Latest / Shiny New Object - a Marketing Podcast / Episode 308 / Francesco Federico / S&P Global / Chief Marketing Officer
Transcript
- 0:00Francesco Federico: You need to take accountability for your own
- 0:02instincts as well, and you know, and find the courage to make
- 0:05decisions following, you know, your gut feeling and because at
- 0:08the end of the day, marketing is still very much about emotional
- 0:11connection, ingenuity and creativity.
- 0:20Speaker: Are you a fitness loving foodie? Then defy snacks
- 0:23is for you. Defy offers real high quality chocolate you love,
- 0:26combined with functional ingredients, giving you as much
- 0:28as 26 grammes of protein per bag. So you can be both a little
- 0:32bit bad and a little bit good at the same time as a female owned
- 0:35brand, defy dedicates a portion of its profits to other women
- 0:39owned businesses currently sold in the US. You can find us on
- 0:42defy with an i snacks.com and follow us on social at defy
- 0:46snacks.
- 0:53Tom Ollerton: Hello and welcome to the shiny new object podcast.
- 0:56My name is Tom Ollerton. I'm the founder of automated creative,
- 0:59the creative effectiveness ad tech platform, and this is a
- 1:02regular podcast about the future of data driven marketing. Every
- 1:05week or so I have the pleasure and privilege of interviewing
- 1:08one of our industry's leaders, and this week is absolutely no
- 1:10different. I'm on a call with Francesco Federico, who is Chief
- 1:15Marketing Officer at S&P Global. So Francesco, for anyone who
- 1:19doesn't know who you are and what you do. Can you give us a
- 1:21bit of background?
- 1:22Francesco Federico: Yes, of course. And thanks Tom for
- 1:24having me. So as you said, I'm Francesco Federico. I'm
- 1:27originally from Milan, Italy, and now I live in London. As you
- 1:32said, I'm in the CMO for S&P Global, which is this sort of
- 1:37huge heritage brand, 165 years old of history, and oversee
- 1:41brand, the demand, generation, thought leadership across all
- 1:45five divisions of S&P, and in terms of where I come from, I
- 1:50mean maybe 20 seconds pitch on sort of my journey so far, I
- 1:54studied law at university, believe it or not, but then I
- 1:58ended up in marketing when I founded my first sort of
- 2:04venture, small digital marketing sort of consulting firm in
- 2:09Milan, had fun with it for a few years, and then I joined
- 2:13Vodafone, and since then, I've been with large organisations
- 2:16like Acer in consumer electronics, JLL in commercial
- 2:21estate, and now S&P Global. And I think the sort of the
- 2:25connecting point, a fil rouge, if you want, in my career, has
- 2:29always been change and disruption. I always join these
- 2:31brands at the moment of an inflection point in their
- 2:34journey. And I hope I play the role in sort of pivoting either
- 2:39the business model or the marketing organisation to sort
- 2:44of really be fit for, for that new wave of change.
- 2:47Tom Ollerton: So what has been the best investment of your own
- 2:51time, energy or money in your career,
- 2:54Francesco Federico: I think, at the maybe at the expense of
- 2:56being a little bit banal, because I know Steve Jobs has
- 2:59been quoted saying this multiple times, when he famously
- 3:03referenced to his sort of decision to take calligraphy
- 3:07classes, which later on in life, as he said, like he connected
- 3:12the dots, and when he was working through the sort of user
- 3:15interface of the first Macintosh, all that sort of, all
- 3:20those studies came back to fruition and allowed him to
- 3:23create beautiful interfaces. And I think for me, has always been
- 3:28learning, I mean the biggest investment in time and energy
- 3:31and sometimes money, because I often self funded training that
- 3:35has been by far the best investment, and it's also
- 3:39something that I recommend, sort of young marketers climbing up
- 3:43the ranks, or to be honest, any sort of young professional, is
- 3:46to really stay curious and investigate adjacencies. So not
- 3:52just obviously, study what is your field, your discipline, but
- 3:55be curious about adjacencies. And even potentially, just like
- 3:58in the case of Steve Jobs, even things that are completely out
- 4:02the remit of your day job, those are the areas that's going to
- 4:07probably bring you, sort of the best inspiration in the future.
