INTRO
This is the OPS Game Changers podcast where we share real life stories from remarkable leaders that are transforming the world of service operations in financial services, healthcare and BPOs. In this series, we explore some of the key challenges faced by service operations as they strive to deliver more, more capacity, more productivity, and more business impact. In each episode, we delve into how these OPS Game Changers have addressed a key challenge or issue. Our guests share how they went about unlocking significant value, adopting best practices, and experiencing some game changing results.
Bhavesh Vaghela: Hello and welcome to OPS Game Changers, the podcast that shines a light on the crucial role that operations teams have been driving business performance and improving staff and customer experience. My name is Bhavesh, I’m the CMO at ActiveOps and I’ll be your host today. Now, in each episode, I hone in on one KPI or challenge that a business has solved, delve into how they went about doing this, and explore the outcomes achieved.
Now today, we’re actually doing this live. I’m at the Nedbank offices in Johannesburg, and I’m delighted to be joined by Brinsley Du Plessis, who’s the Managing Executive for the Transformation and Program Management Office over here at Nedbank. And they’ve been on quite a journey. They’ve been leveraging data, operational data to identify and optimize processes to drive their customer experience and their customer outcomes. So we’re going to find out a little bit more about that. So before we jump in the topic, let’s welcome Brinsley to the podcast. Brinsley, welcome!
Brinsley du Plessis: Yes. Hi Bhavesh! Really a privilege. Thank you for inviting me to the podcast.
Bhavesh: Well, thank you for allowing us to come and film in this fabulous office here in Johannesburg. So you know, the whole topic of this is all around how do you use data to drive operational efficiencies like processes? Every Ops team, every transformational program in some shape or form is all looking at straight through processing, removing complexity and driving those better outcomes. So we’ll delve into that a little bit more today because you’ve got a really interesting journey in terms of how you’ve gone about doing this. Well, before we do that, tell me a little bit about yourself. How did you end up in your role where you are today?
Brinsley: Well, it started well, 35 years. I’m with the bank for 35 years. Yeah, I started back in 1990. You know, I started as a teller at the Perm interestingly. And I mean, as you know, Perm was bought by Nedbank back then in the early 90s. And yeah, I just kind of got to know everything, worked through the branch, all the functions.
And then retail and business banking was integrated in around 2015. I just obviously had various executive or division exec roles and I was then closer to where we are now and kind of the topic for today. In about 2018, I took on the role of the transformation office and program management office. So really coming out of line business and really to drive a lot of the strategic program office work and transformation agenda or business transformation agenda.
Then late in 2020, back then we had product verticals within retail and business banking. What we found there is that we had bespoke product areas and they kind of went from product board design, understanding CVPs, but also they had the credit function, the operations function and also middle back office, the processing of applications and servicing as well. So kind of end to end from strategic to operational, but with that obviously bespoke systems, bespoke processes and we started to pivot as a bank to really being, you know, pricing for the client, credit for the client.
Bhavesh: So really trying to drive more customer centric?
Brinsley: Yes. Then Project Phoenix emerged and I was tasked to lead the reorg design back then, beginning of 2021, really going from vertical product lines to horizontal solution.
Bhavesh: Understood!
Brinsley: That meant taking all the Ops areas, really pure operational middle back office work and getting a centre of excellence across the product and also the channel lines. Some of the product areas had their own channels as well and we had a branch infrastructure and context. So really bringing and we created what we call the Nedbank Integrated Channel with all the channel competencies and capabilities coming together. We had all the ops areas coming together if you think pure middle and back office operations. And then also we had an integration of all the credit deal underwriting, modelling and pricing. So you can start to get closer to pricing for the client holistically, credit holistically etc. So that was really where I kind of drove the reorg for retail and business banking back then.
Brinsley: And then obviously Phoenix is now starting to be wrapped up and, you know, I’ve got vast experience on both retail and business banking in my ops days etc. and I was then asked to come back in, come back into line business and roll up my sleeves, right.
Bhavesh: How did you think about that? Think about getting back in there, rolling your sleeves up?
Brinsley: You know, I thought, you know, it’s what I’ve done before. You know, one thing that I did miss when I moved from a pure line business into the transformation, because it’s smaller teams, it’s more SMEs. I enjoy interfacing with people on the ground, you know, with the guys where the, you know, the rubber hits the road, material and resource and really working and engaging the teams and mobilizing them at the end of the day. So that’s when I then in 2023, about Q3 of 2023, about September, October, I then took on the Ops, but I kept my business transformation team with me. It was Ops and BT, that’s where it comes together, right. And that’s how I got into the role. And I then obviously started to work again with John and Sarah and so forth. That’s how I got back into the Ops world.
