MTN Group – 2023 Digital Award Winner, Source-to-Purchase – Transcript

Andrew Savage:

Procurement is always going to be a people business. You’re dealing with stakeholders, you’re dealing with suppliers, so there’s always that element of soft skills. If you augment that with the intelligence and the knowledge that data can bring you, you’re a very, very powerful procurement practitioner.

Announcer:

Welcome to The Hackett Group’s “Business Excelleration Podcast.” Week after week, you’ll hear from top experts on how to avoid obstacles, manage detours and celebrate milestones on the journey to Digital World Class^® performance.

Vin Kumar:

Hello, and welcome to the Hackett “Business Excelleration Podcast.” I’m your host for today, Vin Kumar. I lead the AI and Digital Operations practice here at Hackett. Today, we’ll be talking to one of our 2023 Digital Awards winners. My guest today is Andrew Savage. He leads the procurement COE at MTN. Welcome, Andrew.

Andrew Savage:

Thanks, Vin. Great to be here.

Vin Kumar:

So first of all, Andrew, I wanted to congratulate MTN and your team on winning the Digital Awards in 2023 for the source-to-purchase – a solution that you had put together. Again, I know you’re a repeat winner. Your team is working in complex environments and coming up with fantastic solutions to address that. So, again, congratulations on winning the award again.

Andrew Savage:

Thanks very much, Vin. Honestly, it meant an awful lot to us. When we got the notification of the Hackett Award, it was really a super proud moment for me. Given the strength of your organization, it really is a fine feather in our cap when we received the news about the Hackett Award, so thank you very much for that.

Vin Kumar:

Thank you, Andrew. So let’s begin, Andrew. Maybe if you can take a minute or so to kind of describe the challenge you were facing that you took it up to kind of build the automation solution for that?

Andrew Savage:

Sure. And I need to go back quite a few years, Vin, to be honest, and I’ve been running the procurement excellence COE for quite a few years. And I always had this ambition to build what I called at the time the procurement cockpit. And I wanted some kind of a system, which was really easy to use, which drove a lot of value through our procurement organization, and that brought the kind of data to the fingertips element to our sourcing buying teams.

So I had this idea, and because data was all over the place – different systems – we work with 17 different markets across Africa and the Middle East, different ERP systems, spend analytics solution, in some markets we have quite I would say relatively immature data processes, so not always clean data. So I wanted a solution to bring all of this data together, which would ultimately guide data-driven decision-making. That was sort of the key challenge that we faced.

People would have to use Excel. They would have to go into a different ERP system – a span-analytic solution – and they didn’t really have the real-time insights or the intelligence to be able to use in negotiations. So what we wanted to create was something that would enable us on the go – on the fly – to be able to analyze data quickly and to bring us that decision support that we needed through the data that we’ve got available to us.

Vin Kumar:

For the audience, could you give the complexity of the technology environment that you had because you mentioned you had 17 systems, but I know the challenge we had spoken previously Andrew was the technology and process is not so easy to run, not even just prepare for the negotiation discussions, but also what are the insights that you had from all the various systems and externals? So anything on the challenges that you had in your environment that you were trying to overcome?

Andrew Savage:

Yeah. So we have a lot of challenges at MTN anyway from a geographical perspective. We operate in 17 markets and a civil war in some of those markets. We’ve got a lot of headwinds from the macro environment perspective. And then from sort of the technology perspective, it was more about solving or bringing that data together to make decisions. So we got all of those different data points residing in different locations. So I could go and pick up a report from one location, but then I’d have to do some number crunching myself – bring those numbers together, bring that dataset together – to be able to come out with something meaningful that would actually drive a decision. So that was sort of the complexity of that.

We got different languages in certain systems – messy data in a lot of our systems, clean data in some systems – but it was really having that sort of central repository and access to data at our fingertips. And then also enhancing the data because what we were also looking for is kind of a competitive advantage when it comes to building negotiation strategies – so understanding how our suppliers negotiate, what sort of information we have available on our suppliers, how we’re using that information.

