Nexo spoke with Daphne Stopforth, Chilled Solutions Manager at for Lucozade Ribena Suntory (LRS), about her experiences investing in connected coolers.
Daphne, thank you very much for talking to us today. Perhaps we could start with you telling us a little bit about your role at LRS and what you are trying to achieve.
My job title is Chilled Solutions Manager for Lucozade Ribena Suntory (LRS); and I lead the team responsible for the equipment solutions that support the business in selling our drinks. One of my responsibilities is making sure that we have great solutions available for our Sales teams. This includes looking at innovation and how we can incorporate innovation to ensure our equipment solutions are best in class.
Everything that we do has to deliver return on investment (ROI), that’s a big focus for us. We have already introduced significant innovation into our vending estate and LRS is ahead of its key competitors with the vending solutions that we offer.
Now it’s time to turn our attention to chillers and how we can improve what we do there. ROI on chillers is very difficult to measure, primarily because there is less accessible data.
Our number one objective is to have a better understanding of our ROI, but we also need to understand how we can improve asset utilization through better knowledge of what is going on in the channels, in the locations of our chillers and inside the chiller itself.
Asset tracking is another significant factor. We know that we lose chillers, but we don’t fully understand the scale of this. Being able to understand this better is another advantage.
Are there particular hurdles for LRS that have driven you to investigate gathering this information from your chillers? What type of information do you feel is currently missing about the ROI from these chillers?
We simply can’t measure very much currently. There are certain locations where we can, such as where we receive EPOS data. For those we are able to measure the impact that a chiller has.
However, many of our chillers are in locations where we have no access to that type of data. It might be an independent convenience store, or perhaps a food service outlet, and there is no data available that enables us to measure impact.
Our only option, in these cases, is to work on assumptions to see whether or not we believe these chillers are going to meet our hurdle rates, and hope that we have it right.
This new solution will provide us with much better data. We can start to use actual data to help us determine which chillers are performing and which aren’t. We can then take action to improve underperforming chillers.
With vending, we’ve had telemetry data for about 10 years. We have an excellent handle on how our vending estate performs and we can take action to improve performance where we need to. This solution will help us move in the same direction with our chillers.
We’ve seen various different types of connected cooler solution deployed across the world. LRS have chosen to go for an ‘Always On’ solution combined with an image recognition solution. Can you tell us why you decided to take this approach and then perhaps discuss what you intend to do with that very rich set of information that you will have access to?
The image recognition piece, for me, is the critical part of the solution. Without it, there is still a significant amount of information, but the image recognition gives you so much more. You can literally see what is going on in the chiller.
Indicative rates of sale will give us a strong indication of how many bottles have been sold, but not which bottles those are. The ability to see exactly what is in that chiller tells us how compliant it is for our products, and therefore if it is actually LRS bottles being sold. Without it we could be making rate of sale assumptions based on sales that are not even from our products.
The camera solution doesn’t just give us the compliance and purity information. It also tells us how well stocked the chiller is, and interestingly it also provides an indication of the location of the chiller.
During the trial, we were able to identify that one of our chillers was in a store stock room. We could tell from the picture that it wasn’t on the shop floor, it was just being used as a back-up storage chiller! It’s actually quite amazing what you can find out from the images that come through.
The camera makes the quality of the information so much more robust. One of the things we haven’t seen yet, but have been told should be possible is the indicative rate of sale by SKU. This will be incredibly powerful, we are looking forward to seeing that.
Another significant aspect of the solution is the predictive breakdown element. One of our other key objectives is to drive ‘up time’ of the chiller. We are measuring up time now, but we can currently only measure it from when a customer phones in to tell us ‘my chiller has broken down’. What we don’t know is how long the chiller had been out of action before they picked up the phone. This solution will give us visibility of whether the chiller is working and whether there is any risk of break-down, in which case we can send somebody to attend to it straight away.
I am pretty sure that a chiller appearing in someone’s stock room is an all too painfully common story for many of our customers!
Considering that you are deploying quite a complex technical solution I would imagine that you have had to engage a range of stakeholders in the initiative. One of the more common problems our customers raise with us, is that they aren’t sure who to involve in deploying such a solution. What approach did you take to making sure that this solution had the support of everybody internally and what did it require?