Bhavesh: So it’s really interesting, right, because you know, the whole Project Phoenix thing is a fascinating example of how an organization can change rapidly to centre around the customer. And I know in the first few years it was all about clearly getting the whole idea of data embedded into the Ops environment, having a standard set of figures and stats that everybody’s consistent with, and really using that information to drive individual performance and drive that productivity and use the capacity more effectively. The second phase of that though, is really then where everybody wants to go, right. I mean, it’s been like this since day dot with transformation and process. Everybody wants to re-engineer the processes. Everybody wants to take the waste away. Everybody wants to straight through, self-service, automate. You know, we’ve had the era of bots and now we’re about to enter this new era of Gen AI and the world is going to change in terms of the ability to drive that customer facing straight through type of interactions. But so you’ve got a great foundation when you came back into Ops, but you started to tackle the processes themselves and trying to figure out how do you streamline and make them more efficient. So let’s delve into that because that’s really what we want to talk about today, because I think it’s a really interesting example of how you’ve been able to use the data to identify. Because I think for me, that whole idea of process transformation is two things, right. First, how are you going to identify which of the bits you’re going to spend your time and effort on? And then secondly, how did you get the outcome? Because a lot of these transformation programs tend to fail because they don’t pick the right area. And secondly, they don’t deliver the business case outcome at the other end. And everybody’s struggling with that, right. So I’m really keen to delve into that. So, you know, how did you use that data to first start identifying the areas to go after?
Brinsley: And Bhavesh, you know, it’s a fascinating question, right. And you know, I’m a firm believer that when you start to tackle the dilemma, as you’ve just mentioned, I want to say it’s almost like taking a picture. Some of us like to take the picture right up front and use it like a passport photo, right. So I said before you get into that for Ops, let’s stand back and take a picture of the whole thing, the whole thing. OK, so let me explain. I mentioned early on retail and business banking, 14,000 staff. But if I kind of break it up into front office, middle office, what we sit with is a branch channel, physical branches out there, you know, just over 400, probably just under 5,000 staff that interface at the coal face with the customer that will take a product order or a service order need. Yes, we’ve got in the centre all our product segment development, constantly understanding customer change needs and wants and they drive the innovation, next best thing etcetera and they evolve and also fix things that are not resonating with customers. But then once the customer comes in, whichever channel they choose, which could be the money app or our mobile, it could be the web, it could be the physical channel of course, or it can be the relationship hubs where the business managers sit etcetera. Now that work or contact centre, yes of course, just over 1,000 contact centre agents. So all of that work is front office work. That’s where the customer sees and feels it. But then a lot of that work where it’s not straight through processing as yet has to go to what I’ll call the middle office. And that’s the area I took over from Nozi. 1,250 of the staff are BT staff, specialist analysts etcetera, project managers, but 1,200 are core operational staff. And you know what is very important when I took the area over, and I told you I like engaging with the staff, I said you know we are no different to the planes we fly in. We see the pilot, we see the air steward, the person doing the ticket, but if that guy doesn’t pick up to bring the staircase, you’re in trouble. You’re not getting on and you’re not getting off, you’re going nowhere. I always say if Brinsley doesn’t come to work, the customer doesn’t care. But I have to tell myself, if you guys don’t come to work, customers are going to feel it. So first and foremost, understand the role you play end to end, OK. And that’s why I’m telling you the story this way. So we’ve got in that 1,200 three mainstream functions, service and support mainly doing a lot of your transactional type products, right, current accounts, forex, investment products, home loans, overdrafts that need fulfilling and conclusion in the middle office. That’s the 1,000. This is where Nozi had initially started with ActiveOps and getting a sense of the work that flows from the front to the middle office. So now we got a good sense of volume, time standards, productivity, etcetera, how the processes behave once it comes from the front to middle. We’ve got MFC, the motor vehicle which is really the dealership out there because a lot of customers when they want to, that segment really, I want to buy a car, I want to drive it off within an hour. Yes, you’ve got to. So the dealer is our customer and then but they then put it through, it comes to us. We’ve got 60 minutes within which to approve or decline to get the podium because they send it to all banks, OK, another 300-400 staff. We do the credit, we do the verification, validation and we want to get a podium in the dealership, right, because they will sway the customer.
And then we’ve got the Regulatory Enablement Center, which is stand-alone because, as you know, same as in the UK and all over, we have to comply with FICA and AML. Enhanced due diligence, ongoing due diligence, high-risk industry, politically exposed, which requires a bit more rigor before you can say yes, the customer can talk on that. So those really are the three functions.
So as you can see, if any of those areas don’t come to work, the plane’s going nowhere.
Bhavesh: You’re not flying.
