- Introduction: meet Kwame, a 15-year veteran in campaign management.
- Career journey: from studying communications to mastering CRM at Vodafone.
- Database marketing: the thrill of working with customer data at banks.
- Tools galore: discussing the ins and outs of Pega, Unica, and Mautic.
- CVM marketing automation challenges: data quality, campaign runs, and ensuring flawless execution.
- CVM ops: the unsung hero of campaign management.
- Implementation woes: migrating tools is a marathon, not a sprint.
- Teamwork: importance of onboarding and managing people through change.
- The proudest moment in campaign rollouts.
TRANSCRIPT
[00:00:00] Kwame: I would say it’s much more a people management thing than a technical management thing. 4 or 5 weeks and then all the connections are made and you have your data and whatever, but then you need to start migrating people. You need to start working new processes, new this, new that, new those kind of things that that is what takes the time. The only thing which keeps you back is your imagination. If you can think of it, you can build it.
[00:00:22] Exacaster: Welcome to CVM stories, the podcast on Customer Value Management. Together we explore how companies can be more successful and the customers happier through the use of latest customer value management techniques. Learn key commercial and analytical insights from telecoms, retail, finance and other industries that drive CVM forward.
[00:00:41] Egidijus: Hi, I’m your host, Egidijus. Today our guest is Kwame, a marketing automation guru. Kwame worked with majority of big marketing automation tools. In this episode, we will dive into deep technicalities of campaign management from processes automation to data quality management to choosing the right tool for the job. It’s a practical episode packed with real examples, deep technical insights and learnings. So let’s meet Kwame. So hi Kwame, thank you for joining us today.
[00:01:20] Kwame: Yes, thanks for having, uh, thanks for having me.
[00:01:24] Egidijus: Uh, so, Kwame, you have been working in campaign management field for the last 15 years. You have probably seen it all done. It all failed, uh, in all, uh, ways possible. So what makes you so passionate about this field? Customer value management. Customer relationship management. Campaign management. Why are you here? You know, for those 15 years, what hooked you?
[00:01:53] Kwame: Well, um, I think for me, it is just a big puzzle, uh, which I get to solve every time. And, um, I like to solve puzzles. Uh, so basically, my work doesn’t feel like work. It feels like I’m getting asked to solve a puzzle. Uh, and to make stuff work. And at the end of the day, uh, at the end of the day, I get, uh, paid for it. Uh, which is like. Yeah. Why why are you why are you paying me for this? I just like doing what I do.
[00:02:34] Egidijus: So it’s like, you know, the same thing I think about basketball. Why do, uh, those basketball players, uh, get paid for? You know, they do what they like and just kind of get salary? Yeah.
[00:02:46] Kwame: Basically. Yes. And, um, yeah. And I think that if you are passionate about what you do, then. Yeah, then then you’re just you’re you’re you’re you’re automatically always take the extra step, the extra mile, uh, and you deliver the, um, the best result. Uh, but yeah, in the end you it is work, but it doesn’t feel like work. So, yeah, that’s basically what kept me, uh, in the game for the last 15 years.
[00:03:17] Egidijus: Um, Mhm. So, uh, could you share a bit of your career path, you know, where did you start? What? Uh, which companies did you work in? Uh, which industries did you touch? Because you have a very diverse, um, uh, perspectives there.
[00:03:32] Kwame: Yeah, yeah. So, so, so basically, um, I think like most people, I just ended up in the game by accident or in the field by accident. So yeah, I studied, uh, I studied communication science and, um, uh, which is a commercial, um, study. Um, and I learned about, uh, marketing and, uh, yeah, the how how do you communicate your message in the best way, those kind of things. And actually, after my, uh, graduation, I started working as a, um, a market researcher. So I did that for one and a half, two years, and I noticed. Yeah, I wanted to be a bit more technical, uh, the work that I’m doing and, um, uh, and then I started looking around, uh, for, for a job opportunities. And then I noticed, um, database marketing. And I thought, oh, wow. Uh, well, working with a lot of, uh, data, um, sitting behind the desk, uh, making calculations, um, I think I think that suits me better than doing the market research. So, uh, I applied for the job, and, uh, well, finally, I got it. So. So that’s basically how I got, uh, how I got started. Um.
[00:04:55] Egidijus: It’s funny. It’s funny part, uh, sorry for interruption. It’s funny. It’s funny part that, uh, those days, uh, personalization was called database marketing, you know?
[00:05:06] Kwame: Yes.
[00:05:09] Kwame: Yeah, yeah. And it’s also funny because the it has many names. Right. So, so sometimes they call it email marketing, database marketing, marketing automation, um, lots of tools. But basically you’re just touching data, which is in a database.
[00:05:28] Egidijus: So, so so you got first job, uh, as a database marketer. So what hooked you there?
[00:05:37] Kwame: Um, well, it was at a bank, and, um. I actually remember that, um, uh, because being a bank employee, um, I could look into all data for all banks, of all people. Right. So I actually, I told my, my manager, I like I’m feeling like a voyeur. I can see all of all these things. And that’s how I’m looking like, um, I don’t know.
