Ocado & Online Grocery Economics

Marc, can you provide a short introduction to your background, please?

I lead grocery ecommerce fulfilment at The Partnering Group, a retail and CPG global consultancy helping, currently, mainly US grocers on their online fulfilment strategy and also to optimize the efficiency of their fulfilment. Prior to that, I had my own consultancy, focused mainly on European food retailers, helping them with online fulfilment strategy and execution. Prior to that, all the way back in 2013, I was a financial analyst for quite a long time – probably slightly too long – all around the world, with the previous 10 years focusing on Europe and the US.

How, exactly, have you been working with grocers moving online, over the last few years?

There are two things. Firstly, there is the strategy of how they want to do the physical picking of the orders, which is the most time-consuming and costly part. That mainly involves optimizing the manual processes and it’s a mixture of processes, technology and people management, to try to get them towards best-in-, which is UK levels of picking efficiency, per hour. In other words, around 130 to 140 units per hour, fully loaded; that’s all inclusive of all the preparation and the staging, not just the physical picking. Just as a benchmark, in the US, the average is around 40 UPH, so there is a lot to be done there.

Also looking at if, how and when to bring in automation. Looking at micro fulfilment solutions, both from existing providers, like Dematic, KNAPP or now with Takeoff, or the more next-generation systems, like that from Alert Innovation, which is working with Walmart, and Fabric and AutoStore. Of course, Ocado is a key part of that, so benchmarking against Ocado which, currently, has the most experience in doing automated fulfilment for grocery, both for itself and for its partner Morrisons and other partners. Looking at whether it makes sense to choose Ocado, as a third party, or to build their own CFC in-house or to test a micro fulfilment solution provider; that’s all part of it.

I have also consulted for a couple of those previously mentioned, automated MFC providers.

What are the decision-making criteria, for the grocers, when choosing between those solutions?

The first question is, is it worth us even looking at automation yet? Probably only the bigger ones think it’s actually worth running the test. You’ve got to have the resources and the people to take that sort of risk because it hasn’t been done. Micro fulfilment, for grocery, has not yet been proven, in terms of the fact that we don’t have the full numbers as to exactly how efficient it can be. Clearly, if you are running at 40 units per hour, like in the US, you are going to get a significant uplift, straightaway. But it’s still in the testing and optimization phase.

Then it’s looking at which solution. Are we going to go with the tried and tested ones, where we know the multi-shuttle systems hardware works, but we don’t know how that works when you scale it down to levels that we haven’t been to before, such as $30 million sales capacity a year? Previously, they had a minimum of $100 - 300 million. Are they going to manage that themselves, by just buying that hardware from the system provider? Are they going to bring in a systems integrator? Are they going to use Takeoff, which has put the software layer into the integration? Or are they going to look at the centralized fulfilment model, from Ocado? There are pros and cons to each.

Because it is a very unknown area, some of the decisions are not necessarily based on pure logic. Humans are not pure logical creatures. Sometimes, whoever is in charge argues, or very strongly feels, that this solution is the one they should go for. Maybe it’s more strategic, so there is a strategic element. Do we want to be very aggressive, to gain share, or do we want to just test and see? There is a whole bunch of criteria depending on the retailer, what their general strategy is, what their risk tolerance is and what their resources are.

What is your view of the Ocado OSP solution?

Ocado’s solution is an extremely good picking system. It can pick from a very broad range of 40,000 to 50,000 SKUs and does it very efficiently and is proven, in terms of the quality of service they provide to customers. It’s also end to end. It’s not just the picking, but it’s the front end and the routing software and people management. They do a fantastic job. If you look at the customer service metrics for both the Ocado brand and also Morrisons Online, they are far ahead of their UK store-based competitors, who are generally picking from stores. You get a better range; you’ve got a much better inventory control and better freshness. It provides a very good service.

But the question is then the cost. At what cost? Who is it suitable for? In my view, for a third party, it is really only suitable if your store base is relatively limited; if your existing share, in the markets where you are going to put your Ocado CFCs is very limited. That means that you will have a good chance, if you have a strong brand. For Marks & Spencer or Waitrose, it makes a lot of sense. Also, for Monoprix in France, it seems to make a lot of sense. They have a limited footprint, they have a very strong brand and, probably, there is a lot of demand, if only people could have access to their products, so they can get a lot of incremental share.

But I really feel it’s not the right solution at all for a major grocer who has significant share in their markets, where they are going to put the Ocado CFCs. The reason for that is, when you are building a dedicated fulfilment center, effectively, you are adding a double running cost center. Whatever sales that would have been fulfilled in your store are, basically, taken out of the store and allocated to that CFC, to cover the running costs. But it means you deleverage your store sales and hence your store profits. Let’s say you’ve taken 10% of sales away from the store and those online orders are now being fulfilled in Ocado, you’ve deleveraged that store’s sales by 10%. Given the fixed cost leverage you have in grocery, your EBIT, your operating profit is mechanically down between 30% and 40%.

In order to offset the cannibalization of your existing store-based sales and profits, you need a very, very high level of incremental share from the CFC. Additionally, you need the CFC to be at least as profitable as your stores, which is not yet fully proven on an EBIT basis. It’s a very, very aggressive strategy. I think it’s more aggressive and risky than a lot of people seem to understand because of the point about deleveraging existing stores. If it was all incremental, fine. But it’s not going to be. We’ve seen that, actually, it’s very difficult. Tesco, in the UK, went very aggressively to become the market leader, online. They used to have about 33% offline share and, in the early to mid-noughties, it went very hard to gain market leadership online and it did very well. They had a share of well north of 50%. But as your competitors react and they start fighting hard to either keep their high-spending, multi-channel customer or steal yours, that incrementality comes down, pretty fast. Now, Tesco online share is not that much higher than its offline share. Your natural customers tend to come back to you.

It’s very difficult to sustain a high level of market share. Initially, you get very excited as you get lots of incremental share and, obviously, retailers love gaining market share. But that tends not to be sustainable and the double running costs of that CFC are going to be offsetting any market share gains and also the fact that you’ve deleveraged your stores.

The other point is, a lot of what you gain in picking efficiency, you lose in delivery cost and time. It’s a centralized fulfilment operation but the cheapest and quickest way to deliver is, obviously, from your stores because you are as close to the customer as you can be. Once you take it away, you’ve got much longer delivery routes and higher costs, so a good portion of what you’ve gained in picking efficiency, you can lose in delivery costs. Also, it’s more difficult to do same day, if you think that is important.

Potentially, the only way it could work is if you have a grocer that has an existing store footprint that moves into a new region or new area, where it is incremental to the existing store base, or they don’t even have that store base in the first place?

Exactly. Often, when you are trying to move into a totally different market, your brand name is less well known. Obviously, you can’t differentiate on service as much as you would in an in-store environment, but you will probably have a broader range than most and you might stand some chance of getting decent share in that market. That’s the only way it can make sense to me.

But don’t all the partners or customers of Ocado today – Kroger, Casino – surely they understand the deleveraging aspect of their store base?