So, if I take, for example, within the procurement space, the reverse auction data that resided with each of our different markets, it wasn’t sort of brought together in one house, and even when it was people were just focusing on the output of that, which was basically a set of numbers, which said this is the final number that came out. So we then took that a step further and said, what do we actually want to gain, or what are the insights that we can gain from this type of information? And that brought us onto sort of looking into the behavioral analytics of our suppliers.

How do they react in a negotiation strategy? Are they waiting till the last minute to bid? Do they bid in increments? What are those increments like? And then putting those suppliers or the supplier data next to each other to understand how we would expect suppliers to negotiate, how they would react in certain situations, and ultimately try to be able to control that environment and change sort of the negotiation strategy and the setup of those actions to be able to create the most optimal outcome.

So it was more about the data being in different locations and not being converged together to be able to drive those specific decision-making support that we needed from a data perspective.

Vin Kumar:

It’s interesting, Andrew, that you have this procurement COE, which is our understanding is a more technology of solutions focused to bring it to procurement practitioners and the team. Coming up with these solutions and recent challenge, is the request coming from the procurement folks or are you taking solutions because they may not even know that this is capable? Are you taking solutions to them and saying this is what may help? How does that work to identify these type of problems?

Andrew Savage:

Yeah. So I think we were a little bit ahead of the curve when it came to employing data scientists within procurement, certainly in our region anyway. Our first data scientist joined about five years ago. And so we’ve been working on really advanced analytics – AI machine learning for quite some time now. And I also give my team quite a lot of autonomy when it comes to developing use cases. So I encourage them to go literally walk the floor, speak to the people in procurement and understand what the problems are.

And you’re right – most of the times nobody’s going to go to a data scientist and say, “I have a solution here that you need to be involved in.” It’s more developing through conversations what are those specific use cases, where are we missing data, or where are we missing insights that we can try and plug those gaps, and then that’s where we sort of start to develop those use cases.

And it starts usually with a small idea from a brainstorming session – from a procurement excellence COE perspective – and then we get bits and bobs from the actual procurement buyers from the buying teams. And then that’s where we start to create something, which is really, really valuable, which is how PX 360, the sort of procurement cockpit, came about.

And with that sort of development, we can just keep on developing these different use cases. And those use cases may be on an automation side, they may be on an advanced analytics, maybe machine learning. So whatever it is, those use cases… And we built a huge use case library actually, which talks to AL, ML – all of those different use cases – and we just sort of develop those, and we bring them into the solution wherever possible.

Vin Kumar:

And maybe that’s a great segue, and maybe if you can take some couple of minutes to explain what that solution – the RPX cockpit solution – looks like, and what the solution itself is and how that will be helpful for our listeners?

Andrew Savage:

Yeah. Sure. So PX 360, so this was what was the original procurement cockpit idea developed into PX 360, which stands for Procurement Excellence 360 – 360 being a 360 view of procurement. And that was created by the data science team using open source technology. They developed the initial first module for it, or the first use case that we developed was around a sourcing calendar. So whilst we had visibility on contract renewals, RFPs also in our markets when someone was running an RFQ or they had some demand, it was all sort of residing in different places.

So let’s say that the data insights that we were missing was the ability as far in advance consolidate those negotiations. So we built this sourcing calendar, which allowed us to very easily just put in a supplier name or a category or a domain, and it would throw out the opportunities for bundling demand.

And so we were very, very quickly able to have a negotiation, which involved multiple category buyers, usually the same vendor, and we would consolidate those RFPs into a single negotiation. So obviously we are benefiting from significant economies of scale by doing that. So that was the very first use case, which we developed.

And so we quite early on moved on from sort of static dashboards and wanted those insights to drive those specific decisions, which is why that came to the forefront, and it was so very quickly adopted and very successful. And then from there, we just sort of built more and more modules as the use cases started to develop.

So, for example, when we’re onboarding a supplier, we use our shared service center to onboard a supplier. They go through a due diligence check. What we enabled there was a very quick automated prequalification, so you put in some of the supplier key details. It runs against sort of 15 different checks. It looks at things like sanctions list, our internal watch lists, whether that supplier is already registered, whether there’s any negative news, etc., and it just spits out a very quick result. And then in the agile mode that we’re in these days, we know that we can already have… Or there’s no sort of major early warning signs with that vendor, and we can continue the onboarding whilst they’re in an RFP situation.