[00:06:05] Kwame: If he’s the best guy for this position. And I was like. And then I was. Oh I shouldn’t have said that because yeah it wasn’t like like I was. Nosing into the data but it was like oh I can see. All these things. Wow this is awesome. And the the times before GDPR, you know. Yeah, it was before GDPR. Yes. You could say anything and, uh. But after the first, the first few days, it was like, ah, well this data again and, uh, it just, it just it’s just data, right? Um, so I.
[00:06:45] Egidijus: Understand that you kind of were, was very inspired with the data part, and then you basically started moving probably from from the banks. Uh, where did you move next?
[00:06:58] Kwame: So so basically. Yeah. So, so my first job was at a bank, uh, three of those bank in the Netherlands, which is a really small bank. Uh, it only has, uh, like 3 or 400,000 customers, which is really, really small. Um, so what? What? Yeah. So it was really it was really a free organization in the sense that, uh, they said, okay, so, uh, we need to have this result and you can organize it in any way that you want. So, um, I worked with SPSS, uh, there does um, and um, I learned how to, how to program in SPSS, which was really awesome. Um, and uh, um, did lots of lots of other automations. Um, and after, let’s say, two and a half years, I noticed. Okay. So, um, I want to, um, I want to keep evolving myself, but this is not possible where I am working at this moment. Um, so, for example, um, at the bank, they had a monthly, um, uh, interview with customers. Um, and those, uh, results were typed out. Um, and I came up with a way to classify that using a Python script, and that was a bit too far because they thought, yeah, we don’t want to go that way. We, we want to keep things manual. Um, and, uh, the work, some other kind of sums. I basically I hit the boundary of my growth there. So I thought, okay, let’s see if it’s possible to, um, to go to a different organization where I can continue my growth path. And then, I, I saw a basically a job opportunity at Vodafone, and, um, I applied and I got, got the job as a, um, CVM, um, um, CVM marketeer, um, and, um. Yeah. So I basically started working there and, um, uh, there I really got to know, uh, marketing automation, um, and basically everything I know of marketing automation was the basis was laid there at Vodafone.
[00:09:21] Egidijus: Okay. So, uh, and from Vodafone, you, uh, if I understand correctly, you had some other career jumps and, uh, you also touched in different, uh, industries there as well. Yeah.
[00:09:39] Kwame: Uh, yes. So, uh, so Vodafone is a telecom, of course. Uh, later we got, uh, Ziggo, we merged with uh, with, with Ziggo. Um, and after um, I’ve worked there for over uh, let’s say 6 to 7 years. I, um, started my own, my own company, and, um, I had customers as, uh, bol.com, which is a major retailer, a major online retailer. I also worked at, uh, Easypark, uh, which is automotive. Um, and um, um, also did something for audio also, uh, um, Telecom. And my final assignment, my last assignment, which I recently finished, was um, at Holland Casino, which is in the, uh, gambling industry. And uh, currently I am, uh, setting up a, an agency, um, uh, with a friend of mine. Um, and it’s, um, with acquisition, uh, so we will help companies with their acquisition and their marketing, and he is doing the sales, and I’m doing the back end part.
[00:10:45] Egidijus: Mhm. Okay. So, um, uh, let’s move now to the let’s say core theme of the podcast, which is about different campaign management tools and the technical aspects of campaign management. Um, I know that over those 15 years, uh, you have tried out countless campaign managements, uh, probably tried out all the biggest ones. Uh, could you, uh, uh, name, uh, kind of a few of campaign management systems that you have worked with and, uh, basically, what are their differences? What did you feel they are doing? Good, bad, etc.. So that kind of, uh, because not everybody has such a diverse experience, you know?
[00:11:37] Kwame: Yeah. Um, so I did, uh, inbound, uh, communication. Um, outbound communication. Um, uh, for the inbound, I worked with Cordiant, uh, in the, uh, back in the days, uh, outbound, I worked with, uh, unica. Um, I also worked with Selligent, uh, outbound. Uh, and I did omnichannel with Pega. Pega marketing. Um, um, I worked within Pega, Um, uh, I worked on the system, the system engineering side. So it was, um, uh, system. What’s it called again? Um. Uh, senior. Yeah, it was a system architect work. Uh, I did, um, worked on flow charts, and, uh, I also worked with, um, uh, Watson, IBM Watson, uh, which is the email, uh, sending uh tool, uh, which you could connect to, to unica. Um, so basically those are the big, the big, uh, the big, um, uh, things I work with.
[00:12:45] Egidijus: So, um, uh, from, uh, from those different tools. Yes. Uh, what were your, let’s say, key learnings, you know, because every, uh, kind of, uh, uh, when teams purchase tools, they expect that, you know, the tools will solve everything and they will not have to work, you know?
[00:13:09] Kwame: No, no, no, no, no. You’re absolutely, absolutely right.
[00:13:12] Kwame: So basically my my key learnings are um, and also for God. One, I also worked with Mautic, um, which is also called Acquia Campaign Studio. I worked with, uh, that at a, um, easypark. So basically what I learned is that all tools basically do the same thing, right? So they’re, uh, connected to a database and, uh, they’re just SQL schedulers. So rather than. But basically that’s it. So rather than that you write the sequel, you have a tool which generates the sequel for you. It creates a list, and then that list gets sent to somewhere and some things happen to that list. But basically you’re just generating a list and that’s that’s what you do. And you can have a big tool which does that for you. Let’s say Pega. Uh, Pega is for the enterprise, uh, companies. Um, but that’s a really a big bulky tool. Or you could have open source like Mautic, um, which also scales to, uh, 1 or 2 million customers easily. But basically, you’re just generating a list of customers you want to send something that is, I think, the key learning of, uh, of what? My last 15 years.