- 4:12And back to sort of Steve Jobs point of connecting the dots,
- 4:15you can really connect the dots together. So for example, in my
- 4:19own sort of very recent example, in 2022 I decided to learn
- 4:24coding because I never had the chance to at the time or desire
- 4:27to really learn. But that was something that actually I said,
- 4:30Okay, why don't I learn Python? It seems sort of easy enough for
- 4:33a first time coder like me to learn. And then after a few
- 4:37months, chat GPT came came out. And so throughout the 2023 I was
- 4:41actually able to play directly with many of these open source
- 4:47generative AI models, and that gave me a profound understanding
- 4:51of the technology. Of course, I'm not I cannot consider myself
- 4:54a coder, but that knowledge of Python really allowed me to
- 4:58explore these technologies. Technologies and then ask the
- 5:02right questions, or maybe probe a little bit more, sort of
- 5:06deeper that I wouldn't otherwise have been able to when I was
- 5:10having conversation with my technology partners in the
- 5:12organisation. So there have been so many other similar examples
- 5:16in my career. So stay curious and really look at adjacent
- 5:19areas to your business that might come useful one day.
- 5:23Tom Ollerton: But how would you fit that in? Right? Because I
- 5:26always say the opposite of busy in marketing is to be
- 5:29unemployed, right? There's always something else to do.
- 5:31There's always something else to learn, and especially with AI
- 5:34and automation, we can do a lot more in less time. But you know
- 5:37the nature we're going to be asked to do more in with the
- 5:40same amount of time, right? As opposed to finishing early and
- 5:42doing less. So how do you how do you find time to do this extra
- 5:45stuff? You got a couple of kids, got big job. How do you make
- 5:50sure that you've got time to study the equivalent of
- 5:53calligraphy or something more adjacent, like coding?
- 5:55Francesco Federico: Well, whenever possible, I try to
- 5:58solve problems that I have. So I not just study for the sake of
- 6:03studying, but studying something that can have a practical
- 6:07application, even my personal life. So back to the example of
- 6:10coding. One way I sort of practised was in sort of have a
- 6:14little bit of a nerd myself. So I have a sort of smart, smart
- 6:17home with sensors and, you know, all the lights, etc. And I
- 6:21decided to apply those skills I was learning in Python to
- 6:25further customise my home setup. And so it was a home project for
- 6:30the whole family. And so to this extent, it wasn't necessarily
- 6:35eating up family time, because I found also ways, maybe, to bring
- 6:39in some of my family members along the journey and and also
- 6:43had an outcome in mind that also gave me motivation to really
- 6:46continue learning. So I think, obviously, you can't always do
- 6:51that. Obviously, you know, I spent some days during the
- 6:54Easter break working on retrieval of mental generate
- 6:58generation. At one point, my wife told me, You need to stop
- 7:01and just, you know, close the computer. So there are those
- 7:04moments. But the advice I would give is try to do something,
- 7:09learn something, that has a practical outlet, practical
- 7:12application in your day to day life, to make it a little bit
- 7:16more compatible with all the other duties.
- 7:19Tom Ollerton: Yeah, that's a really nice idea, and that's
- 7:22something I struggle with quite a lot, actually, because I make
- 7:27music, but I make it in exactly the same place where I work in,
- 7:27like, in front of my laptops, you know, close the laptop,
- 7:27whatever time, and then right it's time for a hobby, and you
- 7:30just open up the same laptop when you're sat in the same
- 7:34room. And it's quite, quite tricky that I find, but I love
- 7:37that idea of solve a problem, and then you're, yeah, you're
- 7:39getting, you're getting double the double the benefit. So
- 7:42getting back to digital marketing now, and away from
- 7:45this smart home setup, which I'll ask you more about later.
- 7:47But what is your best bit of advice to become a better data
- 7:51driven marketer?
