PegaWorld iNspire 2024: Capgemini and Wind Tre: Next Best Action Implementation and Adoption

Thank you. Five seconds of introduction. And so who I am. Daniele Mazzoni, head of product product development in winter. At the beginning of this program, I was in it. And so today I will have two heads of product development in it. Then the head of Customer Engagement and Value management, with the responsibility of managing customers in order to engage them and make their value as much higher as possible. The Grand Canyon Head of Telecom segment leader in Capgemini Italia, partner of Winter on this transformation. Okay.

So before we start with the presentation, I take five minutes to introduce where we were at the beginning of this program, and so to be effective, probably the best thing to do is to explain what happened. And so it was late 2021. We were merging. We are completing the merge between three and wind, and we are moving all the customer base of three into the stack. The final stack. The stack. Stack. Stack. And I receive a call from her and her boss and say, okay, since now we have all the customer base consolidated, we want to change the way in which we communicate to our customer.

First of all, we were in the middle of the of the migration. And, you know, it was we want to change. And we wanted to have a different approach, an approach that is real time. I was 18, and so I you see Mirko and Veera. And so we consolidate the data once a month. You can run a campaign each, each day. I think we have 400 of ETL that moves data here and there. So I think it's quite difficult to have a real time. Daniele.

You have to think of that. And for me, Mirko is impossible. Well, Mirko said to me to imagine that it's possible it's not the only requirement that we have. So we have another requirement. That is, once the data is there, we want that the channel changes each time that the customer does something. And so I say okay, so you want real data and then change the data when the customer is there. And so with our CMS that is a static CMS and so on. Yes. Impossible.

And so one impossible to impossible Third. So. But it's not only that we would like to have an experience that is omnichannel. And so with all the channels I say guys, no way. We have four catalogs that are for all the channels that we have. It's impossible to merge them. It's impossible. Three impossible. Okay, let's start the journey and let's see what the team that was a team composed by Winter, that was done by it, product development and commercial.

I want to say that in this program, we had more or less more than 100 people involved. Okay. And then there was a partner with Capgemini that helped us as an umbrella, let me say partner that helped to to run the program together with Pega, obviously. And then there was also the help of McKinsey see that and you will see in the presentation. Help us to reshape our organization. Okay. And so what we are going to see, we are going to see what we did and so on, where we started, what we are, how we approach it and what we created. Elvira. So who is winter?

So winter was a telco operator, 100% owned by Hutchison, with the biggest client base on mobile in the B2C market, with a revenue share of 25% running two brands Winter, which is the brand targeting families, and the very, which is the no frills digital brand with a footprint in terms of shop of five 4000 point of sales covering all the country. We offer wireless and wireless connection plus selling. We also sell sales devices to support connectivity so smartphones and modems. But in the last two years, we decided to change our strategy, switching from being a simple, simple telco operator. We decided to become a multi-service operator, entering in new markets like energy security and insurance. Why we decided to do that? Because the Italian telco market is an industry decreasing in terms of value year by year. For the mobile, you can see that the now the average rpu of Italian customers is lower than €9. That is more or less $1,010, dropping by 20% in last years, with the usage increasing a lot four times in terms of data traffic.

Why this trend? Because Italian market, even if the population Italian population is more or less 60 million, is a compete with the 27 operator, three main and the others are low cost or second brand competing based on price. So pushing down prices. This operator sell bundles with the cost lower than €10, with the amount of giga embedded more or less of 200, so mainly unlimited. The market is mainly prepaid. People are without any kind of commitment. They are free to live with mobile number portability in 48 hours, and the retention is not allowed from regulation on the fixed side is almost the same, even if the spending is quite, quite steady, just slightly decreasing. The number of people connected with fiber increased by two times. So we have lost the possibility of monetizing the technology upgrade.

And that is why of this for the competition, because there are 13 players competing thanks to a wholesale agreement and providing fiber at a price starting from €2,020. So what we decided to do, of course we decided to change our strategy. We were aware of having three important assets, even four assets. First of all is our base. It is a loyal base. And they they rely they rely on us for for one of the main need of everybody that is connectivity. We have a footprint of 404,000 shops able to get in touch physically with people, providing assurance, providing support. We have a brand equity very powerful, so that we made some researches telling us that we are reliable and credible, also entering in new markets, and we are used to compete in a very dynamic market arena so that we are fast, we are flexible so that we can enter in this market that are not so used to competition becoming, bringing disruption. So which is the three pillars of our strategy , first of all, to be simple, to bring simplicity to our customers, to make our customers, but also our A call center or sales agent.