[00:14:33] Egidijus: Back to the roots. You know, back to the, uh. Basically the database marketing.
[00:14:38] Kwame: Yeah. Basically the same things I did at Triodos Bank, because the three of those I was making lists in SPSS and then at Vodafone, I was making lists in unica. Uh, and then afterwards I was making lists in Pega and basically just making lists. Of course it becomes, um, more complex, more complicated. So, uh, so if you compare, let’s say, a database of, uh, Holland Casino, which is in the gambling industry, if you compare that to, um, Odo or T-Mobile, um, as it was formerly called, that is way more complex, uh, because then you have massive amounts of data which you need to, uh, join, and you need to make sure that that you don’t make mistakes.
[00:15:25] Egidijus: Uh, but I guess it gets a bit more nuanced. Yeah. So if you need to add more data, if you need to make some things more in real time, etc., then the tool starts, uh, mattering probably. Uh uh, but uh, kind of.
[00:15:41] Kwame: Yeah, it’s, it isn’t that black and white? Of course. Uh, I mean, if you take a unica, then you have way more options than you have with, let’s say, uh, intelligent, uh, uh, or different tool. Uh, so, uh, let’s say, for example, you want to calculate the price offer in real time, or let’s say not real time in runtime, right. Um, then you want to have the ability, for example, to create, um, custom variables. Um, but if you want to have a custom variable which only lives in the flowchart at um, uh, runtime, then you need to have a tool which is capable of, of doing that. Uh, so unique is capable of doing that. Peg is capable of doing that. Um, uh, intelligent isn’t, uh, capable of of well, it gets a bit more nuanced. Intelligent. You can have variables, but but, but you can only access them within the flow of the, um, within the flow. But so so, so, so if you want to touch your variable at a different part of the flow chart, that’s not possible. So then it becomes really, um, it becomes vital in how you organize your data. So let’s say um, the, the the less capabilities you have in your tool, the more you need to do your data prep. Um, and with your data prep, you can solve lots of issues.
[00:17:09] Egidijus: So, uh, now, uh, we will go deep into the woods, you know. If you don’t mind. So as a person who run those marketing campaigns with different tools in different companies, uh, what would you say were the top three challenges that you were dealing on a daily basis? Basically, from a technical perspective?
[00:17:38] Kwame: Yeah. So from a technical perspective, uh, data, uh, quality. Um, so, uh, how do you make sure that the, uh, that the data that. Yeah, that that the data that you have is is correct. Um, something else is, um, the run, the, uh, the making sure that everything runs, uh, and that you also, uh, so, so if you have, like, let’s say five campaigns, this is then you can just see, okay, they’ve all run, but if you have tens or hundreds of campaigns which are running, then you make you need to have a system which informs you that, uh, the campaigns have actually run, um, and uh, sometimes this is in the tooling, sometimes the tooling doesn’t have, uh, those kind of things. And you need to create something, uh, which, which does the monitoring for you.
[00:18:32] Egidijus: You said that data quality is important. Yes. So could you tell us why data quality is so important? And what happens when the data quality is bad? And what are your top tricks to make sure that the quality is good, you know.
[00:18:49] Kwame: Yeah, yeah. So one of the problems can be that, um, uh, for example, if you, uh, if you need to send an, uh, a mail or, I mean, if you need to send a postcard, uh, uh, then you need to have all of the information to be able to, to send it towards, uh, a customer. Um, so, uh, or if you want to make an offer, um, then you need to know what. Current, uh, subscription someone is on, uh, that, that and that those fields are filled. So basically the the the the neatest trick, the best trick that I’ve come up with is actually, uh, writing a program which checks the output of the, um, uh, of the, of the information that you generate of the file that you generate. So you can, you can tackle, um, data quality issues earlier. Uh, let’s say, uh, when you are defining your data in your database and you’re defining a data view, then you say, okay, only, uh, records, which, um, uh, which, uh, which have all information in them are allowed to be in my campaign and you filter them out with the data view. But then if there is a data quality issue, you will not see that because the data is already filtered out. So if you check it after the, uh, campaign is run, then basically you have an end to end test and then you check if the data is there. Then you can make sure that all fields are that all information which is needed is present. And then you can send out your campaign.
[00:20:31] Egidijus: So when we say for example, data quality is good, you know, what are the tests that are mandatory to do on on the data? Because it’s like the data is invisible thing. You know, there are millions of records somewhere in the database. And what does it mean, a good quality of data, you know. So how do you check it?
[00:20:54] Kwame: That’s a good that’s a really good point actually. So, um, uh, so ideally it starts at, um, when a customer becomes a new customer. Right then, then then for example, that the first name is written with a capital, and then the last name is written with a capital, and that all of the, uh, basically that, that, that, that makes that, that works. But we all know that in the real world your data isn’t like that. So you need to restructure your. The names. You need to restructure the phone numbers for example that that they all have the in the Netherlands we have ten positions. Um but in other countries you have other positions. So those kind of things you need to make sure um, that is either handled by uh, by, by for example. Um, but then you also need to do that in accordance with by because basically the, the IT department or by department, they look at data differently. They’re like okay, so I have my post codes here. You need to have your post codes. There they are there. So you’re fine right. I’m not fine. They all need to start, be ordered like this. And then the process starts at getting your data correct. Mhm. Um, and those, those those things. Yeah.