Brinsley: So really, you know, to now go and say, OK, how do we optimize our processes? You need to understand the full end. And one of the key things that I did when I took over and went around to the teams, even though I was in Ops many years ago, things have changed. I did, you know, Genchi Genbutsu — Japanese — go see for yourself. Just going to, OK, what’s your process doing? I’ll sit with a guy and say, OK, what you’re doing? OK, this is what you’re doing at the box. I see you counting it here, OK, I see you. Got a good sense, just to re-familiarize myself, just immerse myself, get back into it. What are you doing in this? You have lots of paper or lots of handoffs. And I got a good sense, right.
But importantly, I made each and every staff member doing the work, their goal commitments or their performance indicators, were linked to three top-of-house. So Ciko’s our ME, he obsesses about three things: are we growing revenue market share, are we managing our expenses judiciously — are we optimal — and are we managing impairments?
Bhavesh: Got it, right. Yeah, so those are the three things that matter to him.
Brinsley: So if you were doing the processing of an OD application and payout, you were linked to revenue, but also to our cost, because if you do it wrong, there could be a loss, right. And if you were doing a bit of credit, you know, are you doing it judiciously that there’s no… So everything pointed to one of those three good purposes.
Bhavesh: So you’re giving all your teams, I guess you’d call it, clear line of sight — what they do day-to-day and how that impacts.
Brinsley: Clear, clear line of sight. Yeah. Now to your question. So when I joined, Sarah said basically, you know, can we orientate you with ActiveOps? I said yes, because I just did an introduction with John. I didn’t get into the detail, that was Nozi, and she got that off the ground. And then I think it was Damien from the UK who then spent a couple of hours with me and he started to navigate me, to show me. I said, take me through the mechanics. You know, I sat with the staff member, I see they do something, then they go onto ActiveOps and they say one or two.
Bhavesh: Got it.
Brinsley: And I see the team guys are starting to play. Just maybe take me into the engine, yeah, and see what’s going on behind that. Just take me through it. And he said, OK, you’ll probably use this report and whatever, etcetera and so forth. And he said, OK, but you’ll use this one. And then he kind of, you know, went quickly over the Pareto report, right. And I said, oh, pause, come back, reverse. And I said, OK, can I see what’s in there? And he clicked on it, and I said, this is the ring that opens it all up. It’s like gold for me and my team, right, because I wanted to apply a strategic lens onto it now.
So I said, OK, I then got a couple of the BT analysts. I said, guys, can you take this Pareto data and bring me insights, right? And you know, because I’m a firm believer, you’ve got data, but data is also meaningless until it becomes information. So what we saw on the Pareto is lots of information flows for the different products and services etcetera. Now we needed to string some of those data things together and say what are the insights. And that for me was where I took it over from. Without that foundational capability, I would not have been in that position.
The team and I then used that as a starting point to point us to where to go and look, where we were suboptimal, and to select the processes end-to-end, not just in our middle office.
Bhavesh: Full on.
Brinsley: That’s where we needed to collaborate, OK, to be optimal and get the best benefit for the customer as well as for our clients that were buying our services.
And one thing I forgot to mention — we took a lot of the data points and the insights that we got, and then I gave it to my finance officer. I said, now can you take the aggregated cost of every department and function and give me a cost per minute?
Bhavesh: So you boiled it down to the actual…
Brinsley: So now we knew that department was 5 rand, that department was 6. And you started to create a bit of competitiveness. I said, now we put a commercial lens on it. Yeah, we are selling us. Why must the segment and the channel come to us to process the home loan? Are we optimal because we’re charging 5.12 cents a minute? If you multiply that output, you can see what we started to do is really show them how they play across all those three. And it just changes the mindset of how they now engage with the data.
Bhavesh: Of course, because there’s an understanding there in terms of what the outcome is. Now, one of the interesting things about the data set that you identified, which is the Pareto data, is that it looks at the tasks that are taking the most effort.
Brinsley: 20/80 was very much alive.
Bhavesh: And it’s really interesting because, you know, when we look at general process transformational type programs, there’s lots of data you could use, right. You can use your systems data, bring it all together, and then help you identify the process changes and the process elements and give you the whole picture. You can do a whole bunch of deep dives with process mining and there are so many things you can do. But in many cases, it’s that effort data that makes the difference. It gives you a different lens on actually, I’ve got all these areas I could focus on. Well, let me go — and it’s already telling me this is where I should go and focus. So that foundation clearly helped you think about the areas to go and focus on.
And with the other work that you did around boiling it down to, kind of the common denominator, which is the cost per minute across all of it, it allowed you to understand, I guess in some ways, it helped you build your business case, I would imagine, right. If I went down to that task that took the most effort, I’m going to save X minutes. And then that made you — was that how you made your decisions?
Brinsley: And you know, when we put up the cost per minute, it was actually quite fascinating. The respective functional Ops executives said, oh, I see you’re 6.12. MFC, you’re just under 5 rand. They were the most optimal, right.
Bhavesh: How are you doing?