So those sort of use cases just keep getting developed, and then we just add them onto PX 360. Now we’ve got 10 or 15 different modules – everything from supplier news, supplier insights, through to the bundling opportunities, etc. So it brings it all into one place, and I can easily quickly log onto to the platform, search for a supplier, and I can get all of their information. So whether that’s spend savings, which tenders are ongoing, supplier intelligence, supplier news, all of that information is literally at the buyer’s fingertips, so it’s really, really useful.

Include that negotiation behavior or the behavioral analytics, which I can then use when I’m in a negotiation, so very quickly. If I’m talking to a supplier, I can already see what the org structure of that supplier looks like. I can look at what the spend is, what savings have been delivered, what ongoing negotiations we’ve got, what future negotiations we’ve got planned. So it’s very, very valuable for our procurement buying team to be able to tap into that whilst they’re in a negotiation.

Vin Kumar:

I mean this is fascinating that you’re able to do, and especially given some of the geographical markets you operate where getting this type of information is maybe harder. There is no consolidated source to go and get … for example, here in America, they may go to someone like a Dun & Bradstreet who would provide that, or a fax that would provide that external perspectives and stuff. So you are operating in some of these geographies where it’s harder if we get. Is there something you could share on the technology solution that you built this on? Any insights you can share there on, Andrew?

Andrew Savage:

The variety of markets that we operate in, and they range from low maturity levels from a procurement – from a data perspective – to relatively high maturity levels. So we’ve got a full range of maturity levels and that means in some of those less mature markets there’s less attention paid to the data, and therefore you get quite a lot of areas, so you need to do more cleansing, more automation to increase that accuracy, through to the higher levels of maturity, where we know that we can rely more on that data.

Now the technology that we use is just bringing those data sources together, so it’s bringing all of that data into a single data lake. Now what we can’t solve necessarily with technology is data, which doesn’t exist, and we’re never going to get away from the fact that some data doesn’t exist. We might be working with a supplier in Sudan, for example, and there’s very, very little data on that supplier at all, so I can’t just create that data, so what we need to do is be a bit imaginative with how do we get that data.

So we do a lot of supplier visits. We actually collect data from the suppliers. So whether that’s certificates, etc., we make sure that that data’s onboarded into our tool set so we can’t necessarily solve all of the issues, but whatever data we know exists. We also track a lot of data, which is open source as well, so if there is open source data we’ll pull it. We do use the likes of DMB and other operations like BVD as well. We use the Moody’s Analytics to also get data. So we are looking at sanctions watch lists, we’re looking at a variety of data sources when we work with suppliers and when we understand who their shareholders are, who the directors are, etc. But we can’t necessarily just sort of magic up the data if it’s actually missing. So we just need to augment wherever possible.

So the technology is a kind of a … it’s not the silver bullet, let’s say. It doesn’t solve all of the issues. So we’re always going to have those issues, which we just need to augment the data with whatever we can through manual collection. But then once we’ve got that, we can ensure that the rest of the organization has access to that data so that they don’t need to go and collect it again.

Vin Kumar:

And that brings to the point of I can see significant value that you have delivered to the organization, but one is we would love to hear from you on kind of the value you have delivered and has been recognized by Dirk, your CFO/CPO, and the rest of the organization, so it’s not only that the procurement function is becoming more efficient, but what is the impact it has given to the business from an experience perspective – from an effectiveness perspective? Can you share on how you’ve tracked and what the value you’ve measured that the PX 360 has delivered?

Andrew Savage:

Yeah. So a return on investment for procurement excellence functions is really critical for me. And I think there’s a lot of procurement excellence functions or a COE function, which is not actually generating an ROI or is actually even tracking an ROI. For me that’s very important, whether that’s time, efficiency savings, whether it’s cost savings, whatever that is. So we run a constant check on what savings we’re delivering to the business.

Now, if I just take the bundling, for example, I can look at any point across that sourcing calendar and I can see maybe 50-60 bundling opportunities. So we then task the buying teams to then consolidate that demand and deliver the savings, so all of those savings are separately logged as a bundling activity. Then if it comes to RPA, for example, we’re constantly tracking the man-hours saved, so huge, huge time-saving efficiencies through automation, but also for our go-to-market strategy as well.

