The Biggest AI Mistake? Waiting for Perfect Data | 5IM

Foreign.

Joining us now for five insightful minutes is Tom Cooper.

Tom Cooper is the CEO of Duvo and he's here to give us an update on what he's seeing in the market in terms of retailers abilities to adapt AI solutions to their organizational benefit.

Tom, how you doing?

Yeah, I'm good.

Good to see you again, Chris.

How are you?

You too.

Yeah, no, it's been.

I'm really excited about this interview and we are also shop Europe too in the speakers lounge.

So we're coming at you fresh and bright eyed and bushy tailed on this Wednesday morning.

All right, Tom, let's start with this.

I haven't stopped talking about our interview.

You know, the one we did back in March, the one we did earlier this year.

So remind our audience of who you are and what it is that Duvo does first.

Yeah.

So I'm Tom and I'm the founder of great European retailer 2 billion and delivering amazing food to people where I've learned that.

And what is the name of that company?

It's called Rohlik.

I mean, like nobody knows that some.

Random $2 billion European retailer.

But I've seen so much pain, you know, when I observed people working for that company.

You know, it's essentially endless hell of emails and Excel files and dashboards and Chrome tabs and SAP.

So I also co founded Duvo and Duvo solves all of that.

Right.

So we identify processes that are automatable in companies, we suggest how to optimize them and we just run them after that.

Right?

That's what Duvo does.

Yeah.

And AI is actually making that possible.

And that's what stood out to me when we spoke back in March.

Because my big takeaway was that, you know, you help to take away the least glamorous tasks in retail.

The things that people either forget to do or don't want to do.

And why is that such an important element of what it is that Duvo's doing?

Because, you know, ultimately this is where value is not created.

Right.

And I would assume from what we observed that at least half of the time super smart people spend time on is this is being human API between systems that don't talk to each other.

Right, Right.

And I think that that's why, because if you, if you think about this, if you double the capacity of your smart people, how much better retailer you could actually be.

Right.

And then if you think about it further, what if the AI could do things that you don't do today at all?

Right.

You know, so many processes get overlooked so many things don't do or so many processes are not run at scale.

You know, you maybe look at the biggest fires, but not all of the cases.

So like, I think this is why it's important.

It just, it just gives you the unlimited workforce in a way, both on a human side, but also on the AI side.

Yeah, that's interesting too.

You said we don't create value from that work, but we leak a lot of value from that work, I imagine too.

Right.

All right, so I'm curious too.

I wanted to ask you this.

So founder, retailer, tech company, what vantage point does that give you relative to how the marketplace is operating in terms of how they're thinking about AI?

I think ultimately I'm not naive, right.

Like running AI in production at large scale, where you actually rely on the AI not to break the company.

I think that's something you have to understand from inside.

It's not like, hey, hey, I have a couple agents, do you want to kind of deploy them in the company?

It may work, may not, let's see, that's the posture that I see a lot in the market.

Let's do a poc.

Great.

But ultimately, if you deploy AI on your most critical business processes, it needs to run 10,000 times per day, super reliably for years.

Right.

And I think that's just very unique vantage point.

Also, there's a clear advantage understanding that nobody wants to change their systems.

Right.

So it's kind of.

I see a lot of startups, right, hey, so just change your system.

I have this new erp, and your problems will be gone.

And I'm like, that's gonna cost me $50 million and two years of time.

No, thank you.

Right.

So it's kind of just like we always knew we have to work on top of existing systems, Right.

Because once you talk about some SAP integration change of erp, no sane enterprise will touch you.

Yeah.

So I think that's also fairly unique understanding.

Yeah.

That's funny because I was going to ask.

That was going to be.

My next question is, okay, you have that unique kind of blend of roles here.

And I was curious what advantage it plays on the software side too.

And I think you hit it is like you have an understanding of what you're actually going to be able to sell into a retailer.

I think there's one more.

Okay,.

I understand there's still gonna be people in the company.

Right.