Brinsley: No, they started looking at the competition, you know, am I going to be able to sell my service because they look a bit cheaper than me, right. So you just heighten, you lift it to a strategic conversation, right. That’s where I wanted it to start. And then the how was pretty easy. I stayed away from the how.
Bhavesh: OK, got somebody else to do that.
Brinsley: Because I know I’m clueless to the how. I just figured out the what and the why. So what we did was, I told you, the guys did the Pareto analysis and with no surprise, 20/80 is still alive and well. The theory still lives on from 40 years ago. We found that when we took — so to put it into perspective — across our three functions, we managed 95 service-level processes, service-level agreements that would impact the customer outcome. Either a service request — I need an extra-long statement because I can only get from year to year on WhatsApp — or I need to register a mortgage, I’m buying it.
Bhavesh: And some of those are incredibly critical because they are related to regulation and foreign compliance, and there are big costs involved and fines involved in some of that stuff.
Brinsley: So 95 of these are critical. We are selling these processes. If you had to now, if I put the commercial and the average cost per minute, if I aggregated to the top of house, it was 6.12 cents at the end of 2023. That was our baseline.
Bhavesh: Got it.
Brinsley: Just at an average, because every department’s different. And then I told the team, if I take 6.12 and I multiply it… OK, let me not run ahead of myself here. We had 95 SLAs and 855 subtasks. You could have one with 10, one with 100 — 855 subtasks across the 95 SLAs across those three operations.
86 million minutes per annum is what you were taking to do all of that. Multiply by 6.12, it was coming to 60% of my total cost base. 300 million total cost base before transfer pricing — what I pay for this building or whatever — was 591 million for the year.
Bhavesh: Got it.
Brinsley: That means lots of overheads because it’s paper-based or manual. There are lots of check-for-checkers, so lots of overheads. I’ve got people, I’ve got to pay for premises, etcetera. But if you take direct cost per minute, I’ve got about 40% waste. If I have to think about it, yes, you want to probably get to about 80. You’re always going to have overheads of course, but 40%. So that’s the first KPI.
Bhavesh: So you obviously know at that point there’s a lot of efficiencies to be had there.
Brinsley: Correct. 1,200 staff doing this work. Then we said, OK, now take each of those 855 across the 95 business processes. I don’t know whether one has got 10 or one has 100. Draw the line at the 20%. So if there were 10 tasks to make this thing up, draw the line at two. It’s 20/80. And guess what — the time and the volume was anywhere between 73 and 84% depending on the department, of course. So we said, guys, we then add a theme — we’re going to attack the minutes.
Bhavesh: And that was known to the staff — to attack the minutes. Can I call it the ‘Attack the minutes’ mantra perhaps? Was that therefore known to the staff?
Brinsley: So we did the roadshow, we showed them the volumes, and then we invited them to the big conversation.
Bhavesh: So they felt part of this journey that you’re on.
Brinsley: We said, guys, you know, we’ve got various problems. We’ve got to load balance, we found. So there were almost two domains — it’s running the business and the challenge I put to the staff once we showed them the data and the volumes etcetera that they were capturing daily on that box. And now we showed them the big picture.
Bhavesh: So they could understand the part that they cost per minute and whatever.
Brinsley: And we then said, guys, we have two halves of the story. The first is run the business. So daily we have to meet SLAs — 95 SLAs — whether it’s got a 90% pass mark, a 95 or 98 depending on how much paper versus digital etcetera.
And I said, guys, just imagine this. If you are doing 100,000 widgets a month, 10% is a lot that can go wrong — it’s 10,000. Some of the guys are doing a million things a month. A 10% error rate is a lot. Just imagine buying a McDonald’s burger with 10%.
Bhavesh: Then we’ve got an unhappy customer.
Brinsley: Exactly, given the size, you know. So the guys started to think — OK, listen here, we can’t accept 90. So guys, we need to move most of our SLAs to 95–98, and we must get to Six Sigma level eventually, as far as possible. And people then started to understand because you’re having a big conversation with them.
Bhavesh: Yeah, they’ve got context now, right.
Brinsley: Then we said, guys, you’ve done a lot of good work in terms of productivity and capacity management within the vertical functions. To some degree we’re moving in the sub-functions between each other — load balancing and using ActiveOps. But there were some people that were starting to go across, but it was kind of a process. We had to meet with unions, then we had to meet with HR. I said, guys, the skill set, the profile of the person — whether they’re here or here, whether you’re doing home loans — is the same. How do you make one big Ops and make it more fluid?
So we showed the guys some of the businesses that were burning volume time standards. We said, you can see they’re not going to make SLA. You can see the SLA is red. So I asked, how can we solve this? And people started putting their hands up — we should automate, better straight-through processing. Most of them said flexible work practices. I said, what does that mean? They said, you know, where we’re quiet on these days, we’ve closed the taps because of impairments on personal loans. So we’re doing less of these, more of the other. Why can’t we go and work there? I said, great.