[00:22:19] Egidijus: Uh, so if I understand correctly basically what are uh, from the technical campaign management perspective, what you are doing is you are fixing technical issues from all systems that generate customer data. Uh, and you need to make sure that the data is correct. So for example, if you if you have a post, uh, this uh, postal number. Yeah. Somehow you need to make sure that this is a valid postal code.
[00:22:58] Kwame: Yes.
[00:23:01] Egidijus: So these are so this is like one challenge. So making sure you know that the date of birth is in the right format, that the names are in the proper format, making sure that the name is not, you know, a curse word and making sure that the name is not an X, X, X, etc. those, uh, those kind of things.
[00:23:22] Egidijus: Uh, and if I understand correctly, there is another set of challenges. So, for example, yesterday you had uh, uh, 20,000 of people who used iPhone, and today you have 300,000. Yeah. Uh, so, uh, kind of the list size might be very different due to, uh, data challenges. Yes. How do you check on that?
[00:23:51] Kwame: Yeah. So what you can. So, so so basically what we what we had at, um, uh, at Vodafone at one point, um, was we had a B tables, so um, we had uh, yeah. So basically you have so you say, okay. So the data we have now is correct. Um, this will be the A and then the um, then new data comes in the next day and you compare the data on, um, on several, uh, basically on quantity. And then if the delta is too big, then you don’t load the beat, uh, table. You, you stick with the A because then apparently something is wrong. If it passes, then you, uh, load the B, uh, table, and then, uh, the next day you switch. Of course.
[00:24:41] Egidijus: Okay, so, uh, marketing has a lot of data tests. Maybe more than you could imagine. Yeah. So, uh, so, uh, after we make sure that all the data is correct, and when we use that data for personalization, all the information in those personalized emails, messages, uh, etc. are always are all correct. You also said that we need to make sure that everything runs. Yeah. So what would be your basically, uh, tests, uh, or the set of tests that you would say are necessary to feel confident that, you know, that things are running.
[00:25:34] Kwame: Yeah. So, um, uh, ideally you need to, uh, know what you will send out that day. So if you, for example, uh, you can go into system tables of your, um, of the tool that you are using and then you can see, okay, which campaigns are scheduled to run for today. Um, and then um, if uh, and then uh, then, then you check which campaigns are scheduled to run for today. And then at the end of the day, uh, you see, okay, have these campaigns actually run. You compare those two lists, uh, with each other, and then if they match 100%, then you’re fine. If they don’t match, then you need to go in and fix it.
[00:26:19] Egidijus: Okay. So this is kind of a perspective number one that actually the campaigns ran. Yes. Uh then what about kind of, I don’t know, notifications of failures. Uh, what if campaign ran too long. So, for example, you expected that it would be running like for an hour, but it’s running for ten hours. It’s, you know, uh, are there any more nuances, you know?
[00:26:45] Kwame: Yes, yes, yes, it can be. It can be basically it can become it. So this is where the, uh, the difference in tooling actually starts to, to matter. Um, because the more advanced tooling you use, the more, uh, freedom of the more freedom you have within your tool. Right? So basically, if you have a tool like Selligent, um, you, the only notification that you get is, um, that it started and if it failed. But if you get a tool like, like Pega, for example, you can customize that, uh, you can fully customize it and, and then you can also start monitoring it. Indeed what you said on runtime, on outputs, um, on uh, input, those, those kind of things. Um, and then um, it really depends on what kind of operation you’re running. And if you’re running something as complex as this, uh, telecom and the telecom. Uh, then you need to make sure that there are a lot of moving parts, right? Um, so, uh, one of one of the things that you, uh, for example, uh, if you have an omnichannel, uh, campaign, um, set up running, then you have an API call which is coming in from a website, um, uh, for, uh, if a customer visits the website website, you need to make sure that, uh, um, that, that the correct offer is served so that API call can fail for whatever reason. Um, if it fails, you need to fix it. Um, then, uh, you can also get a, um, another call coming in from, uh, let’s say the, um, um, how do you say the, um, the phone, um, um, uh, um, the, uh. Perfect. Yeah. The call center. So. So basically, the call center wants to make an offer to a customer that’s also, uh, coming in. So there are lots of moving parts, and you need to make sure that everything works. Um. Uh, so, but that is, that’s from a technical perspective. But the organizational perspective is also important because who is going to pick up.
[00:28:57] Kwame: All these. Who who’s going to fix it. Right. If you have a team of, let’s say, uh, ten people working and uh, and you have all of everyone is developing campaigns and they make sure that everything, uh, goes out and they need to test it and they need to, uh, have and they need to discuss this with the marketeers. Um, if something fails, who will pick it up? So you will. So you would need to have someone who is on an, um, ops duty to make sure that if, if something fails, that, that the operation keeps on working.
[00:29:33] Exacaster: If you are interested in customer value management, check out our Customer Value Management body of knowledge. Cv and Bog is a comprehensive guide for CVM professionals offering tips, tools and best practices to help you in your job. Visit CVM Book.com for more.
