Retail's "iPhone Moment" & How Intelligent Systems Are Detasking Stores | Julian Mills, Quorso CEO

Foreigning us now for five insightful minutes is Julian Mills, a frequent omnitalk guest and the CEO of Corso.

And Julian is here to share with us some of the key lessons he learned from the recent Intelligent Management forum he and Corso just hosted.

Julian, it's great to have you back on five insightful minutes.

Let's dive straight, straight in.

You brought together some of the top minds in retail at this forum.

What was the biggest aha moment that came out of the event for you?

We had about 30 SVPs, VPs from 20 of the largest grocers, convenience stores, apparel chains, etc.

Coming together really to talk about how data and AI can be used to guide and connect the daily work of everyone from the store associate up to the EVP stores.

There's a great group.

And in terms of the aha moment, I think one retail exec said it best when they said to me, this is kind of retail's iPhone moment in that for ages we've been spending time sending out hundreds of different kind of tasks and comms and walks, et cetera, to stores.

It's all been very kind of overwhelming.

It's all coming in different channels, et cetera.

And the stores hate it.

It doesn't move the needle and it can't frankly cost a bunch of money.

And actually where we're moving to is having more of an intelligent backbone that is personally prioritizing daily work for everyone in our business.

Julian, you mentioned to us too that one of the other themes that kind of follows along with that is that you're trying to de task the store, dive into that a little bit and what that actually looks like in practice.

I think there's a general sense that retailers push work to stores that may not always be very value adding.

Okay, so let me give you four examples.

So the first one is retailers send out a bunch of tasks that are just annoying.

Yeah, go submit your labor schedule.

Well, you know what?

I've been doing it every week for the last six months.

Okay, I can remember that.

Second one is they're sending out stuff that's repetitive through lots of different channels.

So one retailer we work with has nine different comms channels for the stores.

Guess what?

The task might get sent two or three times via different channels.

The third one is they're sending out tasks that can't necessarily be done.

So one retailer we work with sent out a task saying, please go and set up this pop.

And 90% of stores said, yes, we've done that.

And then a couple of days later the vendor sent an email saying, sorry, we haven't sent you the pop yet.

Okay.

Oh, wow.

So we're sending up tasks that can't be done, and people are wasting their time ticking off checklists saying, yes, I've done that.

And then the fourth thing.

And then, Chris, this will resonate, I think, a lot for you, is we're asking people to go and check stuff visually that you can check better using data or using data or potentially computer vision.

So dm, go check that these planograms are up to date.

Well, guess what?

The data can tell you that.

So why are you paying someone to walk around and check that?

Someone who could much better be spent spending their time coaching the team.

Yeah, so I think it's detasking is about trying to get rid of those types of work and to use data and exceptions and AI to basically focus people on the things that I personally, in my role at this particular store, need to do today.

Amen, brother.

I mean, like, yeah, that was one of my least favorite jobs, and I was a merchant too, and I love planographs, but that was one of my least favorite jobs because at the end of the day, there was more efficient uses of my time.

All right, so another term that's making its rounds across the industry is this idea of single.

Single pane of glass.

It's not new.

It's been around for a while.

But I'm curious, like, what's your perspective on that term in general and what does it actually mean in practice to you?

Concept of.

It is very appealing.

It's a single place where.

Where every employee can go, and it gives them just what they need to do to do that job.

Okay.

I don't think anyone has ever done it.

No.

Yeah.

So I don't think it exists.

Having said that, I think at Corso, we're about as far down that road for store leaders and for area leaders as anyone's gone.

Okay.

So we are bringing together all the work that or most of the work that they need to do, whether it's, you know, what historically would have been called a task or a walk or an audit or an alert or an exception or a ticket or a maintenance ticket or a customer callback.

All of those can be done in a single workflow that's driven by data in an intelligent way in Corso.

Having said that, there are lots of things we're not doing.

Like, you can't check your pay slips, and I don't think you ever would be able to do that.

Okay, so we Think, you know, you might be able to bring 70, 80% of it into a single pane of glass.

But the vision of having everything in one place for every role in the company, I think is aspirational.

We're talking about technology that's helping the store teams get smarter.

What do you think that means for the role of the field leaders?

How is that going to evolve?

This is changing very fast.

So what we're seeing is that as you use data and AI to push work and decision making and action taking down into the store, the role of the district leader is evolving and becoming more what I think what most district leaders would like, which is more of a kind of a coach and a person who's there to help when people really get stuck.

Yeah.

So if you think about it today, how does it work?

So, for example, a couple of weeks ago I was touring stores with a market director and he said, look, here are my 79 KPIs, here's all my dashboards.

I'm somehow meant to walk into the store and know that they got a problem with hams.

Yeah.

But if I can detect it's got a problem with hams, and I want perhaps the store to have fixed it before I get there.

And I should only really be there to help them if they're getting stuck on things that they don't know how to fix.

Yeah, right.

So I think that kind of diagnostic role is pushing down into the stores and the field leaders, becoming more of a coach and more of having a broader kind of more strategic role.

And I felt that pain every single day.

Like, you know, you'd be expected to go in and diagnose these problems.

And the part you said about too, like the dashboards.

The dashboards are just overwhelming the number of data and like.

Yeah.

You just can't possibly check the whole thing.

And so it just doesn't make sense and it's a lot of wasted energy.

But, you know, in the perfect segment, the segue of all time that I've always wanted to make, I want to go from ham to AI Julian.

So I want to close with this.

I want to get us out of here on this.

So, you know, AI, you, you mentioned it in that last statement, actually.

So, in all seriousness.

So, so, but the question is how far should retailers let AI optimize their store operations?

And, and, and how should they think about that in terms of the dichotomy of what still requires human oversight?

That's the question I want to talk to you about.

So what are you hearing from executives at this event.

What did they.

How do they think about balancing AI versus human interaction or human responsibility in the store?

So two years ago, if you'd asked me that question, I'd said most retailers are firmly towards the we must remain in control of everything.

So they probably battle one out of five on a scale of, you know, human control to AI running everything.

I think today we're about three and a half.

Okay.

So there's been a rapid shift.

Having said that, everyone at the event, when you present them the logical conclusions of doing absolutely everything through AI, felt that they weren't quite ready for that yet, or it might not be appropriate for that particular bit of the problem.

Let me give you a very specific example.

Corsa uses lots of different types of AI across the platform.

For one of the things we do, we essentially watch lots of operational data and trigger an alert when something goes wrong.

Okay.

For something like a product recall, they're saying we absolutely need to know that that product is being taken off the shelf.

So, you know, we want to use a more kind of deterministic model for that.

You know, we want a more rules based, machine learning approach to that.

But then there are other areas, for example, where we're much happier to have, you know, an LLM or the equivalent actually optimizing stuff for us.

So, for example, an LLM might be helpful in sifting through all our kind of SOPs and come up with a personalized plan on how to fix a particular issue.

Yeah, so I think what we're seeing is it's a very nuanced solution.

In some places you want a much more deterministic kind of rules based solution.

In some places ML is great, and in some places Genai can work magic, but you need to be doing all of them and bring them all together in one place and be very transparent around what you're doing.

Right.

So net net, there's no one right way.

You've always got to have a balance, which is, which is why we love talking to you, Julian.

I mean, you and the team at Corso do such a great job of thinking about the next level of where store operations is going.

So thank you for joining us today.

It was great pleasure.

Thank you for having me.