We had a picture of a boat and two guys sitting there saying, that’s not my problem, and two guys sinking, throwing the water out. I said, are the orange guys after the purple guys? And the guys said, OK. Do you know what actually happened? Guess what the challenge was — is there going to be a fair process to select me to go there?
Bhavesh: So it’s already created this idea that, yeah, I want to be part of that.
Brinsley: Yeah. Now the question was, are you going to choose who? Because people then also started to realize, but then a lot of the recommendations were straight-through processing and automation. But they also know that means it’s going to drop here. But if I can make myself future-fit and also go and learn there, you’ve got a career path, you’ve got a future. So that for me was the big thing — bringing them into the big conversation.
And I used the example. I said, guys, now you see 86 million minutes, we’ve got it 20/80. I asked the teams — we’ve got our top performers, and these guys want to add value once you give them purpose. I’d bring them into the room and say, guys, which of these 86 million minutes in the 20/80 are we going to focus on? We created a new language. You guys said STP, robotics, and Gen AI — some of the younger ones said Gen AI.
So I said, OK, which of our processes are STP-able — that’s the language — and which are robotics-able, and which are Gen AI-able? We added a bit of fun in there. Can you guys go and have a look at it, do the mapping, and tell me which are candidates for each of those?
Bottom up, 24 million minutes were STP-able, 8 million minutes robotics. That means it’s got to come to us and then we do the robotics from here, whereas STP means you’re at the start of the channel — that means we’ve got to go and work in the channel to say why the work mustn’t even come to us.
When we stood back and got the bottom-up input, the divisional execs were there, MANCO, and they had worked with analysts and their teams to get input and come up with ideas. We worked out that if we could do the 23 or 24 million, if I had a magic wand…
Bhavesh: Then what would it mean?
Brinsley: We would have 40 fewer SLAs to worry about.
Bhavesh: OK, that’s huge.
Brinsley: Why do you have an SLA? Because there’s work flowing. If there’s no work flowing, there’s nothing to worry about — OK, 40. The rest would have been taking a ten-step process, making it five steps, making it shorter, and also getting closer to 98–99% SLA performance.
So we then, as a collective — now that we had the bottom-up view — set ourselves a goal that we were going to…Over and above the work that Nozi and the team had already done, because they had done a lot of heavy lifting prior on productivity, the next horizon was to attack the minutes, attack those 23 million minutes across the functions between 2024 and 2026
We did 8 million last year. We earned about 5.9 million this year, and we still have the balance of our 24 between now and the end of the year. And then the balance — if you added it all up, 24, 8, and 7 — was about 40, 27, 28. So we carved it up into short, medium, and long term.
Bhavesh: Fantastic.
Brinsley: So yeah, that was the journey of how we zeroed in. And oh yes, one more thing — we then went to our segment owners and to our channels to say, OK, if I cut this time, what do you want? Do you want me to drop the cost per minute? And an exec from RMB said, no, no, that one SLA that you have a 90% pass mark — make it 100. Or that four hours, take the people and redeploy them and make it two hours because that’s a key one for me.
So we had a balanced view of redeployment to make it go faster, but still within the remaining steps in the process. We did an overlay from channel and segment so that we didn’t just give them the option of the cost.
And just to give you a sense of it, we targeted 92 headcount for 2024 associated with the 8 million minutes. We banked 71. The rest was redeployment — to go from 90 to 95, or in some cases to give them a quicker turnaround time or dedicated queues, etcetera.
Bhavesh: So one of the things that you hear a lot when you’re talking about these kinds of programs is the general fear from the staff that my job’s going to be displaced. But what you’ve done here, from the sounds of it, is you’ve brought the customer and the team along the journey with you from the bottom up, as you described. And they feel part of that outcome because they realize what that outcome means from the SLA point of view and I guess the savings.
Brinsley: And from the pocket, that’s from the pocket. And I’ll explain to you how we did that job. But finish up.
Bhavesh: Yeah. So does that mean therefore then inevitably in these kind of scenarios, you potentially need less head count and those kind of things happen and you’re redeploying these individuals, another part of the organization?
Brinsley: Yeah, Our, our, our retrenchments. We reduce that with no retrenchments, right. Because if you are on the front foot and you have foresight, listen, now I’ve got these initiatives, it’s going to land in six months or whatever. As you start to get people natural attrition, you start just slowly freezing it, get contractors.
Bhavesh: Exactly. You don’t need to.
Brinsley: Then as people contract you, they eventually say, OK, you reduce that. Here is another big conversation. So I explained to each and every staff member how their bonuses, how do the SDI sort them in this incentives work out. It’s related to headline earnings. I shared with you earlier. The group is 15 V 5.6. So you’ve kind of joined it all up all the way. So I said guys, you know, bonus pools are set based on our Q0 budget. So if you said you were going to do 10 at Q0, did you do 10, then your bonus pool is intact. If you did 8, your bonus pool starts to dilute.