[00:29:50] Egidijus: Do I understand correctly that, uh, you. Once you are running those robust operations, you need to think about, uh, basically CVM ops as a separate, uh, um, kind of operations that sometimes even need to be on duty, you know, over weekends and, uh, and holidays. Yeah.
[00:30:11] Kwame: Yes, yes. Um, so, um, at, uh, in the telco, at Vodafone Ziggo, we had, uh, CVM ops, actually, we also called it CVM ops, and it was a dedicated role, uh, which was, um, rotating between the team members. And uh, the person on ops, uh, actually had less campaigns to develop that week because 50% of the time was assigned to UPS. Um, okay. Yeah. And in the weekends, we, uh, we we didn’t, uh, send out campaigns, but we did have someone on standby to make sure CVM marketing automation run smoothly.
[00:30:54] Kwame: Just to make sure, just in case, because you’re running a 24 hour operation.
[00:30:59] Egidijus: Yeah. Because there are still some real time activities going on. Customer is making top ups or purchasing something you need to and visiting your website.
[00:31:10] Kwame: Yes. Yeah. And also, um, for example, twice a year you have a server upgrade. The server upgrade only happens at uh, between uh, three and four at night on a Saturday and Sunday. So yeah. Who is going to check if the server update works. Those kind of things. You know.
[00:31:30] Egidijus: I bet that nobody told you about server updates when you were, uh, in university learning about marketing.
[00:31:43] Kwame: I didn’t even know what a server was. It’s like. I thought, oh well, everything works by itself. It’s okay, man, it’s automatic automated. Uh, no one knows.
[00:31:56] Egidijus: But then suddenly, uh, some network rules were changed in the network, and your campaigns are not going through and and everybody are shouting on you, you know.
[00:32:08] Kwame: Yeah, but it’s not like they close. The gate. They close the port, those kind of things. Yes, I know, but you.
[00:32:15] Egidijus: Have the problem. You know, the problem sits on your desk. People are shouting at you, you know.
[00:32:19] Kwame: Yeah, yeah, yeah. But there’s one valuable thing which I’ve also learned in the last 15 years, and that is if, um, um, that. Well, two things actually. In the telecom, uh, world, um, everyone always says this as prior and I thought, yeah, but if everyone says it has Priya, then nothing has Priya. So because you cannot have ten priors, so you need to keep your head cool. And the other thing is that, um, if somebody, someone says it is important, I’ve only actually, uh, had one time that something was really important, that I was literally working at an other department. Um, this was this was when I was working at the bank. I was working at different departments. I needed some, some, some me time to, to concentrate. And at one point I launched the campaign and the co-worker was actually standing beside my desk. Uh, they said yes. Uh, uh, can you help me? Okay. What is it? Yes. Uh, you need to stop running the campaign. I was like, what? No. Yeah. You need to stop running it now. Okay? And then it was a it was a. There were two, actually. So one, one of the marketeers, she was crying and the other one was saying you need to stop it now. And I was okay. Yes, I know. And I was like, oh my God, what did I, what did I do? But it was actually, um, the, the, the, the thing, what was wrong was that there was a wrong variable delivered which was loaded into the email, which caused, uh, customers of the bank to be really concerned about. Oh, my God, I don’t have any money anymore in my bank account. Started phoning up the bank, those kind of things. So if something is really important, people will find you and they will come to your desk and they will make sure that you do it. Now, that’s. Otherwise, it’s not important. Well, it is important, but it’s not urgent. It’s not urgent.
[00:34:20] Egidijus: Okay. This is a very good lesson, you know. Uh, so, uh, getting back to the CVM ops part, uh, uh, how would you describe the CVM ops? Uh, function? Like, in a couple of sentences. What is the essence? Because you really defined the essence of campaign management very well. Okay. Can you try to uh, uh, define the CVM ops as well?
[00:34:51] Kwame: Yeah, I think the the CVM ops is the most underrated, but also one of the most important parts of the CVM team. Um, and, um, so, so basically the job of, of CVM ops is to make sure that stuff gets sent out and that stuff keeps on working. Um, and, um, there are lots of moving parts. Um, I’ve worked in many industries, uh, but, uh, banking, uh, is complex. Telcos also complex. Uh, but the difference is that telco goes way faster than banking. Um, and because of the um, uh, the speed which is needed to get your campaigns running, um, um, you have the risk of things falling over and if something falls over, someone needs to pick it up. And that is the part of ops. And without CVM ops, your operation, uh, or your campaign operation will not function. It will not function as well as you think it does.
[00:35:57] Egidijus: Thank you a lot. I think, uh, I didn’t find a better definition so far.
[00:36:09] Egidijus: So, Kwame, uh, if if you would have to choose a campaign management system right now. Yes. And you have seen it all, done it all, you know, uh, what would be the most important aspects that you would take into account when making a decision to purchase a campaign management system? Because making a change in organization probably is a huge thing. Everything will change, you know, with a new campaign management system. So how do you make it? Uh, how do you make sure that you chose the right thing?
