“Marrying Sensory and Data Analysis to Create the Ultimate Retail Experience” with Jerry Chen | Ep 4
Updated: Mar 9
The Group CEO of USEN Southeast Asia, Jerry Chen, joins this episode of The Future of Retail Asia. He shares his experience with Jun and Imran of having to grow his business from the ground up to the one today that provides the full sensory experience in the retail sector. Jerry also conveys the importance of collecting data, crafting relevant hypotheses, and conducting appropriate tests to further enhance the customer experience through the senses.
Episode 4: “Marrying Sensory and Data Analysis to Create the Ultimate Retail Experience” with Jerry Chen
-- Start Transcript --
IMRAN: Hello, welcome back to The Future of Retail Asia. My name is Imran.
JUN: And I’m Jun. Joining us today is someone very interesting. I think he started off selling royalty-free music to retailers, but now is an expert in influencing customer behaviour through the senses.
IMRAN: He is none other than Jerry Chen, the group CEO of USEN South East Asia. Jerry started off in 2009 with Express IN Music, a music crowdsourcing platform for jingles and voiceovers. And then the business expanded to providing background music and other services for retailers, F&B and malls, around Southeast Asia and APAC. Welcome Jerry!
JERRY: Hey! Thanks for having me here.
JUN: I think that is where I personally very interested about what you are doing, Jerry. I still remember the first time we met, we were talking about what we are doing. Sorry to say that when the first time you mentioned that you were doing background music, my first impression is, is it the Spotify for B2B? After we discussed, then I realised that you are actually using music to try and influence people behaviour in the different scenarios. Very curious. How did you manage to get into this niche market? I mean, background music is everywhere, but using the music to influence the people, the consumer’s spending behaviour.
JERRY: I will start off by sharing, back then the very popular Spotify, Deezer, and KKBox coming into play back then. We were brainstorming what type of music company do we want to be? From there, we looked at, needs vs wants, B2C. Consumers may want to have certain ad-free music, but do they need to? Even for myself, in the music industry, I have been listening to ads, kind of, on Spotify, until very recently that I paid for my monthly fee.
JERRY: Finally! After some promotions and discounts. We understand that we cannot fight the big boys, like Spotify, Deezer, KKBox. Therefore, we actually looked into who else needs to have music, and what is the pain point that they have. And we focused on non-consumers, which are the B2B commercial part of things, and that’s where the F&B, the retailers, the shopping malls, hotels, that need to have music curated ad-free. That’s where I went into this business, and from there we actually grow exponentially.
IMRAN: Awesome. When you were growing, you shared a lot of stories with us before this conversation. You shared that, in the malls, you played different music and different locations. In the corridor vs toilet, and even in the car parks. You play music to make it safer there. Could you share with us the thinking or the practical theory behind the customer psychology?
JERRY: Every single part of the mall, the shopping mall is big right, so every part of it is very different. People go to toilet for, you know, purpose. We go to car parks for another purpose. Corridors at certain parts, certain levels, they are all segmentised. They’re already being categorised in our mind, in our way. For example, when you mentioned about car parks earlier on. What is the purpose of having background music, there can be a few. One of which some of the research papers has shown, it is for the safety, for security purposes, giving you the sense of security. In the Maslow's hierarchy of needs, that is one of the most basic. There is the science that shows if you have empty spaces that is too quiet, you will be paranoid. You don’t feel safe, you look around. Is there somebody going to attack me from out of nowhere? You watch too many of this type of dramas, TV shows, crimes happening, robbery and such. It could happen to you because that’s where the mind starts to wander. But when we have music that fills the void, that space, we are able to see how we are able, in terms of psychologically, calm ourselves down subconsciously, and then we feel safe. Once we feel safe, this mall is associated for example, to being a place that I would go to.