Make their life simpler to be relevant. So we want to be able to tell customers to bring to customers exactly what they need at the time they need them. And we are multi-service. So we are now targeting families, providing not only connectivity, but providing energy, providing security solutions and providing insurance or covering something that is around the 15% of family spending with our products, which are our main goals. First of all, increase loyalty because as I told you before, the customer base is our asset and it's important to preserve reducing as much as possible churn. The second target is increase cross-selling. So increase the share of value we have from our families, switching reaching from people to households. So targeting households and trying to get the value not only from our customers but from a relatives of our customers. And the last point is to increase the penetration of new core on our base, which are our levers, first of all, customization, personalization in terms of products, in terms of price, in terms of contents, in terms of communication, applying the generative AI to build a customized and compelling sales pitch, able to maximize the conversion of any single interaction.

The second point is the contextuality. So be able to understand when a need something happens, able to reveal a customer need, and then to have the possibility the capability to act in real time. Targeting at the customer, with the proper solution covering the specific need. And the last one is to be fast in deploying the innovation, bringing to market our products and our services. So this is this was our vision after the call and after trying possible you know marketing guys typically push push till bring it guys where they want. And then this is what we have in life okay. And so they are impossible. But with with the marketing that was pushing. And so let's see what we realized.

And so I would like to divide into two parts. One that is the front what I call the front stage. And then the backstage. Let's see the front the front page the front stage. And so you remember the three. Impossible impossible, impossible. Okay, let's see where we. should transform it . Impossible.

And so on. First of all, we had we started to think to the customer that has the possibility to have the content that were the same for any kind of, of channels. And so we deployed and also the offer were pushed by the machine with an orchestrator that was common to all the or to all the channels. And so we started with the, the app. Then we went, went, went on with the dealer station and then with the agent station, but all the channels and we will see also other small small channels. But all the channels are orchestrated and see the same stuff everywhere, each customer. And anytime that the customer does something that is relevant for, let me say, the orchestrator, the customer will see in all the touch points , everything that is different. And so what is the best proposal that we can do is immediately propagated not propagated, seen in the in all the channels. And so this is what we when we said that we want something that was fully personalized.

Okay. So the first impossible became possible . Let's see what is in the back end. This is the transformation that we did. Let's start from the bottom where we started. We started with all the data that were present in our Google Data Platform part. And so we have a GCP that were that were already with all the data. All right. Starting from that we harmonized harmonized what was the data model.

Right. In order to feed the the second part. And so the data platform have all the data. And then we move. And we also integrated all the data that were relevant data in real time data in order to trigger the personalized offer. Okay. And so the second impossible became possible. And so in real time we intercept what is an event. And we create immediately the trigger to offer to the customer something that is new.

The orchestration layer that is the orchestration layer that is always on. And so we have a unique brain that works on all the channels always on. So for the inbound and outbound okay. That manages what we manage, we configure more or less more than 1000 actions. Wow 1000. 2000. 2000. And so let me say it's not a very huge amount. We integrated all the channels and we will see later.

And then we did also the part of the content in another way using Adobe. And so we create a personalized view. Okay. Immediately. As soon as the machine proposes the the action to the customer. This is valid for all the touch points. And so the most important web app posts and contact center, but also for other channels that are outbound channels. Okay. So this is the second from impossible that became possible.

The the second possible that that we did. Okay. Let's move to the last. The last possible. Impossible become impossible. So as as Daniela said that Pega is in the center between our product catalog, translated in more than 1000 actions and our touch points and our customers. Doing what? So, starting from the whole catalog and applying the engagement rules, we are able to bring to our customers a fully customized shop. So any customer has its own shop based on more or less 100 actions that fit perfectly the customer profile the customer needs, what the customer really wants to get from us.

Then, thanks to the arbitration, we are able to rank this action based on customer propensity based on action, value based on business priority. In order to highlight the three five best offer to recommend this. This recommendation is directly displayed to the customer on the app, but is also provided to our physical channels like call center and shops, so that as soon as the the customer called the call center or enters into a shop. The agent is able to see which is the recommendation, so that in order to to to check to rank directly all the possibilities. He also knows which are the goals. So the easy things to sell for that specific customer and all the touch points provide feedback to Pega so that they are able to feed the propensity models, training it in order to become more and more effective and efficient in predicting customer, customer needs and customer interest. On top to this, we also have our triggers so that we have a platform able to collect data in real time. Transform this data in events, revealing a specific customer context that could be a need, could be a risk, could be an opportunity so that we react in real time, pushing something that is relevant for that specific context. This solution is completely flexible and future proof because when we want to add new actions or when we want to push a new a new business, we can do adding action to the to to pega.