Bhavesh: And therefore it’s going to hit my pocket.
Brinsley: Yeah, correct. And based on so when the guy said no, no, but we heard that we grew a 7% aspect, that’s year on year. Q0. We were supposed to grow 30. So when I started to walk in, they used to ask me ask Q0.
Bhavesh: So now they’ve got a vested interest.
Brinsley: Yeah. So now they’re interested, they’re not interested in year on year growth. Talk to me about Q0 because now they know that you know, our short term incentives. Obviously there’s lots of moving parts, right that I tried to make it as practical is linked to the bank firstly achieving its objectives ROI headline earnings, then down to the cluster and all the way. How do they fit into those three revenue manage your cost effectively and optimally effective and efficient and then the impairment. So those are linked to the impairment parts, right. So people had an appreciation or so why they have a vested interest.
Bhavesh: And did you find that? So I know you do kind of pulse surveys and things like that. Did you find that the employee engagement empowerment, attrition went down? Did you find those kind of things that happening?
Brinsley: So I’ve got probably 2 metrics there. So we, we track on, we track on the pulse survey. It’s similar to like a client NPS, you know, are we a great place to bank and are we a great place to work? And there’s a whole lot of dimensions and things that that get asked etcetera. So probably one of my very proud moments in this is that we took it from 5 in 2023 and as at the last check, we were close to 25 and went from there to 20 and then to 25. We’re even higher than the cluster average. A great place to bank was also in the kind of teens, it’s like 32 how they think about the bank. So and then another data point that I was tracking is sick leave year on year, our sick leave, 23 on 24, anywhere between 25 and 30% down year on year. So number of days per person on average coming, yeah, coming, yeah, coming down, which gives me a sense of well-being, you know, morale, etcetera, because that’s something that I look at. And then at the same time we were also tracking are people taking leave the minimum of 10 days, Right. Yeah. So you didn’t want, you know, that to be dropping in your kind of slave driving that was that didn’t drop, you know, people that spiked in December. So those were kind of how we triangulated some of the measures because you can’t just look at one, you’ve got to look at a couple right at the end of the day. So so there that was quite pleasing that you know, so I always say it’s like a five day Test cricket game. You know, it’s not a one day game, this thing around bringing people on the journey, etcetera. But you know, the big thing is, was for me and the leadership of the divisional execs looking after their OPS, right, is to focus on the what and the why and the teams figure out how.
Bhavesh: So I was going to ask you, you know, looking back at this journey, what are you most proud of? There’s a lot there that you can pick on, but what what would you say is your, you, you from a personal point of view, what are you most proud of.
Brinsley: I think what I’m probably proud of was to see the team achieve the results because I mean, our job as leaders and at the senior level, we achieve results through others. That’s the bottom line. Yes, OK. And for me to have seen that Pulse grow, but more importantly to see the guys hit the objectives right. And if I just, you know, look at some of the, the, the, the key milestones, we went from 6.12 cents per per minute aggregated top of hours to 5.83 from 2023 to 2024. Our Q0 we penciled in at 592,000,000 for 2024. We came in at 591, which was a 1% growth year on year from 2023. 1% with staff salary increases, etcetera.
Bhavesh: It’s amazing.
Brinsley: And so our TP buyers, our transit guys have paid transit pricing for the work that we do. They should have been happier, right? Yes. On our 95 Slas at the beginning of 2024, about 23% of them were failing the pass mark, whether it was a 90 or a 95% pass mark at the end of the year, only 5% were not making their pass mark, desired pass mark. At the end of 2023, we had 25 Slas that had a 90% pass mark. We have none at 95%. At the end of 2024. 16 of them went from 90 to 95, 9 of them went from 95 to 99.
Bhavesh: Amazing. So you’ve seen a general upward trend.
Brinsley: So you look at the multiple, remember, I say optimization has 2 parts, are you effective? And then are you? Because you can be efficient and hard to be effective, right? And I always said, guys, you know, we must when we do things, whether you’re attacking the minutes, we must be prudent, right? It mustn’t be at all costs, right? Just saving headcount, etcetera, must be effective at the end of the day.
Bhavesh: Because your, your outcome is a delivery.
Brinsley: That’s correct. So if you asked me about that was kind of seeing the team own that and being able to play back to them. Guys, this is what you did. You know, at the end of the day.