IMRAN: So you talk about security right, and I wanted to elaborate on that. Is it possible to use the music to actually then activate the wants? So security is basic, but now I want stuff. I want to feel comfortable, I want to look or feel a certain way. How does that activate that need or that desire to spend?
JERRY: When we talk about more than just music as part of our sensory perception. From your sight, in terms of mood lighting, warm lighting, bright. It can enhance certain types of perceptions. Example, when we have our meals at certain western restaurants. You say, “This place, kind of cheap la.” Because the utensils are kind of light. “Cheap, but then I pay so much. Not worth it la''. People say, “Bo hua”. So it’s all part of their perception that we create, what type of music ambience. Some people say, “Oh, pop songs. Because it is what people like to hear.'' But does that mean that it suits your entire perception and filling the value that you want to create as an outlook to your targeted audience. I know a lot of businesses say I want to hit the masses, everybody and anybody is the best. It is always that 20% that brings in that 80% of the profit margins. And that’s where you look into that niche audience and create that ambience from the specific ambience music, to the mood lighting, digital signages, and even fragrances that we want to portray. You cannot make everybody 100% happy, there will be that few X% that will be like “Ahh, I don’t really like your place.” Well, they are not your targeted audience then. So look into that 20 that brings you that 80% profit margin.
IMRAN: I think that brings a very interesting topic right. For today, it is the case where we are still under tightened measures, in retail. I think even over this weekend, there is a restaurateur that was sharing that their business was suffering very bad even though it is opened up. Partially it is because there is no music to create ambience. Is the customer experience today, in Covid context, in New Normal context, even more important than ever? The sensory experience?
JERRY: I think this sometimes when you lose something, then you realise that something is important. When it comes to this context, it’s about managing the best to our capabilities and also not just generally of music is good. People know that generally it’s that. But different times of the day, got different sets of audiences. So how do you set your time slots, how do you schedule it. Weekday crowds and weekend crowds is different. No music versus got music, there is a difference but what type of music makes that difference for your specific audience. And that is that punchline you need to look into. Now you say no music, it hurts businesses, like the Straits Times reported on the news. There is this research that I have with me, it’s quite a chunk out there. All this type of thick research paper is what our headquarters, USEN-Next Holding in Japan, actually curated and gathered sources from the world, from US, Europe, Asia. We also have our own group of researchers, scientists, professors that are looking at this. For example recently, we looked at lab studies that show low volume, not no, it’s low volume. Low volume actually increases sales towards healthy foods.
IMRAN: Low volume of …?
JERRY: Low volume of background music.
IMRAN: Ah… I see.
JERRY: And if you look into higher volume in contrast, unhealthy food. Even the certain type of pitch, tempo, will affect the way this food is being perceived. That is why when we eat on the plane, It’s kind of bland. But do note that the ambience sound is white noise. There is a lot of different type of, the food can be presented the same, but you eat at a different environment, at a different place, you feel differently. “Eh, this is so expensive.” “Ah.. This is so cheap, so value for money” Actually can be the same thing. Why and how do people perceive it, just like what you mentioned just now, the restaurant owner that commented. It actually plays an important part, but more so the very specific type of music or the environment that you want to create from mood lighting to digital signages. You can say I got digital signage, but I flashed what type of food like, “Oh this is my hamburger, this is my cheese fries.” That is A. If you do AB testing, for example, you will look into, “Oh, this is the way I describe my cheese fries. How juicy the beef patty is for my burger.” It flows out, it makes you just crave for more. Every bite makes you think of your first love. They way you describe it, itt makes you just want to have it so much, and this makes you more hungry. As compared to the A testing. So that is AB testing to understand better. So digital signage is just a medium, background music is just a medium. You can do it yourself or do you find an expert to curate that ambience for you so that you can tweak it so that the 20% is targeted to bring you that 80% profit margin. Otherwise, you would just be doing 80% of the time, the music that people play, herd mentality. And that is where you get only 20% of that margins.