All this has been possible shaping the organization in a way that is effective and coherent with this new way of working. Apart from the product factory, all the blocks are in the commercial area. So this is our first step in martech application because we are independent in managing the CMA and we are also managing directly some pieces of operation. From one side we have the marketing proposition. So people responsible for building the proposition for any, for any persona. So based on our personas. Then they pass us. They provide us these products, this proposition and the customer engagement. That is my team.

Help in in putting this this proposition in the system, in this central orchestration and cascading on all touchpoints across all stages of the customer journey. We work in agile. We have squads that are responsible, accountable for a single stage of the customer journey, grouping all the skills that are needed for managing end to end processes. So they are independent from the concept up to the delivery, but they are orchestrated by a central team, which has the responsibility of providing coherence across all the action defined by the squads, and checking if the taxonomy that we have to build in pega is is correct in order to let the system work properly, taking bringing synergies, including which are the most relevant skills for making this activity that are CMA designer. So people configuring directly in the system, the action data scientists and our editor who work on Adobe for building the fragment that are required for making the contents that are available on our touch points. And we work directly with the channel factory, both for digital and physical channels, in order to be sure that channel are totally able to get our activities and to bring them to our customers. So let's celebrate. Let's celebrate our journey. Yes.

Let's see how we arrived here. Exactly. So we left late 2021. Okay, obviously time for Retender because we have to decide what was the right product and the right partner with which we wanted to run this program. We decided to do with with cap. And the cap was really helpful to make it happen, because with this data, only if there is a big partnership between the teams you can do, otherwise you cannot do. And this is important. And so let's start and see what are the. So I signed the tender six months to build the foundation.

What does it mean to create and to to create the heart of the of the system. And so with the data, with the logic, with the logic of feeding the channels. And so the first six months were spent like this starting from January the 1st. We also introduced we deliver the app because we decided that the first channel was the app. Imagine that during this journey, her team had to do the double work because they had to feed the whole campaign system, right? Because obviously, guys, at the end of the day, we are a company. We are not. And also we need to do money and we need to have the the whole campaign running in parallel. Right.

And so we decided to start with the app. We started with 10,000 customers. And then we ramp up and we complete totally the app in May because it was May. In parallel we run also with the with the POS. And so what we call the POS that is the dealer station. The dealer station has. We took part of the action that we already delivered for the for the app. That was all the customer base part and we added all the parts. That was the gross parts because we are managing not only the customer base with this application, but we are managing also the prospect and the gross sets.

This is important to say. And so in six months we arrived to complete the also the dealer, the dealer station in this case. Even in this case we decided to to have a ramp up. That was a smooth ramp up. And so I remember that we implemented a button with which the dealer could go back to the old to the old POS. All right. And so to give the possibility to come back in case they found that it was something that was that was wrong or right, in any case , we arrive also to complete it. Call center was faster because all the action that we deliver for the POS, all the action that we deliver for the app were obviously useful for the contact center. That has only a few specific actions on top of that, and we deliver very, very fast into drops 30% one week the week after.

Since there was no disruption, 100% of the call center was onboarded in parallel. Always we optimize the machine because obviously we started to use the machine. We see the behavior, we correct the behavior, we introduce new stuff in order to have something that could bring more value. That was in the pipeline. So everything was in the pipe and we implemented it in the during the time. And let me say we we see at the end of the journey. But it was even in this, in this case, something that we run in parallel. We introduce the contextual, what we call sophisticated contextual triggers that are the triggers that tar. All the offer that we give to our customers that are linked to an event, that is an immediate event that we incorporated in, in our in our system.

Okay. And so more or less we are talking of 18 months, right. That was very tough. But we did it. So. Impossible. Impossible. Impossible. Possible.

Possible. Possible. Plus, I would say because at the end of the day, we are very satisfied with and proud of what we did. Okay. Let's see some numbers that celebrate. Celebrate. Let's celebrate. Let's celebrate what we did. We are still learning.

Exactly. We are. Still learning. We are still learning, driving this very sophisticated machine. But the earlier results are very promising. We are fast. We are able to bring to market a new proposition in 24 hours. So we are very flexible and fast trigger based campaigns. So campaigns delivered based on events has a click rate that is three even five times higher than standard.

