RBR, Intel & meldCX: Turn data stories into actionable insights

Published on
June 8, 2021
No items found.

This is a recap of RBR's Self-Service Banking 2021 webinar that took place on the 25-27 May. Brought to you by RBR, Intel & meldCX. For more information on upcoming events, please sign up to our newsletter or follow us on LinkedIn.

Banking businesses spend big money on digital signage & in-store advertising but have very little to no visibility on how it is performing.

Through Viana™ Audience Measurement & Content Effectiveness platform, Westpac Banking Corporation is able to improve customer engagement in their digital content strategy — tracking pieces of content and optimising its delivery time, length and placement.



Gillian Shaw (RBR): Hello, my name is Gillian Shaw. Welcome to this session of self-service banking 2021. I'm delighted to be joined by Intel's Dr. SB Ngoo, or SB for short. SB is the General Manager of the banking vertical market and responsible for IoTG business for banking worldwide. 

It's also a great pleasure to be joined by Thor Turrecha, EVP of Global Software as a Service at meldCX. SB and Thor, a very warm welcome to you. Now you. Good morning. Hello, really nice to see you. 

Now at last year's conference, SB and I started a conversation about how technology is shaping the future of banking. In this session, we are going to develop the theme by taking a closer look at AI, one of the industry's hottest topics. The question really is how can banks use AI to drive digital transformation? To get us started Thor, I believe you have a case study to share with us.

Thor Turrecha (meldCX): Thank you, Gillian and SB for having me today. We appreciate the opportunity to share our story, and potentially be able to help a few of our audience in transforming their branches ready for the future. 

So at meldCX, you know, we develop several products, you know, however, for this session, we're going to put a spotlight on Viana, which is short for vision analytics. Now in this session today, you will discover how we've leveraged Intel open vino technology to be able to track how customers interact with branches, like a website, pretty much. 

The problem

So to begin with, we've got a couple of things in retail that we're looking at. One of the problems, really, that we see is that in a banking environment, you have multiple channels, you have at home, you have in branch and you have on mobile. Now the problem is that at home, you're using the web as a tool and you can do self serve, you could access all the products. And whilst you're browsing that the brand or the bank itself can see everything you're doing. And at the point where you go in branch, unfortunately, only a few of those interactions that the brand or the bank can see. And therefore it's like a blind spot. So by the time you get out of the branch, and you start interacting with the brand again through your phone, you will never transfer the interaction that you've got in branch into the phone. 

So today, we'll unpack that problem statement and see how we basically help that through now. As we were ideating earlier, banking, you know, this industry is quite big. And there's a lot of projects that we're working on. We were ideating What's the best way to present this today. So the team came up with a great idea that instead of going through functional topics, we decided that it's best to showcase Intel and meldCX capabilities if we walk through from the experience perspective. 

Privacy regulations

So today we will showcase a customer journey. However, before we deep dive into the journey, let me introduce to you joy. Now, Joy is a selected persona for this journey. And yes, she is a character inspired from my college Joy. Shout out to Joy for letting us borrow your character today. And when you see the image on the left side to us humans, we see Joy as an image, a portrait. And for Viana or platform, you see Joy as a kind of blurred image with a unique ID. And that's basically our commitment on data privacy and we convert video data to actionable insights without compromising customer data privacy. 

Anonymous, tokenised individual

So to begin, let's talk about the entrance, right. And here, I don't know if you can see my mouse there's Joy coming in. We blurred her face. And you can see that we've associated her with ID a 392KJ002 now at this point, we're tracking Joy's impression meaning Joy's consumption of content if their signage engagement meaning engagement to human beings, therefore staff or other customers and interaction mostly on smart ATM technology that are installed in there it could be self-service kiosk, it could be smart ATM most of the time. So we looking into that based on the number of people going in, so density. During COVID we need to make sure that we can limit the number of customers going in. And at the same time, we understand whether you've got enough staff that would serve as customers that are going in depending on the time of the day.

The customer journey - zone engagement

So here, there's always this typical bank that would have approximately 5000 to 10,000, digital signage screens across their network. Now, they will personalise the content. And that's mostly based on 70 plus products across their portfolio. Now, using our platform, we start tracking how effective their campaigns are, by tracking audience attention towards the screen, we will know if they've glanced at the screen or watch the content until the end. And that's very important for marketers. 

So this is zone engagement kind of use case now in branches, they've got approximately 250 to 450 square metre of space, which most of the time they break up a zone. So there's one that is marked ATMs, there's one for the score results. There's one for the heart zones ,and the mostly the prominent ones at the transaction zones. Now, we're helping them understand the, you know, the intent of the customer based on zone engagement. 

Banks today is more about service, right? They're not just pushing products, they're not, you know, throwing in credit cards everywhere. It's basically, if you go on the branch, it's about how do we provide you the best service possible so that you know, we will be part of that journey every step of the way. So one of the key elements that we're tracking is the queuing in branch, so we track how long customers are waiting to be serviced. So that's a very, very important key, then translate into KPI and figure out how, how well the quality of service that they're rendering per location per zone. 

