth number 501.

That's I think the first starter.

We'll have a mad scientist coming up soon in about the next 10 minutes.

Okay.

So I think sort of come to us at the booth and talk to us.

501.

We also are sponsoring a wine and lunch session today afternoon and I know it's already pretty booked, but do come over and we will try to sort of get you in the room for the wine and lunch session.

We we are also hosting a private client only dinner.

So if we have about three seats available and if some of you are available, feel free to come over but through an introduction preferably and we'll try to absolutely get you to meet with our clients.

The clients essentially are the reason why we are here.

And just one last claim, a little bit of a brag I'll end with because this is not for me, for my team.

Yes, we are actually the so called the performance marketing or AI digital agency with the highest NPA score net promoter score in the US of 84.4.

Wow.

Prior to us there has been no other agency which has gone beyond 80.

So I think it gives the client love that I'm extremely thankful for.

Many of the clients are here today, so thank you very much.

Opportunity again.

If people are listening at home, what's the best way for them to get in touch with you after the show?

I would think just come to our website or just email me uden u D a y n et alexa.com any t elixir com and we are pretty quick in terms of getting back to you.

I can attest to that.

They are great again in the team.

Thank you so much for helping us bring all of our coverage to those listening at home.

Such a pleasure.

And until the next interview, stay tuned.