JUN: Interesting right! In order to understand and implement the customer psychology, just now you mentioned about AB Testing, and not only for F&B, also for the retail space, supermarkets. We look at a lot of data today, camera, people counting systems, a lot of IoT sensors. And here at Aimazing, we can paint a picture of the customer’s transaction data. Do you think that Aimazing can help malls to improve their customers' experience with the retail transactional data.
JERRY: I think that brings me to my memory in army days. My encik always say, “You think, I think, who confirm?” So everyday we want to do certain things in a shopping mall, in a retail outlet, and then we say “Based on my experiences”, “Based on what I believe it is” is not based on true data. And that is why you get all these datasets in place, so that you are able to analyse and push hypothetical analysis, hypothesis out there. You are able to do AB testing in multiple datasets to prove a point, and you do it with a purpose and objective. That is how we are able to meet the first step that analysis the data came from. And then we will push out. Without that, it is all a guessing game. I think, you think, nobody can confirm it. Then you say I should have, I could have. Too bad. This is very important. And it’s step one! If you don’t have that, how are you even going to do step two or three.
IMRAN: So if we are able to go from that guessing game to the confirm game from the no testing to AB testing. How do you think we can merge this whole idea of consumer sensory, consumer psychology and with looking at retail transactional data. How do you think that malls and retailers play to win now?
JERRY: Let’s say in a mall, we have different floors of F&B. And F&B itself is also different types, got western, got ATAS one, premium, for masses. So we need to understand the crowd dynamics, the personalities that are involved. When I go there, how do I, as a consumer, hope to be perceived as. Or is there a specific interest? Typically, with data that is connected, you found out that at a certain time of day, there is a certain type of people that go to the cafe, there is a certain type of groups of people that go as a family, young family, multi-generational family and such, you go there for a different purpose. You capture the trend through such data, and from there we are able to say because of this data, I am able to give you certain types of mood lighting. Because warm lighting and the bright, white light, has a different type of feeling. Ambient music with a different type of voice over that targets, if we know ladies typically are more into white wine, as compared to men, red wine. It is also known that groups of guys come together, four, five guys come together and they drink a tower of beer. As compared to you trying to sell the one pint, a second pint, third pint. It is not as lucrative. It is known that guys in groups of four or five typically come down at 5.30pm to 7.20pm. You can always have a certain type of promotion to upsell, through voice overs, and this with a system, SOP processes, you are able to set that in the right tone. And then at 7.21pm onwards to closing at 10.30 for example, you set another type of ambience, because you know who is already going to come. You cannot be very certain 100%, but you are able to get a high probability of who and what you want to target. If there are more ladies, dessert menus on promotion, your digital signage will have a time schedule being put out there to seduce, attract, the right audience. Imagine those hunky guys after sports typically go to the cafe at certain times, please. People want to be very general and generic. At the end of the day, you are targeting no one, because 80% is targeting at 20%. Oh, by the way. This by the way is you playing on chance. What we talked about, guessing game. With what we mentioned with the solutions that is presented, the dataset, we are able to customise voiceovers, mood lighting, digital signage schedules. How appealing you want it to be to that target audience in a very targeted manner in a retail, F&B, shopping mall setting.
IMRAN: And not just at a static preset level. What you are saying is that even from a real time responsive level as well. Or even an AB level testing level on a daily basis.
JERRY: Yes, correct. It can be done in both ways.
JUN: Do you think there are any malls in the world that are currently able to measure the sales conversation of the mall traffic?
JERRY: I think, this one I cannot confirm. It could be in China. They are very much on big data, and they analyse and scrutinize further. In Singapore, not that I know of. I really hope that this part of the world in Southeast Asia can be more proactive because the amount of treasures and opportunities that they have missed is immense. Once you are able to have that information, it’s just like intel in the Army, that we were in. We know where to attack, when to attack. We cannot guarantee, there could be soldiers that do the guarding. “I spot you!” and then they fire. We do not. The chance of that happening is low. We have intel so as to target things better. We do not see our customers as enemies, but we want to see them as, how to optimise their experience. If you want to target everybody and anybody, then you are in a business for nobody.