Smart ATM interaction

You will see on banks especially here in Australia, there's heavy investment on smart ATMs now, the smart ATMs today can govern most of the low value transactions. However, it is still believed that it's underutilised and the banks are working really, really hard to promote its features. So we're helping them track behaviours related to at least you know, people using the smart ATMs. 

Now, how many are queuing up how many came up to the teller and decided to do it at the smart ATM instead. So remember, the queuing, you know, we can track them at that point. And then by the time they actually go back and say all the queue is very long, I might just do it, the smart ATM. Or there are situations where they're assisted by staff, where right now they're doing a campaign where staff are educating customers and walking customers that say, hey, your transaction that normally you do with the teller, you could actually do it at a smart ATM. Sometimes we do weird use cases as well, like some homeless guy sleeping at night in a 24, seven smart ATM locations, because now in matters of data, we were recognising staff through their uniforms, because you know, very high kind of profile, High Street area in Sydney, for instance, there's not much people going in, potentially, there's 50 to 100. 

Staff filtering

And if you're tracking five staff that are going in and out, that will make your data difficult to digest and not accurate enough to be able to, you know, useful for decision making process. So we filter staff, so we train our models to be able to identify uniforms. 

Insights dashboard

Now, a little bit about, you know, insights that we get, and we simplify everything. Now, we've found that in a banking environment, there's a lot of data science people, a lot of machine learning people that need to crunch the data. And the problem is sometimes you have to wait for a week or sometimes two weeks before insights are provided to marketers at that time, it's too late, you can't do anything. So everything's lag measure. There's no lead measure that allows you to interact with customers real time. 

So these are real time insights that we've got. So start with people counting zone engagement, which zones basically are getting most traffic, their stuff as well from content effectiveness, where we showcase the ranking of different contents that runs and therefore it allows the marketers to understand which amongst my 72 content are actually performing better, and therefore How do I improve it. 

Multi-channel correlation

And then lastly, there's always a conversation around cross relating everything like phygital kind of thing, physical to digital components. So here's one dashboard that we're working on where we're tracking what's happening on the web, what's happening on the mobile, what's happening on the kiosk, what's happening on signage, what's happening on physical space, aggregate everything and use the AI to actually do a pattern recognition that a person goes in, you're trying to understand, what would this person be interested with? 

  • Is it a first homebuyer 
  • Is a person having a baby or 
  • Thinking of travelling overseas or 
  • Loss of loved one or 
  • Sometimes divorce? 

Bridging the gap between retail and digital

I want to basically conclude with this right. So I want to leave with an insight and a challenge. Now, according to our research, the purpose of the physical branches is currently being redefined. Now, regardless of what it looks like, everyone can agree that it will be repurposed to optimise the bank's capability to serve their customers, regardless of where they are in their life. 

The keyword there is service leadership, right humanising the interaction and engagement between the brand and the customer. And that is likely going to be the role of the branch. Now SB and Gillian, you know, we have limited time, and I've made it as visual as possible. 

So I hope this session has been informative to our audience. And, you know, we, we thank you for having us here and allowing us to share our story.

Gillian Shaw (RBR): So thank you so much for that. absolutely excellent presentation and loved the visuals that really helped me understand that you know, how AI can really be harnessed to to monitor customer engagement and branches. Really important topic. And I have a few questions, if I may. 

And actually, I'd like to start with SB, and SB, obviously, we can see that AI has a crucial role in improving the customer journey, as illustrated so beautifully by Thor there. And is AI really a core focus of Intel. And what else are you doing in the retail banking space to power digital transformation?

Dr. SB Ngoo (Intel): Thank you, a great question. Thank you, Gillian, and also a wonderful presentation by Todd just now. So if I think that questions, I will say that the transformation is really nothing new for financial services. BFSI industry is typically the first to innovate and live in a technology transition. Throughout history, you can think of it as a canary in the coal mine for banks. We see things in BFSI before we see them in other industry, AI is similar, has the potential to transform almost every industry and is really making an impact in the financial services industry today. 

AI in banking, in short, will help financial institution divisions improve the efficiency and the services, as well as risk management and regulatory compliance. 

In additions. I think banks are continuing to look for technologies that can help them to reduce the operation costs. I hope that AI is the answer. AI is a feeling the new capability, whether it's online or offline, at the bank branches through their offering a lot of new services with a digital transformation from a cloud AI to the IoT. You know, India cannot really implement all the solutions without a good ecosystem in place. Besides working with the hardware manufacturers, we are putting a lot of effort to support a lot of independent software vendor partners like mausi X, in enabling this new cutting edge solution,

Gillian Shaw (RBR): I'd say really excellent answer SB and obviously highlights the importance of building these strong ecosystems with partners working together. And I know that's a theme that we explored last year as well. So it's great to have meldCX here today to illustrate that point. So clearly, just a point, actually, obviously, we are still in this sadly, in this current pandemic situation. And we're still being encouraged to reduce our physical contacts. And I was just wondering, does AI have a role to play in this? Can it provide some solutions to this dilemma?