If I think about the sourcing teams in the pre-agile mode, when they needed to wait for a full prequalification and onboarding of a vendor before they could start an RFP, for example, we go through, and depending on how fast the supplier is in returning their information, it could take five days, it could take 20 days, and that means that I can’t start my RFP, which means I’m 20 days delayed when I go to market. Now, if I can already run that process with a quick check and then I can run the onboarding in parallel, then I’m cutting out already 20 days out of my onboarding or my overall cycle time with my procurement.

So we are delivering significant value, and it’s not just the … you’ve obviously got the hard measurables, the hard KPIs in terms of the actual cost, in terms of the actual time-efficiency savings as well, but then we’ve also got that ability to be able to negotiate live using that information that I’ve got. So it’s a lot of missed opportunities if you don’t have that data in front of you. I’m in a very, very awkward situation as a buyer if I don’t know exactly what that supplier is already bidding for or what they’re potentially be bidding for in the future. So I think those elements will significantly help our buying teams just from an efficiency and an effectiveness position.

Vin Kumar:

But that’s sometimes tricky to capture the value, right? It’s the lost opportunity. It’s not that you need less category managers or buyers because it’s efficient. Obviously, this helps in the efficiency perspective, too, but what we have seen in procurement is the impact that you could deliver is maybe more significant than actually what the cost savings is there. But how do you kind of measure that lost opportunity? Is there, hey, if you didn’t have this, if I didn’t know … if you do it by case by case, is it more systemic or programmatic way of capturing some of that value?

Andrew Savage:

So two things there, Vin. So, one, on the bundling opportunities, for example, we send out those bundling opportunities. If they’re not used or they’re not converted, then we log that as an opportunity missed. We may not put a dollar value to that because we might not know what the dollar value is. Probably in time we’ll get better and then we’ll be able to put a dollar value against it. But we do always report if an opportunity or an insight has been provided to a procurement team and not used or not converted, then we will always log that and report that back.

And then, secondly, we also request feedback from the procurement teams as well in terms of when they’ve used the tool. So we’ll always do a survey to say have you used this particular tool in a negotiation, and then we always log the number of times that it’s been used in a negotiation. We log the value of that transaction so that we know these are all of the transactions or the total transaction value, which has been in some way, shape, or form impacted by our tool. So it’s been used on the fly. We can most of the time can’t quantify that, but at least we can say it’s actually been actively used and therefore adoption, because for me, adoption is key. If people are using the platform, then it has value.

If nobody is using it … and, of course, we look at how many people are logging in, when they’re logging in, how often, etc., just to ensure that we’re on top of the adoption. And adoption is a really interesting thing for me, because I’m a big fan of sort of marketing procurement in order to drive things like adoption. So we do constant educational sessions. We’ve got a new module. We’ll do a training session on it. We’ll take back feedback from the procurement teams, as well in terms of how can we improve this. Why is the adoption low in this particular module? What do we need to change?

And so we’ll constantly evolve this tool, and because it’s been built in-house, it’s very easy for us to modify something or very easy for us to scrap a module. And we operate in that kind of fail fast environment where we can develop something and then we can quickly sideline it if it’s not working.

Vin Kumar:

And it brings an interesting point. Is your solutions that you are building from the COE changing the profile of procurement talent that the firm needs or hires? It’s not oversimplifying it. In the olden days, it was connections. You knew exactly who the supplier is, you knew the guy you could just call and you did, but today it’s more a different profile, which is … and how is that impacting Dirk, and how he’s trying to change the talent pool of the procurement organization?

Andrew Savage:

That’s a really interesting question, Vin. I have quite an interesting take on that. So I think procurement is always going to be a people business, right? You’re dealing with stakeholders, you’re dealing with suppliers, so there’s always that element of soft skills – the ability to build relationships and deliver value. Now if you augment that with sort of the intelligence and the knowledge that data can bring you, and understanding how to use data to your advantage, I think you’re a very, very powerful procurement practitioner.