And in our shops, where there is the mediation of a human between the company and the customer, next best offers or offers recommended have an adoption rate that is two times higher than the other action. Good, good. Now I put the head of the IT guy. Right? What we did on the on the IT side. On the IT side together with Cup. We did the following. So you remember we had a catalog for each channel. We have only one catalog now that is feeding everything.

Otherwise we had also the problem with the business rules and so on. And so one catalog content management. We had five content management because as it's quite common in the telco and telco market. And it tended to. You have a new channel, right? You put the catalog and CMS on top of that. And so from 5 to 1. All right. What we did that is very, very important.

We simplify a lot the way in which we manage data. And we remember one month to come to collect the data to, to to send the data to the campaign and so on. Now we remove 400 of ETL that were moving data, and we have three centralized process that manage the data and feed the system. And so I would say that even on that side a big simplification on top of that is not written here, but it was part of the strategy of everything in cloud, because at the beginning of the, of the, of the, of the project, half of the system were on prem. Alpha was on cloud. We push everything on cloud. Okay. And finally. Finally.

Let's see what a lesson learned. Because we say that it was a journey. Tough. But it was a journey. The first recommendation that we give to everybody is it's not an IT project. So when you move from a traditional campaign tool to something that is sophisticated as pega, you cannot consider it as a mere transformation. It transformation. You have to start to think together with the business from the very beginning, because the foundation. Do you remember the famous six months that we spent for the foundation is the core part of the program?

Okay. If you start, well, then the life is easier. And we know because obviously we did not so. Easy, but. It's easy, right? It's easier. But in any case, because we did a lot of mistakes, we did a lot of mistake, probably taking a little bit more time to understand how. That was not only a native program. And so just to say, okay, do and transform, probably we had gained something in the, in the last, the last part of the program.

Another important thing that you choose a partner for sure to implement it. But it's important that together with the partner you understand what is pega okay. Because it's relevant that you together with the partner and with Pega, understand how to make the system work. It's very, very important because otherwise they run the risk to go in one direction. And then you find that that this system perform much more better in another direction. Okay, so take your time. As my mother says. Study, study, study I tell to you. Study, study, study, study together with your vendor and with Pega before you start everything.

So that this new approach is completely data driven. So data are the crucial key element for the solution. So you have to assess to have a very, very powerful assessment of which kind of data you have, which kind of data you need in real time to define a data model that is able to let Pega work properly in order to avoid to waste a lot of time bringing an unusual or useful data, or bringing some very, very tough layers that are very consuming in terms of computational National effort. The last point is related to people. It's important to bring people on board in this journey, and to make them aware about the possibility we can unlock with this new solution, because they have to completely change their mindset. They are, well, working. It's very important that they to reskill them, to provide them the capabilities of learning to drive this new machine. And if I may, I would add the last bonus in terms of takeaways. It's much faster to to to make this journey with the with partners with the you have to choose the better partner to to support you.

We have done choosing McKinsey from the business perspective, but also choosing Capgemini in driving us in guiding us in this process, in providing the umbrella to support us during this transformation. So I like to try to explain how. Well, first of all, I'm really happy that we let's say we have chosen us Capgemini as a technological partner for transformation. When? Because when Daniela said impossible become possible, it's really happiness that some contribution to this possible Capgemini has done for sure. I mean, it is a it was very complex project actually, and I could even say one of the most complex projects we did with the client, actually. And so part of the, the, the whole architecture that we. So we jumped into this project from very beginning and we worked with Architecture integration part and Pega was is the, let's say, the key component of the system. But also we work to complete the whole contextual marketing ecosystem.

There were other components like content, digital content management, the social media management platforms and the rest. So we integrated all this with CRM, with data platform, the other components. And the other point is very important that we colleagues already mentioned, and also we contributed in transition, because it's not only the question of rollout of new platform, because you enter in the project when we already have the processes, the systems running as business as usual. So you introduce the new component As a system integrator, you need to help also, from the technological point of view, to to go faster on the new one and also to guarantee the coexistence of the old one and the new one as well. So there is a really we speak about the number of actions, and I really think that we never should. We always should try to go beyond the limit of the product sometimes because the standard implementation implementation is the standard implementation. But if the client need is to have some specific use cases that they have in in mind, of course we need to with the help of also the vendor, we need to expand the product to understand how we can be in line with this expectation. Adoption That very important topic because you know you don't. You need a part explaining the new system, how it works, etc.

, etc. but it is very important. As Elvira said, there are also topics like data science. So because the company already have the data science culture that you need to integrate into the new platform, and this is very important and also adaption in terms of changing paradigm between IT operations from IT operation driven business to the IT business operation business driven. Provided US training. You guided us exactly making new actions. So thank you. Can I say what is the result of this? Consider that nowadays after I think six months, the internal people were internal.