Dr. SB Ngoo (Intel): Yeah, I can take that question. So, this great question, you know, post pandemic, you know, this is the really the hot topic and the hot request by the bfsi end customer bank, which is the touchless of contactless indirection, you know, artificial intelligence come at the right time to enable this function for the industry. We more than this AI solutions deployed, especially on the computer or computer vision, this technology I believe that we can categorise into three key area that to support the pandemic society. 

  1. Now, first one, as a regulator to ensure the physical distancing and maintain through the visions analytic. 
  2. The second, reduce the interaction point, to try lower infection with an intelligent camera and a heat map. 
  3. And the last one, I believe that you will streamline operational processes to simplify the process and at the same time, maintain the security policy, such as utilising computer vision to authenticate the consumer or customers and reduce queuing time when the stabbing into the bank branches as just a call show a while ago, 

I’d also like to touch on an example, you know, one of the use use cases that we have seen during a lot of demand for this physical distancing, or contactless will be the example like integrating the AI functionality into the current transaction or ATM machine to transform it into the interactive teller machine, or kiosk just like what Thor mentioned, the smart ATM, right, which enable the in depth self service engagement for consumer beyond the cash machine. 

But with the video teller, or AI chatbot with a digital assistant or avatar. So, in short, utilise AI will help to smoothen process of the SOP during or post-pandemic, which will be the new normal for hours in the future.

Gillian Shaw (RBR): Yeah, that yeah, that's exactly right. SB it's really going to be a powerful tool. And I do have a thought, though. And this one, I think, I think is for thought we're seeing an acceleration of automation and branches. And really, I'm just wondering, do you think the traditional banking customer is ready for more human less transactions?

Thor Turrecha (meldCX): Yeah, it's a very interesting topic. In banks, their different views to people that are doing smart ATMs, they think there's no, there's no use for having the traditional approach to the people that are selling home loans. Definitely, they think that humans are required to do that. 

And you know, to the people that basically are in the teller, which is basically the grunt of customers going in when someone complains, sometimes they'd like it, sometimes they do. So I think the future will be a hybrid. More and more, I see AI as an ability for humans to be more humans. 

Now, it's quite interesting how I put it like that. The reason why is that there's a lot of transactions that humans do inside banks right now that are low value, low margin, and therefore there's, there's not much association to emotion. So it's like routinary kind of job. So you don't do much, you don't even interact with a person, no one's even looking eye to eye, right. So I think AI would help through that kind of, you know, kind of routine interaction. 

Now, there are things that you cannot, you know, do human lists, meaning, a good example would be 90% of humans that owns a house would only buy a house once, and therefore it's a once in a lifetime event. And therefore when you make that decision of buying the house, it's very personal. You talk to your wife, you talk to your family, you talk to your friends, you will spend six months, years even before you make a decision, you will only make a decision when you talk to another human that knows that space. And when you get that thought leadership, then you're ready to make that investment because it's like another marriage and not 30 years more than that, to get the sort of house right. So I don't see that it would disappear. But I think AI would definitely help humans service humans better moving forward by removing the stuff that humans isn't like doing anyway. So that's, that's my, that's my position.

Gillian Shaw (RBR): So as you quite clearly demonstrated in the presentation, you keep those sort of human transactions for the hearts, which are quite excellent. So actually, back to SB. It's obviously it's fascinating to see just how quickly this AI technology is developing. Could you perhaps share with us some other similar AI projects that Intel is working on in the retail banking space?

Dr. SB Ngoo (Intel): Yep, thanks, Gillian. Good questions. You know, besides the use case for the content effectiveness, anonymous audience measurement, as meldCX has mentioned or shared. We also have seen that and done a few projects that are in different regions, such as I just mentioned a while ago, the AI advisor in that diligence chatbot with the digital avatar you know, the digital assistant in a self-service kiosk machine for the retail bank. And also there are some other new stages like the use of the AI vision analytic for the security civilian and the retail banking premise, which is to complement the physical security today. So before I forget, I would like to do some advertisement here in the upcoming lengthy section, for the self-service banking section for the EMEA and America. We will have another two sharing sections on the self-service kiosk for the digital transformation in detail and also the introduction of AI chatbot or some type of conversational chatbot use cases. I will strongly highly recommend our audience to watch this section on demand to know about these two new use cases if the time zone is a challenge for you.

Gillian Shaw (RBR): Yeah, excellent. SB, we're really looking forward to seeing those use cases at the EMEA and the Americas conference in the next couple of days. Thank you. And we also have the out of time. So it remains just for me to say thank you very much to Thor. And thank you to SB for a wonderful presentation, and a fascinating walk through the power of AI, and how it's transforming the retail banking space. If you'd like to find out a little more, please do visit the Intel booth in our exhibition hall where you can chat to SB and his colleagues and they'll be happy to answer any questions you have. Thank you both very much. Thank you.

Everyone: Thank you.

Latest from meldCX

meldCX's Vision Analytics Solution Showcased at TD Synnex Showroom in Munich, Germany

Mar 19, 2024
2 minutes
Featuring our Content at the Right Opportunity (COATRO) solution for digital signage. In partnership with Intel and Signagelive.

Get the latest meldCX news and insights right to your inbox!