JUN: Yes, exactly. I think this is what we should, and could create the first offering together, right?
JUN: Using the sound and music in the background, and immediately you can see the real time sales conversion rate, just for an example.
IMRAN: Before this, we were also having quite a number of discussions leading up to this discussion today. You were talking about an idea of activating a physical retail ad network. What is your idea with this, if we are able to capture the retail data? What is this idea of an ad network that we can do together?
JERRY: Currently we have almost all the supermarkets in Singapore. We have petrol stations. We have convenience stores, pharmacies in our network of clients for the BGM, background music, which have the function of scheduling in our advertisements. Example, we talk about SPC vs Exxon Mobil vs Shell. These are all petrol kiosks, but they are not similar, when it comes to targeting very niche audiences. It is understanding their niche customers, from like Sheng Siong vs Fairprice vs Haomart. Or even in this case, let’s talk about Fairprice. Fairprice already got their Cheers Convenience, Fairprice normal ones, they got Finest, and they got Xtra. These are all very, very different niches. They have different people who go to Xtra, for example in Exxon Mobil, they do it fast. Grab and go. They know what they want. Or maybe it’s just opportunistic. So the target messages from the voice overs at a certain time, at a certain day of the week, will be narrowed down to that audience. So you have a better chance of activating them. But how do we know it? Once again, it’s from data, and not just the guesstimates.
IMRAN: Right, right. Where we could actually play the music, or the ad all throughout the network. And we can actually measure how much SKUs are actually moved.
IMRAN: That’s very interesting.
JERRY: I would like to share an example from our head office in Tokyo. There was this Christmas season back then and they were selling log cakes. If I am not wrong, it was 120 stores in Tokyo, Japan, around Japan countrywide. Half of them we did AB testing. Voice Overs were streamed in to promote Christmas log cakes. The other half, no such promotions. Those that have promotions, have an increase of more than 20% in sales. Out of which of those people who buy, more than 90% did not claim that they consciously notice the voiceover, but they buy.
JERRY: Why did they buy? Because we made them to. In that manner that we want to create that right ambience sound. The right message. And when it feels right, it’s just part of the experience. t are very irrational when it comes to retail therapy, when it comes to food even! Now we buy things at the spur of the moment, it’s that feeling that we always talk about. It is so subtle that it can only be connected with data with the triggering mechanism, like the voiceovers, the digital signages, the fragrances, the scent, the smell, that will bring you the result that you desire, so that you can elevate that purchase decision.
IMRAN: Right, so there is a lot of complex thinking and systems behind it. But in it’s best form, the customer thinks it’s my own idea to buy.
JERRY: The feeling, that you entice them with that experience.
JUN: Very interesting, I think in the past few episodes. We are talking about data collection from the tenants for the GTO, rental calculation. After that we talk about loyalty programmes, how to provide a good user experience for a shopper to shop across a shopping mall. Today we actually bring in one more, background music and sound, to improve the shopper experience, even in the mall. Just now you mentioned about the different petrol stations that have a different target audience. So this can also be implemented in different shopping malls right? I believe that CBD shopping malls will be very different from the neighbourhood one. And how can they use the different background music or sound to provide their shopper journey more smooth, or have a better shopper experience.
IMRAN: Are there any interesting local stories around that?