[00:36:49] Kwame: Uh, I think, uh, well, first of all, you’re right. Everything changes. Uh, I’ve, uh, seen the I’ve worked on the implementation of Pega, uh, at, uh. Vodafone, Ziggo. Um, I’ve also done, um, a Mautic implementation at Florence’s, which was a company, a customer of mine. Um, and if you implement a campaign management tool, everything changes, um, throughout the entire organization. And I would say that the most important aspect to keep in account is that you need to know what do you want to do? Because yeah.
[00:37:35] Kwame: Mean yeah the definition. So so let’s say let’s say you want to um, you want to uh, to offer omnichannel, real time, uh, offers. Yeah. That is I wouldn’t personally then I wouldn’t go with Mautic as an open source. I mean, it’s not that open source is bad, but it’s just not made up for that. It’s not made for that. But if your goal is to send batch campaigns in a timely fashion, matter, matter, then Mautic is awesome. Um. Mhm. Uh, but then Pega is not your place to go because that would mean that it’s just complete overkill. I think the the yeah, yeah, you definitely don’t want to go with that. So I would say I would say a good comparison is, um, uh, enterprise level tools are like formulas, formula, formula one cars, F1 cars, they go really fast and they’re really fine tuned.
[00:38:35] Kwame: Awesome machines. But if you want to go to the city center of, um, Amsterdam or, uh, Berlin, you cannot get there with the formula one car. You need to have a Fiat Panda or a Opel Kadett or whatever. It will actually get you there. Or or let’s say, uh, if you, uh, if you want to, if you need to get. A class of school children, you need to get them to the museum. You don’t show up with your car. You need the bus. Because then in a bus you can actually. Go and get them there. So it really depends on your purpose. What do you want to do? And as soon as you have defined that, then you can make your tool selection. But you need to do it in that order.
[00:39:24] Egidijus: Okay. Um that’s a really, uh, uh, valuable piece of information, you know, because, uh, I haven’t seen many companies that start from the purpose. Uh, many companies start from a beautiful idea. You know, I will do everything in real time. Uh, I will do everything with machine learning, everything AI, etc. and then kind of you get lost down the way. Yes. Um, now the other important aspect is let’s touch a bit about the change. You have seen several campaign management systems being changed. I will ask you two questions. How much time does it take to purchase a new campaign management system? And how much time does it take to actually implement a new campaign management system from start till the legacy system is shut down?
[00:40:28] Kwame: Wow. I think the sales side. I think the sales cycle. So to purchase it can can take easily a year. So let’s say one one year probably because you need to go through all of the options that well, first you need to define the goal, the purpose, what you want to do. Then you, um, go through the options and then you, uh, get the contracts, those kind of things. So it’s still cycle is easily one year, perhaps one and a half years. Then the implementation part can take I think two to between 2 and 4 years. Something, something like that.
[00:41:09] Kwame: It tool a lot of time. And sometimes, sometimes you cannot even face out the. But but this is this is okay. So now I’m talking about the complex organization. So you have uh, we had 100 hundreds of campaigns running and you need to migrate everything. Um, and then you also have a, um, a period that, uh, tooling, uh, is besides each other. So you have your legacy system and you have your new system, and when do you actually switch it off? And then you also need to compare the outcomes of the new system with the old system, because you need to make sure that everything is migrated correctly. Um, so um, so let’s say the implementation, it uh, it took like, yeah, uh. It started when I was there and it was still not finished when I, when I left.
[00:41:54] Kwame: And switching off, switching off the legacy tool. Um, sometimes it’s not possible to switch off legacy tools because I know companies who have purchased a new tool and they still need to use the legacy tool because the new tool isn’t capable of, uh, doing stuff the legacy tool could do.
[00:42:19] Egidijus: So this is how you end up by having, you know, a zoo park. Uh, it’s like, uh.
[00:42:24] Kwame: Yes, yes. You have a zoo of multiple legacy stuff and those kind of things. But, uh, but I would also say that, um, that that it depends on the complexity, but also on the political part of the company. Because you have a company, politics and people want to do this, people want to do that. And that can be that can be very time consuming.
[00:42:48] Egidijus: Mhm.
[00:42:49] Egidijus: Okay. Um, so now if we are switching from one tool to another, we are setting up ourselves for 2 to 4 years. Project. Uh, what, uh, what would be your advice or what would be your learnings from previous projects? Uh, what is important to make this project a success or what are the challenges basically during the migration part?
[00:43:20] Kwame: Yeah. Well, I would say that one of the things is that, um, um, uh, people, people can underestimate how long it takes. So it can be that the manager. The manager who, uh, says, okay, fine, we need to migrate, uh, isn’t there for the entire migration part. Um, so then the next manager who comes in isn’t responsible for the for the for for for the migration or doesn’t feel responsible. So that is a risk. Um, then also the, um. Uh, the people who, who, who you have working, you need to, um, take them with you with the change, and you need to make sure that your own staff are actually learning the new tool. Um, so even though I’m now working as an independent consultant, I’m still saying you need to have people on the payroll. You cannot outsource your primary, uh, process. Uh, you can have. Of course, you can have experts who help you for a short period, but you don’t want to be dependent on them. So it is really important that you take your own people and you learn them working with the new tool from the first possible moment. So that you teach them and you also, you also need to, um, make sure that they don’t get the feeling that they will be made redundant because perhaps they don’t want to learn it, or perhaps they think that they cannot learn it. Um, and I would say the last thing is that, um, uh, a new tool can be daunting, right? And people can say, okay, this is a new shiny, fleshy tool. And then the consultants who are coming in, they are protective of their knowledge, so they don’t want to share it, and they make it a really big thing. But actually if you make it really small, again, it’s just a relational database which produces lists, which just with a with the shiny user interface. That’s it. If you can make sure that your people also understand that it’s not more than that, then you can take them with you on your new journey and you can make the transformation.