Bhavesh: That’s amazing. And I think, you know, my sense is, and I’ve had a chance to talk to a lot of net bank colleagues in over the last few years and there’s a general sense of of staff well-being here, which is definitely a differentiator in terms of what you guys have. And this is another demonstration of that. Really. Can I just take you back a little bit? Just I’m curious about a couple of things. You know, you we talked about, we touched the AI question now at this point where everybody wants to do AI and I know AI means different things to different people depend on what you describe it as. But you you kind of we talked about this idea that some bottom up what is straight–throughable, what is a Gen AI–able and I’ve already forgotten what the robotics.
Brinsley: The robotics one.
Bhavesh: Yeah, exactly the robotics 1. So when you now look where you are today or when you were at that point, did you see, did you use any a Gen AI in there or was it predominantly bots and straight through process? Is that where most of the head?
Brinsley: So the initial was predominantly the bots and straight through. I’ll explain. So what we did in last year, we started exploring a few use cases for Gen AI because obviously still early days, early days, yeah, a lot in our channels like for our chat bot, etcetera, like NB chat bot on the app, a bit more front end Gen AI helping customers, you know, the same with card delivery as well. When the customer is starting to get the WhatsApp. So we’ve tried to test there that, you know, before it was an SMS, they don’t see the SMS. Now it’s through WhatsApp and the bot helps to a degree until the customer comes up here to say, you know, firstly, I don’t like this card. Well, OK, we’re asking about your renewal of your card. Then it’s got to kick off into a call agent, etcetera, right. So we that we’ve just now launched now this year. So all our card renewals now are starting to work through WhatsApp with a bit of bot, but then also with workflow to kick it into an agent if a customer is not within the script etcetera. In the OPS world, most of our Gen AIwas focused within the regulatory enablement center. So we’ve got a few use cases there where so as you know, if you are an agent doing enhanced due diligence or you need to see is a person politically exposed, a high risk individual, there will be things where you now need to go and check adverse media and so forth. And you know, when you sit with one of those guys, flip it, it can take them two days just to go through a enterprise that’s got all these associated parties and now you must go and see, OK, Bhavesh, are we picked up a bit of negative now you must read the media flip. It can take you like a couple of days to do that one, right? So that’s where we saw how do we start to use the AI to help those individuals be more effective, do a lot of the grunt work that they can then just focus on, OK, was this really an ad and make the decision, etcetera. So that’s kind of where it’s where it’s at. I don’t have the answers as to how they’re doing because as you know, I told you about Phoenix coming to OPS and as you know, our group has now decided to also reorg and I’ve been asked to do a lot of the leading there. So since the beginning of the year, I’ve been kind of focused on that. And a while something that has been caretaking the OPS and Nozi now obviously taking the reins again there. But that is kind of where we have pointed the Gen AI focus because it’s complementary to making the human being and the staff more effective and that’s where we are pointing it towards.
Bhavesh: So can I ask you your view then? So you’re now walking into you started with the Phoenix, which was, you know, customer centric, bring the classes together and then you went from the people to the process element fundamentally within the OPS side, connecting from the middle and streamlining the whole process from front, middle and back. And, you know, attacking the minutes.
Brinsley: Are we still busy with that too?
Bhavesh: Of course, the road is long. That’s never never really finished right? And that’s a constant, constant improvement. And you’re now faced with a much larger reorg that’s going on within the organization. I’m curious when you’re looking at the the next phase of reorg, do you have in the back of your mind or is there a driver from the the board and C level that says how do we use AI more effectively? Is that a at that level up are you looking at it go, OK, how do I now start thinking about the next five years and how AI is going to fundamentally change my organization.? Is that how it does that?
Brinsley: I mean I was giving a reference specifically to the OPS will bet in a lot of our other frontline businesses potentially a bit more advanced in terms of starting to extract value in terms of AI. But generally in terms of our group technology that supports us with any technology requirements, you know, because technology is a means to an end, right. Once you’ve identified your end, you know, what does it mean for the customer, right and and which customer that you’re targeting and how you’re going to win within that in that customer segment in market, you then deploy your respective technology. So there has been a lot of advancements in focus in that regard. I know even in our group technology that they started exploring using a lot of the AI to help them with routine business analytics developments, you know, also to improve the effectiveness of developing code and things like that, right. So definitely it’s a key imperative and one of the technology levers over by robotics and straight through processing. Obviously first prize generally is straight through processing, you know, whenever you look at something, once you’ve, you’ve understood the end, what’s going to woo the customer from the competitor to yourself, make you distinctive.
Bhavesh: And do you, do you think that’s a really interesting point? But do you think that if I look at the, the market industry and peers who are doing this in other areas, do you think the Gen AI thing is a bit of a smokescreen? Do you think that actually there’s so much stuff underneath it, like straight through person, like bots that can do so much and we’re only just scratching the surface of what those are capable of doing before you even think about the Gen AI stuff? Do you think we’re almost taking a leap without figuring out how you solve some of the other areas?
Brinsley: Yeah. And I think once you understand what the capabilities of each of them are.
Bhavesh: And then you know where to play them.
