In this 5 Insightful Minutes episode, Rita Kerbaj, Chief Strategy Officer at the Rohlik Group, joins Omni Talk to discuss the evolving economics of online grocery and what retailers need to do to build a profitable e-grocery business for the long term.

Drawing from Rohlik Group's success operating one of Europe's leading online grocery businesses, Rita explains why the industry's early struggles with profitability stemmed from inefficient fulfillment, small basket sizes, and costly last-mile operations. She also shares how retailers can leverage automation, fulfillment centers, and integrated omnichannel strategies to meet growing customer expectations while improving unit economics and building sustainable long-term growth.

Key Topics Covered:

• Why online grocery was historically considered structurally unprofitable

• How larger basket sizes and dedicated fulfillment networks improve unit economics

• Why same-day grocery delivery can be profitable when supported by the right infrastructure

• The role of dark stores, micro-fulfillment centers, and automated fulfillment operations

• When store-based picking still makes sense and when retailers should move beyond it

• Why order density and demand concentration are critical to profitability

• How Rohlik Group built a successful online grocery business without physical stores

• The importance of creating a unified omnichannel operating model

• How AI, automation, and connected systems are transforming grocery fulfillment

• Why the future winners in grocery will optimize both customer experience and operational efficiency

P.S. To explore more conversations from the OmniTalk Five Insightful Minutes Series, head here.

Music by hooksounds.com

*Sponsored Content*

Foreign.

Joining us for today's five insightful minute segment is Rita Karbage.

Rita is the Chief Strategy Officer at the Rolla Group, as I know many omnitalk fans and listeners are already aware.

And she is here to talk with us about the current state and the evolving economics of online grocery and what it really takes to make the model work the long term.

Rita, let's get you started off with this question.

Online grocery, I mean I've heard it a lot.

It's been called structurally unprofitable for years now.

But what has actually changed of late and what still does need to change for the math to actually work?

Yeah, that label was real painful, but real.

We're talking about small baskets, manual store picking, inefficient last mile.

And that meant grocers were losing somewhere between five and ten dollars on every order.

Nobody wants a business like that.

The operators really making it work today are the ones who have shifted to a full basket, really delivering a weekly shop at over $100 as an average basket.

And they've rebuilt their whole infrastructure around online.

So they've rebuilt compact, high throughput fulfillment centers with the right level of automation.

We're talking about 500 to 100,000 square feet that are close to where the customers are with concentrated demand and not these oversized monuments that we saw in the early days.

These are really the winners who have changed this discipline and stopped promising everything everywhere for free to their customers and started to rebuild their networks around something that really works both for the customer promise and the unit economics.

So Rita, I'm curious because we've been seeing same day grocery delivery, particularly over here in the States.

I think it's even more so in Europe.

It's quickly becoming the default grocery expectation.

And I can think back to just in the past month alone we've seen big announcements from Walmart and Amazon to that end and they're doing it by way of quote unquote, the dark store concept, which is the fulfillment center that is, you know, close to the consumer locally where the, where they know the demand is going to be.

And to me, you all at Veloc and Roelic are the proof point that that model actually works.

So, so how do you, how do you think about that?

Is that a requirement, the dark store approach?

Is that a requirement to making this model work in the long term to get to same day?

I'm a firm believer in same day.

For a long time people said customers don't necessarily care about same day, but they actually do and they want it as short as three hours or even 30 minutes these days in the U.S. it can be profitable.

As you said, you're thinking about density and demand.

So really building those sites very, very efficiently around where the customers are and automating everything from picking to, to labor management to how the route is distributed for the last mile.

And it can really work.

The promise itself isn't the problem, it's how you execute it.

Yeah.

And to get, to get the efficiency of same day right.

They don't need to be necessarily located in the store, which is where the industry at least over here was trying to go initially.

Because you guys don't have stores, you know, at the roll it group and you're doing it successfully.

Absolutely.

We have 12 different fulfillment centers.

We don't have any physical stores.

We've really built the model around online and this is where the grocers who win will do it.

It's about really building the model around delivery and not trying to repurpose your infrastructure from the store to deliver online.

So let's roll it back a little bit.

Let's talk about store based picking.

Does that model still make sense in certain instances and how do you decide when you need to go away from it and create a dedicated fulfillment center like we were just discussing?

I think there's still a case for store based picking when you're talking about an online penetration that's still slow.

When you have good store coverage, you have spare capacity in that store and you're promising next day delivery or delivery in 8, 10 hours.

Reusing that store space and your staff and your stock can be really a low entry point to online.

You start to need more capacity when you're exceeding about 300 orders per day.

When you're hitting about 500, you have to start thinking about MFCs because what happens is your physical shoppers start to compete with your pickers.

And that's about experience for everyone.

The numbers start to really make sense when you go over 500 and that's when you can really improve your UPH.

You can really, really reduce your picking labor hours.

You can significantly improve your cost per order as well.

Rita, is it all about order volume too?

Or does the control of the customer journey also factor into when you think you need to kind of unplug from the third party platforms too?

The way that we've thought about it at Rolik Group is we've built everything around the customer proposition and this is what drives the service that we're creating at Veloc.

Right.

If we're promising the customer same day.

We need to build an infrastructure that delivers on same day.

If you're a physical grocery store, you need to be thinking about the in store experience for your customer just as well as the online experience for those customers.

If you want to win in the future, those channels have to work side by side and not in silos.

You have to think about the entire customer journey because your omnichannel has to be one operating system and they can't compete.

The minute they start competing, you're losing the customer both in store and online.

All right, well, let's get you out here on this then.

I'm curious, what do you think is going to define the winners in online grocery over the next decade?

Chris, Online grocery is not going away.

So we need to really start thinking as omnichannel as one operating system and not three silos.

You know, shoppers are already mixing online within store.

So winners really have to think about that entire customer experience and start optimizing for both.

Everyone's talking about AI these days, right?

We've talked about fulfillment centers and automation for years.

But what we need to think about right now is how all of those systems interact together.

So from demand planning to replenishment to labor planning to routing, to robotic picking, to price and promo optimization, to the last mile and the routing, all of those have to play together so that you're delivering on your promise.

And if that promise is same day, then that's what you're optimizing for.

I think that the winners really will have made clear choices on how they use their stores, where they add dark stores, where they build an FC and what that looks like.

I think that's a great nugget to end on because you're saying there's the infrastructural question and then there's also the software side of this that is going to become even more important to make the model work.

And it all comes back to what you said at the throughout this discussion, which is you have to build the model that meets the demand, not of just of your customer today, but of where your customer is going in the future.

So thank you so much, Rita.

That was excellent.

Thank you so much for having me, Chris.