[00:45:38] Egidijus: Okay. And, uh, could you share, uh, let’s say, what is the feeling from being inside, uh, and going through that transformation? Because at some point, you know, you are running your old campaign management system. All processes are running. You need to be kind of on, on the stand by, stand by. You know, you are operating those campaigns and you are moving to new campaign management system. You are now operating two tools, creating campaigns in two different environments. Need to, uh, you know, you need operations and both of those environments, you know, how does it feel from the inside?
[00:46:20] Kwame: It can be quite chaotic. Uh, and, uh, if the, um, uh, if the if the lines are not clear, then you can get, um, you can also get confusion. Right? So, um, uh, so when we did the migration of unica to, uh, to Pega, um, not all team members were, um, onboarded at the same time, which is also not possible, of course. But it can create a feeling of anxiety of why am I not allowed to touch the new tool? Um, but also what we also experienced was that, um, some, some people who were on the payroll weren’t allowed to do stuff inside of the new tool. Uh, okay. But only the consultants, they, the external consultants were, um, that allowed to do that. And then you get. Yeah, but. Hello. I’ve been working here for, like, 4 or 5 years.
[00:47:18] Kwame: I actually run the current, the current campaigns. Right. If I’m not here, then things break down. So, yeah, I know what I’m doing. So please let me touch it. Those kind of things. Right. So that’s also something that you need to make sure. And uh, and it can just get a feeling of anxiety and, uh, um, and you need to make sure that you don’t, that you don’t lose your people along the way. Um, I think it’s more, much more, I would say it’s much more a people management thing. Than a technical management thing, because getting a system up and running is not that difficult, right? I mean, that takes like 4 or 5 weeks and then all the connections are made and you have your data and whatever, but then you need to start migrating. People need to start working new processes, new this, new that, new those kind of things that that is what takes the time.
[00:48:31] Egidijus: So uh, for example, if you, if you would start the project what would be your first thing. So imagine you purchased a new campaign management system. You need to roll it out. What would be your number one task? Basically.
[00:48:49] Kwame: Um, my number one task would be to, uh, to onboard the, um, uh, um, the. Yeah. So, so, so basically so basically you you want to produce stuff with your, um, new tool, right? So you want to onboard the people who, um, who will work with the products. You would. So let’s say, um, uh, in this case, marketeers. And then we’d say, yeah, marketeers do not have anything to do with it. No, but if, if stuff changes, for example, if, uh, if you go from, if you go to an NBA and express action, um, then you say okay, but then the system decides what will be offered. So then all of your KPIs will become worthless because yeah, you don’t have any, um, any say in it anymore. So that means that the organization changes. So then if you if you first implement the tool and then you onboard the people whose work will be impacted, then you have struggle on struggle. But you need to start with onboarding them. So I would say the number one thing is basically to get the people on board, because it’s the people who need to work with the, the tool or also who are impacted by the outcome of the tool that those are the people you need to onboard. And that’s that’s really secondary to all of the technical stuff, because the technical stuff is because software, software nowadays also is really good. It’s just the click, clickety clack and then an API call and then data starts flowing. It’s like, yeah, it’s not the same as 30 years ago, right.
[00:50:54] Egidijus: Okay. So, um, it’s it’s been a very interesting conversation about, uh, campaign management part. Uh, let’s slowly move to the last part of our conversation. Uh, it’s, uh, uh, the three questions that we always ask, uh, our participants. So the question number one, uh, is what was your proudest moment in your career? You know.
[00:51:26] Kwame: Proudest moment in my career was that, um, when I was working, um, many, many years ago. Um, I, um, so, so, so the campaign tool that we had, um, the the file was, uh, sent to a channel partner, and then it was checked. And then if the if the information was missing, we got it back. And, uh, at one point, uh, I got it back at the end of the day on a Friday, and I needed to fix it. And it was really upset because I missed dinner at home. And I don’t like missing dinner at home. So I was riding back home and I was like, ah, I missed dinner. And I was like, hey! But actually, what does the tool do? Okay, it checks for missing columns, blah, blah, blah. I think I can actually make that myself. So over the weekend I, um, I created, um, a proof of concept. And then I spent one and a half years perfecting that. And then basically we could, uh in-source everything. So I’m really, really proud of achieving achieving that at every every communication up on this. Until this day, every communication sent by Vodafone or Ziggo to any person in the Netherlands is checked by that piece of software which I wrote.
[00:52:44] Egidijus: That’s amazing. Uh, okay. That’s a very inspiring part. Um, and it’s really interesting that, you know, progress happens at, uh, the biggest point of frustration, you know.
[00:53:00] Kwame: Yes. Yeah. Yeah. You’re right. Yeah. Yeah. That, that’s that’s. Yeah. Because then you think, oh no, how can I fix this?
[00:53:09] Kwame: And then you start and you start thinking.