Brinsley: And like when the guy said, how do we determine what is straight through processing? I said, well, you know, I used, so I used the example when I was a cashier at OK bazaars. It’s a supermarket, one of the supermarket chains in South Africa in 1987. I was in my final school year and I was a cashier on a Friday and Saturday and Sundays. And back then, you know, I had one of my buddies used to used to have a gun. That’s the guy with the price. And then we want to compete and it’s 89. Then he’s got to put the sticker over. Now here’s me pure cashier, you know, like this is 89. And then I never put supervisor because I got it wrong or, you know, because I had to type 99 where we today they failproof the process straight. There’s now barcodes. Is that now? I wish I could have been a cache. I could have gone and partied a bit more. I just had to scan. I just had to scan.
Bhavesh: You’d have to concentrate much, right? Yes, of course. Fail proof, right?
Brinsley: So STP really is…If the customer starts on the app or they start in the channel, try and design that so it doesn’t even go to the middle office.
Bhavesh: So it goes straight through and you don’t worry about it.
Brinsley: And have to wait straight through. Yeah, right. Yeah, because the more handles you are as you pass down the line, the greater the chance of a knock on an error human, yes. So really in the things that are high volume repetitive, you’ve got to go STP as far as possible. But where that’s not then can you wear is and can a robot help with some of the screen dumping of the things that are common and then you just perform and then the more intense we a Gen AI can. So once you understand what’s a candidate, you can then deploy your energy in the right place. Of course, we’ve picked up that most of our 7,000,000 minutes for Gen AI would have been effective in the regulatory because of the nature of the work. Yeah, so and the volume they’ve got to go through etcetera.
Bhavesh: I think it’s a very valid… actually everybody’s kind of trying to figure this out, right, because it’s a relatively new emergent technology. And, you know, there’s also, you know, the whole idea of the human in the loop to always check the Gen AI as well. So , we’re on the start of a long journey of a transformational journey at this point.
Brinsley: And really weird. I mean, if I look at our chat bot, when customers chat to us, etcetera, you know, lots of, you know, the Gen AI is sitting behind there, right? But the importance there is to for it to know and for you to learn of when the customer comes something off script or you know, that your call agent can intervene, you know, at the right point in time and that the workflow is there to say, OK, now I need a bit more higher thought, thought leadership or thinking and a human being must come in, you know, otherwise, you know, I might fight, respond or in a way that you don’t want to do anymore.
Bhavesh: Of course. So I mean, I think, you know, just kind of concluding what you’re talking about, what you’ve shared with us today. It’s been a fascinating journey to look at how… I love the attack of the minutes. I love those kinds of things that people can hang on to. And, you know, the way you’ve brought your staff along the journey and the transformational changes that you’ve made, What advice would you give somebody if they’re doing something similar? What would you say?
Brinsley: Yeah. So I think I said it early on, you know, once you’ve got your foundational data threads, right, you’ve got to know what it is you need to be tracking, right, What those key drivers are, ensure that you scope the information flow. So where must these data flow, where must it go to and who’s going to do what with it? Because generally it’s the strings of information that you tie together that gives you the insights.
Bhavesh: Which allows you to do this.
Brinsley: And it’s the insights that gives you the knowledge and tells you the story that gives you the AHA. Or you can make the wise decision from right? So ensure that you’ve got that continuum. And at the end of the day, have the big conversations with your team. That’s a big thing.
Bhavesh: Bring the team along with you, right?
Brinsley: I say, if you employ somebody, why won’t you trust them? It’s a fundamental, right? And guess what, if you are clear on the what and the why, they will surprise you 9 out of 10 times.
Bhavesh: It’s a very good, it’s a very good observation. I think we should close it off. That’s amazing. Thank you, Brinsley. First of all, thank you for the opportunity to have a chat with you. You could chat for hours. I’m not talking about this topic. You’re incredibly passionate about the journey you’ve been on.
I think that a lot of our listeners who are going through similar type of programs will take a lot of great nuggets of what you’ve shared with us today. So thank you so much for doing that.
So, all that remains for me to say is thank our listeners for joining us. Do check in, catch up some more of our episodes of Ops Game Changes to hear more fascinating insights from other leaders like Brinsley and how they go about tackling some of their operational challenges. But thanks again for allowing us to use these facilities. And it’s great to be able to do a podcast face to face rather than virtually. Thank you all, and bye for now.
OUTRO
Today’s Ops Game Changes podcast was brought to you by ActiveOps. We’d love to hear your thoughts. You can do that by sharing them on our Facebook, LinkedIn or X feeds, or you can comment in the comments section of the YouTube version of this podcast. And if you’ve enjoyed what you’ve heard, please do like and subscribe and check back soon to hear more valuable insights from other inspiring leaders in operations. Thanks again for listening and goodbye.