So about two years ago, we changed our operating model into a new agile operating model, and we introduced new job cards. We moved away from what we call job descriptions into job cards, and we look at something called the house of capabilities, which covers the procure-to-pay cycle of all the capabilities we expect to see, not for today, but for sort of five to 10 years’ time.

And so we built in a lot of the digital – the data analytics capabilities. We’re not getting away from the old traditional sense of what you need to be a good procurement practitioner. We’re just augmenting it and saying this is how we build the perfect procurement person. And we’ve spent a lot of time, and we put heavy emphasis on upskilling people and reskilling people when it comes to the importance of data – how to use digital tools. It’s inexcusable for us – for somebody – to enter a negotiation without a full negotiation strategy with all of that data in their hand to understand what the negotiation analytics are, what’s been done before, what’s happening in the future, so they can build that full picture. We like our procurement teams to be very, very well-prepared with all of the data.

So I think it’s very interesting, but I think it’s a real combination of the old school ways of working, or the soft skill side of it, and then the technical and the data analytics as well. You don’t need to be a data scientist, but you just need to know how to use these tools. And I think the benefit of PX 360 is it’s a very simple tool. I don’t need to be a data scientist to use it. I don’t need to have real deep data analytics understanding to be able to use it. We’ve kind of made it idiot proof just so that we can use it very quickly and not overcomplicate things.

Vin Kumar:

No. Excellent. I mean because I think by tracking the adoption, you’re seeing that is OK – how it’s used. And you’re right, they have the domain knowledge. They don’t need to have the data science. That’s what … your solution is making it easier and giving the insights, and they can see the insights and use it how they would like to best use it for the business outcome that they’re trying to achieve.

What’s the future? I mean what is in your pipeline? As you say, you are adding more use cases or more modules. What’s the future for this, and maybe what have you done in the last year? I know when we submitted this was probably first quarter of 2023. What have you done the remaining of the year – this year – and what are you seeing the product kind of life cycle for next year for the PX 360?

Andrew Savage:

So we’ve got a list of different modules that we’re planning on introducing, and I think since we made the submission, that’s where we’ve really started to build out those behavioral analytics and understanding which of the suppliers we’re working with – what their backgrounds are, who the people are, how do they negotiate – and really bringing that data science element in to augment this. So before we were getting the insights of what had happened, what’s going to happen, but this is really … sort of the next level is around the behavioral analytics. So it’s almost like a game of chess.

When I build my negotiation strategy, I can say what are my negotiation levers? I know what’s upcoming for this particular supplier. I can perhaps use some of those as levers. I might have some market intelligence around that supplier, but I’m getting it from all different places.

Now we have that very holistically brought together, so we’ve got very, very powerful in that. I can put in the supplier name and I can see who I’m negotiating with, what the negotiation tactics are, what the strategy is, so it makes me very powerful as a procurement buyer, and I think we will continue to develop that. We don’t have enough data to be honest. We do have a lot of data, but I think what we’ve not done is stress-tested this in different environments, because we’ve run a lot of reverse auctions, for example, but we would tend to run them with a kind of predictability.

Now when you want to see how a supplier negotiates under different parameters, you obviously need to change those parameters. Now we may not have done that in the past, so what we need to do is implement different ways of running negotiations to understand how our supplier behaves under those certain circumstances, so we can just build more and more data points, so, again, the more data we get and the stronger we get as well, because we’re just collecting more and more data points.

Things like spending time looking at the data during COVID, for example, is very interesting, because that’s a time which we’ve never really seen before – we don’t have data on. So seeing how people react, seeing how prices react within that sort of time period is very, very interesting for us and helps us to build. If there’s a future disruption like that, we already know what to expect in those circumstances. So I think we’ll continue to develop more and more modules based on the different use cases that we establish.

I know you asked a question about Gen AI in the beginning. I mean we have incorporated Gen AI. I honestly struggle a little bit with use cases when it comes to procurement. I do use the likes of ChatGPT myself occasionally, but not for very sort of advanced use cases. I think the next step would be to build our own Gen AI, and we’ve developed sort of things like chatbots internally, but it’s a little bit more gimmicky than actually driving a lot of value. But I do think that that will be sort of the next step, will be to really develop the Gen AI on our own data. I think that’s probably the most logical next step.