People were able to do everything by themselves. Really, really. It was an achievement that is very, very important. This means that there is a partner that helps you to understand how it works. There is inside the people that wants to change their mind because it's a new job. But in six months we were totally independent. And the last but not least is continuous optimization of the platform, because it's not only the work you do before go live during the go live, but also you need to continue continuously to to improve the performance from the technology point of view, because we were technological partner, the KPI, technological KPIs as a response time, for example, but also to try to optimize the platform in a way that the Google consumption also go down. This is also important. So I think complex at the beginning, but at the end, all together we work very well.

And I think that this was also the main key for the for the success of the story. Okay. Okay. And well, to to to do all this, we have the developed Pega practice. Pega is one of our strong partners. Probably you already heard that we we we nominate. We we won also the partner of the year. The communication of the year. Winner of of these days we have some big number of professionals architects certified.

We have accelerators to help to do let's say to to execute the projects. And we have solutions not only on Pega, not only in technology and not only in telecommunication industry, but also in the, let's say, in all the industries that Capgemini operates. So it's a finance, manufacturing, telecom and and also public sector as well. So I will leave a little bit time for the questions. Okay. If somebody you have microphones on the left. On the right. Remember that we are not mother tongue. And so questions low and okay.

Hi there. Thanks very much for the insight and presentation. Quick question for me when you were saying study, study, study and then six months to six months to be on your own parallel running on the different systems. How did you do balance that out in terms of additional resources or leave the b.a.u. Like under a change freeze? Yeah. About how you were basically what was the effort I suppose, in terms of study, study, study. Good question. Working harder.

Good question. Yes. To run in parallel, obviously you have to squeeze your organization. All right. Because you have. it's important that you don't focus on what you are doing and so on. The business as usual. All right. But you you must take the person that is able and capable to, to use the system.

And so they knew that the business rules and bring in the program. We had also the chance to have a partner that helped us to also feed the resources. But for sure that is very, very is, is is difficult to manage. As a matter of fact, as Elvira was saying, there was a people that went crazy during that period because obviously they have. To work very, very hard, working ten, 12 hours per day and asking people to do to do this extra work. That is why it is important to keep motivation, motivation as high as possible, to make them aware of what we are, where we are doing, why we are doing this kind of transformation, which possibilities we will have thanks to this new approach, so that they are motivated to to work harder. Motivation. Motivation. You must you you have to to to to explain to them that they are part of the program.

So they doubled the their energies, at least in Italy is like this usually. But I think it's all over the world. Also, just maybe add from the experience that we had as Capgemini. Also, Elvira said at the beginning, I think it the the implementation of Pega platform from it point of view, usually the effort that the client spend is on the business side. Adoption is the, let's say the the effort that you need to because technology in some sense. With some effort, additional, etc. it works. It works very well, but the adaption that people work in one way and they need to switch and start to work another way. In all projects that I heard with our clients , the adoption was the main, let's say, point of attention.

Okay. Other question. The same. We are not mother tongue in five minutes. Me neither. Okay. Okay. So in Italiano. No.

I'm joking. From. From Poland, actually. But one question related to your architecture. In the middle of the picture, you had pega as the brain. And Pega obviously has its own adaptive models. But on the lower side of the picture, you also had additional gen AI driven models. So could you elaborate a little bit of where you're using Pega adaptive models and enhancing them with some other AI driven models, or how does it work? What's the relationship here?

Then I leave the floor to to to Elvira. But you have to consider that when we start this program, we were not, let me say, Greenfield, we already had our predictive model. We had our all, let me say the KPI, the business KPI and so on that we already had inside the winter that were realized and maintained by an internal team, that that is our data scientist team. Right. And so we didn't have we didn't start from scratch on that side. All right. And so for sure, part of the cleverness is on the Google Data platform that feed Pega. All right. And part of the data and let me say the way in which you orchestrate its impact.

All right. But then I leave also the floor to to to Elvira. It's a mix of both. We decide to do things where are much more effective in terms of resources. So we have a team of data scientists and they build some kind of models like churn predictive models, like things like that. They build they bring on pega. They use this model to feed back a propensity, so that using the outcome of these models in the predictive variables on which pega propensity model is based, and we use pega when we know that pega is much more effective than our internal solution. So we balance both. Okay.

Okay. Thank you. Okay. No other question. I thank the crew. Thank you. Thank you.