This was a challenge, because there are so many people to involve and there are a number of aspects we had to consider. Internally within LRS, we initially engaged with the Sales teams and Finance.
For the Sales teams, we needed to demonstrate how the solution could help them do their jobs more effectively. We needed to get them excited about the innovation and technology, but also illustrate the potential pay-back, which is difficult to measure before the solution has been deployed.
I can’t, at this stage, say I am 100% certain what the pay back will be on this, but we created our model based on the trial and that was strong enough for us to put forward the case justifying the cost and to get sign off to move forward.
The Category and Shopper teams have also been involved, because they can see that they will get additional data to help them understand better what is going on in store. The Data & Analytics team are now getting fully involved too and thinking about how they can use the information.
The only other people we’ve engaged, at this point, are our external equipment partner. From a technical perspective they have a massive role to play in setting this up for us and managing the technical side, as well as getting the customer masterdata to Nexo and responding to some of the suite of alerts.
Once we are further along in the process and have more data, we’ll be able to share more with other parts of the business, such as Marketing, to see how it can help them too.
Because everyone is so busy, it can be difficult to get enough of everybody’s time to really get their engagement and their input. Even me, I have had so many different projects that, in hindsight, I didn’t give this as much time and focus as would have been ideal to move the project on as quickly as we could have. So time is probably the biggest challenge, but people in the business are very excited about this.
That moves us quite nicely onto another question. You have been on a considerable journey to get it to this stage. What do you know now about deploying this kind of solution that you didn’t when you started; that perhaps would be valuable for anyone setting out on their journey now?
This is definitely a question to ask me again in six months!
My initial thoughts are that I have found it more difficult than I thought it would be. I am not a technical person, so I think there has been something of a language barrier. I don’t speak ‘techy’! Some of the emails and documents have been quite technical and difficult to translate into layman’s terms.
We have found that retrofitting to our existing coolers is quite challenging and more costly than we had anticipated. As we move forward and buy new coolers this will be less of an issue, because the solution will be fitted at source.
Also, we are now at the stage where we are working through the implementation of the suite of alerts. Unfortunately, the solution doesn’t appear to be quite as flexible as we had been led to believe. Plus, the various Sales teams want different things. Consequently, conversations have turned more towards ‘We’ll have to check on how easy it is to do and what is involved in doing it’, rather than setting these up quickly and easily.
I’m afraid, in terms of what I didn’t know then, compared to what I know now, it’s mainly negatives. However, I think if you ask me six months from now, or in twelve months, then there will be more positives, because we’ll be so much further along the process, and more established, and may even be identifying things that we didn’t expect.
Whilst positives are nice to hear, your honest feedback is important to us, because then we know what we need to do to improve.
Let’s go back to a previous question, where you had finished by talking about the various teams that were going to be involved with the deployment, from Sales to Data & Analytics, to the Shopper teams, etc. There are obviously a wide range of uses where you can put the solution to work. Considering all the different teams, how do you intend to make the data work within the organization?
This is one of the things we have been looking at with the Data & Analytics team in particular; the best way to use and share the data. We don’t yet have a complete answer to this, but we know it will allow us to make better informed decisions to help shape our strategy.
Another thing we’re thinking about already is how we could use the whole solution in the future. We are already starting to think about how we can extend the use in subsequent phases, such as use of the i-Beacon and potentially adding further digital add-ons.
Let me wrap-up with a final question. You talked earlier about the business case not yet being realized, but that you have modelled it. I would be interested to know how you have dealt with those who are skeptical about rolling out this kind of solution. Would you have any advice to people who face negativity about deploying such a solution in their own organization?
In the overall context, the costs are not massive. I believe the best option is to trial it, as we did with you. That way you soon get an idea of, or can at least model, an indicative business case.
I know we have spoken about it already, but the image recognition for me, is one of the key pieces – the visibility of the chiller and the contents of the chiller. The camera and the connectivity are an extra cost, but it’s cheaper than regularly sending someone to look at what is going on in that chiller. Simply modelling against that could demonstrate a saving. Although you still need people for follow-up actions, you can use your people resource more effectively by replacing some of what they do with these cameras and directing their activity based on the alerts.