What I see a lot of tech companies kind of assuming is, hey, you know, guess what, I'm just gonna automate your 20 billion enterprise with agents, so you can get rid of everybody.

And that's funny.

Right?

Right.

And if you have that understanding, there's still gonna be people.

Right.

You have to build both for people and agents.

Right.

And you're trying to create the environment where people are becoming almost like agent operators.

Right.

As opposed to replacing everybody or automating everything.

And that's, I think, also driven by knowing the company, like, retail company from inside.

You know full well when you're inside, you're not going to automate everything.

Yeah.

And that's what I learned from you the first time, particularly in the first wave of AI.

Like, the first wave of AI to do this.

Right.

Is going to make your people better at what they do.

And that's the right way to think about it first, versus, like, going at it from, like, how do I just reduce headcount with deployment of AI?

It's really about how do you make your people do more work more efficiently to find the value that they're not creating and also the value that they're leaking.

Back to the point we said at the beginning.

Yeah.

And maybe deploy AI on processes you've never run because you didn't have enough people.

Right.

But I think retail is inherently quite efficient.

I mean, we're leaking tons of value.

Right.

But everybody has been through decades of optimizations.

Right.

Like, the margin is so thin that there is super hard to find if you go to banking.

I mean, if you have 3,000 people more than you should have, like, the P and L is not gonna really know that.

Right, Right.

But in retail, that's the difference between making a profit and making a loss.

So I think inherently trying to find ROI in headcount savings is a false pretense here.

Yeah, well, that's really interesting, too.

Yeah.

And the other thing about retail, too, that I think is fascinating is just there's so many touch points and so many different things that are happening.

Cause people that are new to retail always say that to me.

Like, I had no idea how complicated it is, because it is.

It's probably one of the most complicated industries when you think about production, getting things to the shelf and getting things home to the customer.

All right, so on that note, the last question I'm gonna ask you, and because we were talking beforehand, and I know you guys are getting some good traction, you've got some big clients already underway.

What is the biggest mistake you're seeing clients make in regards to trying to implement the ideas that you're putting in front of them?

I think Ultimately, when, when we're involved.

Yeah.

The mistakes don't happen.

Okay, Right.

But, but what I, what I see retailers getting wrong is, is doing these, you know, 10 year strategies, you know, two year data cleanups, you know, and, and the world is moving too fast.

Yeah.

Like, you know, by the time you're done with your 10 year old strategy, you're going to be left behind.

And I think that that's the biggest mistake.

Like there is a technology very capable today of doing what you need.

Right.

Yes, it's important to have strategy, but like the technology will be different in two years.

Right.

So you're going to redo that anyway.

So maybe let's do one year strategy, two year strategy, you know, let's figure out tools for the job right now and then take it from there.

I think that's the biggest mistake also AI because it can reason doesn't need such a clean data as maybe the previous wave of technology.

Right.

So I think it's kind of like every retailer is six months away from clean database.

Right.

Like never gonna happen.

I'm tired of talking about it, to be honest.

You don't need that.

You don't need that with current technology and it's just holding you back if you think about it.

So I think that's also a huge mistake that I'm seeing.

So when you go into those meetings and you're pitching this to folks, do you have to kind of rewire their brains around that?

Particularly around the time horizon or are people starting to get that?

I think our pitch is in a way fairly unique because what we say is, what's your messiest process you really hate?

And then we show them how we automate it in like three, four weeks.

Okay.

And that rewires the brain.

Like talking about things.

Don't rewire the brain.

Like seeing is believing.

And that's what we are trying to show people in retail management.

Okay, awesome.

Well, Tom, thank you.

Thanks for making time out of your busy schedule at Shop Talk Europe.

You're on stage later, right?

Yeah, yeah, yeah.

Excited and looking forward.

Yeah.

Can't wait to see what they're saying.

Thanks for having me.

Yeah, great.

Anytime.

Welcome back.

Anytime, man.

Best of luck.

Thank you, thank you.