Vin Kumar:

The procurement COE has been really valuable for the procurement organizations. Do you see this kind of model … has it been, or do you see other functions adopting this type of model – be it in finance or HR or legal or any other kind of your SG&A function in sales or service, adopting this kind of solution/technology kind of COE for that function where you have the domain and you’ve got the capabilities to deliver these type of solutions?

Andrew Savage:

To be honest, I’ve not seen it so much. I mean there’s a ton of use cases where kind of a COE can support those business functions. I think more on the sales side, where you’re actually generating revenue, and you’re looking for sort of the insights to negotiate again. But when it comes to finance, for example, I mean it’s prime for automation. There’s so much automation, which you can deliver in the finance function. So it would make complete sense to have a center of excellence.

Now our centers of excellence are more on … they’re more holistic across the entire group. So you maybe have a software center of excellence, which is looking at sort of the move to cloud, for example, or a data science COE, which is covering multiple parts of the telecom business, but you’re also looking on the sales side there as well.

Obviously, within procurement it’s relatively popular, I would say. But, again, I don’t think it comes back to the ROI. I think a procurement excellence function can be very easily seen as just putting the policies in place – just putting the systems in place – but not really using their tools and their skill set to generate an ROI, which is all around driving their decisions. So I’d like to see more and more COEs, and spending more time really upskilling people, identifying those use cases and developing those use cases.

So I think their scope for COEs – whether functions would actually go down that route – is yet to be seen from my perspective, but I see an awful lot of value in this. If you can have an enablement function, which is driving and really cracking the whip on people operating at their optimum, I think a COE is perfect for that.

Vin Kumar:

Again, thank you. This was really insightful. I’m sure the listeners are going to see, and I know we’re going to be including some of the links to your actual case study and all of that too in this podcast. Plans for 2024? Are you planning to apply again for a Digital Award and be three times in a row? Is that in the future here, Andrew?

Andrew Savage:

Yeah. Absolutely, Vin. I mean I would love to. I always say with awards, for us, it’s kind of a validation. I use it as a benchmarking. When I look at the sorts of organizations, which are awarded by yourselves, I look at the interview process that you guys go through … I was even talking about it today with the team about how sort of intense it is, and the interview process for your awards is really good. So it tests us. It’s an external benchmarking for me. If the team is delivering something which is winning awards, it means that externally they’re doing really well.

Internally it’s great validation as well because our stakeholders look at it and say these guys are actually competing with market leaders out there from another procurement organization. So that gives them the comfort that we’re actually delivering something that’s also world class. So for me, I’m a huge believer in awards – all that validation.

And, of course, it’s a real hook for talent as well because there’s a talent shortage, and it’s always difficult to attract talent. So if you’ve got awards behind you and somebody picks that up and says, “Oh, wow, they’ve been awarded for something by The Hackett Group,” that’s fantastic. People want to work there. For me, it’s a big tool for attracting the best talent as well in the market, and also sort of the next generation of talent as well, because it is all about the digital future, and how I use my tools and my skills around data and data analytics.

Vin Kumar:

Hey, Andrew, again thank you so much for taking the time to do the podcast, and again, thank your team, your CPO Doug, for kind of investing in this and continuing to invest in the COE there, and wish you guys all the success for 2024.

Andrew Savage:

Thank you, Vin. Much appreciated. Thanks for having us on again, and, yeah, I look forward to catching up with you soon.

Vin Kumar:

Thanks, Andrew. You can find more about the case study, the Digital Awards, what’s the application process, what’s the solution, and the case study that MTN implemented and the value they delivered on our website, thehackettgroup.com.

Announcer:

Thanks for listening. You can find audio, a transcript of each episode and other resources at podcast.thehackettgroup.com. If you liked this episode, please share it. You can also subscribe on your favorite listening app so you never miss an episode. We welcome your feedback by rating this or any episode, or send us an email at podcast@thehackettgroup.com.

The Hackett Group is a leading benchmarking research advisory and strategic consultancy that enables organizations to achieve Digital World Class^® performance. Learn how we can assist with your transformation journey at www.thehackettgroup.com.