[00:53:13] Egidijus: Together with, uh, you know, the proud moments. They are failure moments, uh, in the career, uh, especially in customer value management because they are very visible, you know, could you share some of your, uh, kind of failures as well?
[00:53:30] Kwame: I have many. Yeah. I would say, uh, one of the biggest was, um, uh, when I was making a selection for when I was working at the bank and, and in the bank, your your data needs to be 100% because people get very, um, how do you say, uh, um, nervous if you get the wrong communication from your bank. Uh, so, um, I was making a selection for the, um, uh, for, uh, for Ford, for the for the coming out. How do you say it? For the magazine. For the magazine of the bank. And I had this, uh, list and, uh, it went to the printer, and then everything was fine, and I checked it, and my manager checked it. Everyone checked it. Everyone’s fine. La la la la la. Okay, I was was the first time. Actually, I did it myself. Uh, and I was working, like one year or something like that. And then at the last, I don’t make it up. The last address that my manager checked, he said, hey, Eric Young, this guy was called Eric. Hey, Eric Young, you’re living with, uh, this is your this is this this. And this is your address, right? And, um, you’re living with this and this and this and this person. And then Eric Young said, no. That’s my ex-wife. And I was like, no. I just still feel I can still get the feeling because it was the last time it was the last address was checked before over 200,000 records were sent to the printer, and everyone would get the distributed. So, in the bank you have, uh, in the current account, right? You have your own account, you have your shared account, and, uh, you can have multiple accounts. And then at one point you close the account and. Yeah. And I took the names and addresses of closed and opened accounts. I didn’t filter out the closed accounts. And then the match is made at random. So at this point at his closed account with his ex-wife, was matched at his current at his name and his address. And that went almost horribly wrong. I think that because I was also on the temporary contractor and I had kids and I had a wife and a house, and I was like. Oh my God, I’m going to lose it. No. Yeah.
[00:55:55] Egidijus: Oh my God. But you are lucky, you know that this was not printed, you know?
[00:56:00] Kwame: Yeah, yeah, I think I could have sold. My house, actually. No, no, I was, I was in the. Yeah. No, but yeah, I was really, I was really lucky and and yeah it got, it got really, really like okay I need to make okay. So this is real data, real people. I need to make really sure that what I’m doing is, is just 100% correct. But I was I was really lucky. Until this day when I’m, you know, this, when I’m telling this story.
[00:56:27] Kwame: I get the feeling and I’m just like, oh, God.
[00:56:33] Egidijus: There is one very important, uh, piece about customer value management. You know, you need to be 100% right, because, uh, uh, it’s like your mistakes are very, very widely spread and, and, uh, the last part is, um, what learning resource would you recommend for, uh, any fellows, uh, who are either working in this field or would like to, uh, step up and start their career here?
[00:57:08] Kwame: There isn’t a, uh, study of it. Right. So you you, um, so, so you cannot go to university and learn it. Um, so the way I learned it is through, uh, talking to people, talking to, uh, uh, experienced coworkers and, um, basically, because it is an ever evolving field, um, I would say the way, the way to learn it is to think that, um, uh, I’ve, I’ve said several times, uh, it is a, uh, your, your generating SQL on a database, but that’s only one part. The other part is that you’re working with a computer so you can calculate stuff. You can build stuff. So the only thing which keeps you back is your imagination. If you can think of it, you can build it. And that is I think, the key way of learning it, because. Then things become possible, like, um, uh, if you if you, if you want to, uh, send out an email for a sauna, for example, and then you can also say, okay, fine, I want to, uh, include the weather because then people might want to know if it’s shiny, sun is shining or whatever. And then you can think, oh, wait, but then I can make an API call to this. That’s such. So, um, but but basically you need to you need to keep in mind it is a computer and you can program stuff with it. And then then you can learn it.
[00:58:52] Egidijus: And then the magic happens. Yeah.
[00:58:54] Kwame: Then the magic happens. Yes.
[00:58:56] Egidijus: So thank you, Kwame, for a really inspiring, fun and deep, uh, conversation today about CVM marketing automation.
[00:59:05] Kwame: It was fun. Uh. It was fun. I don’t know the inspiring part. I leave it to others. But it was. It was fun.
[00:59:14] Egidijus: Yeah, I imagine that there are plenty of people, you know, who are going through the same struggles.
[00:59:20] Kwame: So yeah, I can also imagine it is. This is quite a complex field, but it’s really a fun field. And I think that we should always keep in mind that if you can’t laugh about what you’re doing, then you should.
[00:59:36] Kwame: You should perhaps find another field, but you should just be laughing and, uh enjoying yourself and, uh, not sending magazines to ex-wives. That.
[00:59:55] Egidijus: This is what I also love about this job, you know, because it’s like, if you can’t laugh, it’s it’s not a good job.
[01:00:02] Kwame: No. Yeah. You’re right. Yeah. You definitely. Do. Definitely go and do something else. Yeah.
[01:00:07] Exacaster: CVM stories is produced by Exacaster. We help companies take their customer value management to the next level. To stay updated on our latest episodes, subscribe to the podcast or sign up for an email newsletter@exacaster.com/cvmstories.
#CustomerValueManagement #TelecomInnovations #CVMStrategy #marketingautomation #DigitalMarketing #Personalization #MultiChannelIntegration