JERRY: Too many! I will share a few. For example, if you were to go Orchard. Orchard has a lot of shopping malls, a lot of shopping malls are out there. When you go to Orchard, Wah very atas. There is a different level of that ATAS-ness, the premium, the level is seen very differently because they have different target audience. From Paragon to Cathay to even Plaza Singapura, PS itself. This is very very different. You need to understand the entire customer journey, it’s like a circle. When they want to engage, it can be an O2O that a lot of shopping malls want to use. When it comes to O2O strategy, you always talk about the online world is getting so much datasets because this particular person got cash involved. You want to understand if he or she used to buy a certain thing. In the customer journey, he or she experience it in a different way. This colour is more appealing to this particular user online. So how is it being applied to on the offline world, is also this entire customer journey, how you engage them from that beginning where you may not want to go into a shopping mall, but because of that fragrance, it lures you in, it just brings you in and when it brings you in, you just enter it, and feel so relaxed, because of certain scents. And when you walk in, there is a voice over promotion that is targeted at you at 2.30pm. Where it says that there is a one-for-one coffee break promotion. And it’s only for that one hour, 2.30pm to 3.30pm. And you are there at 3.25pm, you only have 5 minutes.
IMRAN: What would you do?
JERRY: What would you do, right? So you go in, you buy a coffee, you settle down and it is said that during that lull period in the afternoon, let’s say 3+ 4pm, where there is not a lot of transaction and traffic going on. There could be a type of collaboration between that cafe and let’s say a retail apparel brand. With your receipt that you buy, you just show the coffee receipt and go to a retailer, get 20% off or buy 1, 50% for the next pair of shoes, the next sets of clothes that you want to purchase. That would then be a collaboration, and then that is the entire customer journey. You buy the clothes, you buy the coffee. In the first place, you may not even want to enter that shopping mall. But it is that scent, that voice over, you sit down then that physical receipt. Might as well buy, impulse buy. You buy already, you go out already. You look! Actually, these pants looks not bad. Upsell. Because the digital signage shows that type of clothing that you may be enticed to buy. That’s where we talk about artificial intelligence, machine learning, where camera analytics see that you are dressed in a certain colour. You are dressed in a certain clothing type. Might be a businessman, from a collared shirt, or darker shade of pants. The digital signage can push to you a darker shade of pants, navy blue pants and black pants. They won’t push to you white pants, for example. That is how our machine learning, or artificial intelligence, gathered through data. That is also another form of data. When it rolled out, there is a higher chance for you to buy that particular pants.
JERRY: And there was no promotion for pants! You buy! And when you walk out at the end of the journey, when you scan, when you checkout. Your information is captured. This is where, what we termed as the online strategy we have always been using, capturing your data, can be applied offline in a retail shopping mall, in a retail setting, F&B altogether encompassing this, to know what is the dynamics, what is the movement of our audiences. Because it is always changing, like with Covid times, lockdown, no lockdown, lockdown how long? We do not know until all this comes into a form of datasets for the machine to learn and apply and push out the higher probability for our customers to buy. And when they leave, you create a lingering classical conditioning, of them wanting to be back for more.
IMRAN: Very interesting, because you touch upon the other layer of the ad network with some of the others which is, the receipt transactional layer. There is the sensory layer you talked about, the voice overs and the display, and then there is the purchase of the coffee and here is the receipt offer, which is the SKU of the pants, or the clothing, which is that second layer which is enabled on transactional data. I think this is super interesting and I wish we could carry on. But we do have a few closing questions. I think these are difficult times, and I think one of the things that people have to balance is, I have so many priorities, what do I do now? If you had a mall manager in front of you today, what is your one piece of advice for him or her? And if you had a retail manager today in front of you, what is your one piece of advice for him or her? So one piece of advice each.
JERRY: For the mall manager, or the management level, I will really urge to focus on the bigger picture. Be it your KPI such as your profit margins, your revenue size, or your walk in rates. You can never satisfy everybody, because the ground level will say, for example, I do not like this type of music. Certain person complain about the fragrance. It is smelly. Things like that. Are you going to be because of one person, or a very small group, that dislikes certain things, and you forsake the big picture? You can never, ever satisfy everybody and everyone. For mall managers level, management, that is something that you need to have the data that solidly tell you that most people, majority, or this is the set that will give you the most optimum results.
As for the ground level, it’s to get lots of feedback. Because you are on the ground, you will be able to understand the customers in a very intimate manner. And that conversation flow, that entire service level, you have to feed such information back. Because management, the managers, they may not be able to see what is happening on the ground. With such information, just like any parts of data, it’s good to be feedback to the management, to take a more decisive decision on what is the best outcome that could possibly be achieved.
IMRAN: So on the retailer lever, so we talked about malls, and on the retailer level, any piece of advice?
JERRY: We have been speaking about O2O for a very long time. If the online world is able to do with the data captured from the cash, from the memory and such. And from that data of that user, so can we do it on the offline manner. Just like people say about our ad network. We actually look into how we want to poise ourselves as the offline google for advertisements in the supermarkets, convenience stores, and in the petrol stations. We want to provide that data through camera analytics and push out data or information to the right target audience. Once we have that ability to have open-mindedness of how such data, artificial intelligence, machine learning, can be a form of condition, so that the triggering mechanisms can be automated from mood lighting, digital signage display, and even fragrance. Because we do know what the personality is. So that scent, that smell, it encompasess the whole experience. A lot of people say online is the way to go, e-commerce, m-commerce. I agree, and I also want to state that it is not as cheap as what people think it is. Competition is very, very strong. It is also very costly from the advertisement point of view.
JERRY: You have to push out a lot of e-commerce efforts. It’s not that cheap as compared to your offline stores and retail outlets. And offline, you have one biggest advantage. Experience. Online cannot give you that physical feeling. That ambience. It is not able to give you. You receive things online, “Eh, how come my XS is actually M size.” Because you never know. Fitting different. The shoes or the clothes, the feeling, the texture. OR the food that I ordered. It doesn’t come across that it’s of my expectations. So offline has that capability. So make full use of it. Don’t just become like an online shop. “This is my platform, I just display it.” You want it, you take it. Might as well go online and buy. You already have the physical experience. So enhance it further, with your fragrance, scent. Your digital signage. Your background music. In a smart manner. Data, condition, trigger. Set it as a schedule and you will be able to optimise it further.
JUN: Lastly, do you think retail is king? If not, how can it be king?
JERRY: I thought cash is king. King, queen, prince, princess, is not for us so much to define. Each to its own perception of how we want to manage it. I think we look at (malls) as a kingdom. It is a whole ecosystem, where retail needs to depend on the F&B, F&B needs to depend on the shopping malls system. How they walk, the pathway. All this type of things is all an entire process. The customer journey. As we walk through from place to place. How we integrate, how we collaborate, how retail needs to work with the F&B, F&B working with the retailers out there. I think it is an entire ecosystem. So I would term it as a kingdom, that we are able to encompass our sensory experience, our datasets. Piecing them all together and with your expertise with our specialisation in sensory. We are able to bring this platform and say that this is the place to go. Come to the kingdom, which is the mall, in that manner. Otherwise, people will just end up shopping online, and then getting the wrong things, feeling frustrated. I’m trying to push physical retail offline shopping, that’s why.
IMRAN: It is very cool because we had some guest previously that said that we were the Google of the offline retail, and I think it is just like you said. It is a kingdom with multiple ecosystems, and there will be a few players that will be kind of like, putting together the piece. It is very interesting, a very nuanced picture that you are putting together about stepping back, seeing the big picture with the data. I think that is all the time we have for today, thanks so much Jerry for joining us today. Very insightful for all of us. Listeners, if you have any questions for Jerry, please reach out to us. You can actually comment on our post on LinkedIn. You can even tag Jerry and I’m sure that if you have any questions for him, he will answer. So Jerry, again, thank you so much.
JERRY: Thank you so much for having me here.
IMRAN: We are on Linkedin, Facebook, Instagram, Spotify. And we will see you in the next episode.
-- End Transcript --