- 7:52Francesco Federico: Well, I, if you allow me Tom I might be a
- 7:56little bit controversial here, and because I've been hearing
- 7:59about sort of being data driven and data driven decision making
- 8:03and data driven marketing, etc, for for quite a while, and
- 8:07although obviously, I do not dispute at all the importance of
- 8:10data I also think that sometimes this almost obsession with data
- 8:16driven marketing or data driven decisions can become a refuge
- 8:21for mediocrity, if I may say so, or also a way to escape
- 8:26accountability. And I think data tells you a lot, tells you
- 8:31yesterday's story. Sometimes, if we are all really honest, then
- 8:36we all know how our data stacks made of we don't know, we know
- 8:41what gaps we have in our analytics platforms. So I don't
- 8:45think anyone can really say we have 100% sort of correct data,
- 8:52accurate data, and so you have this sort of data almost God,
- 8:57that is looked at to sort of base on all our decisions on but
- 9:03often is incomplete. And as I said, it's always historical.
- 9:08And so for me that to become a better data driven marketer, I
- 9:13think you need to find that balance between reading the data
- 9:19and making sure that you have the best sort of data layer
- 9:24possible in terms of accuracy and precision, but at the same
- 9:28time, you need to take accountability for your own
- 9:31instincts as well, and you know, and find the courage to make
- 9:35decisions following, you know, your gut feeling and because at
- 9:38the end of the day, marketing is still very much about emotional
- 9:43connection, ingenuity and creativity. Of course, marketing
- 9:47has changed immensely in the past 10-15, years, and I've been
- 9:52saying this for many years, that the role of the CMO has steadily
- 9:57converged with the role of the CIO and vice versa. Also, CIOs
- 10:02are now more and more versed in marketing and customer
- 10:05experience, but that does not negate the need for sort of
- 10:11courage in making decisions. And so I always say you want to be a
- 10:15great data driven marketer, remember to balance the data the
- 10:20rational side of you, sort of your mind with the instinct, the
- 10:26gut feeling and the back to my previous answer, the ability to
- 10:29connect the dots, sometimes in ways that your analytical stack
- 10:32cannot cannot provide.
- 10:34Tom Ollerton: Yeah, reminds me of a previous guest on the show
- 10:36that said, if we only needed dashboards, then you wouldn't
- 10:39need marketers, right?
- 10:40Francesco Federico: Indeed.
- 10:46Tom Ollerton: This episode of the shiny new object podcast is
- 10:50brought to you in partnership with Madfest. Whether it's live
- 10:52in London or streamed online to the global marketing community,
- 10:56you can always expect a distinctive and daring blend of
- 10:59fast paced content, startup innovation pitches and
- 11:02unconventional entertainment from Mad fest events, you'll
- 11:05find me causing trouble on stage, recording live versions
- 11:08of this podcast and sharing a beer with the nicest and most
- 11:11influential people in marketing. Check it out
- 11:15www.madfestlondon.com.
- 11:23So we're going to now move on to your shiny new object, which is
- 11:28agentic AI in marketing. I think I know what those words mean,
- 11:31but could you be specific about why that's what is agentic AI in
- 11:35marketing? Why it's your shiny new object?
- 11:37Francesco Federico: Yeah, of course. Well, let me start maybe
- 11:40from a quote so Marc Benioff, the chief executive of
- 11:44Salesforce earlier this year, said something that was quite
- 11:48striking, as he said, This is going to be the last generation
- 11:52of chief executive officers that will only lead humans. So
- 11:56essentially, what he was saying is, very soon chief executive
- 12:00officers, and I bring this into my work as well. So marketing
- 12:05leaders will have to be able to lead a hybrid workforce which is
- 12:09made of, obviously humans, but also sort of machines, AI,
- 12:13colleagues. And so what we mean by agentic AI is the step beyond
- 12:17the current if you want level of AI, which we call generative AI.
- 12:22So generative AI is sort of this set of technologies that may
- 12:26enable us to create new content or new images or new videos,
- 12:30starting from a data set that has been analysed by various
- 12:35technologies. Agentic AI is that step forward that connects
- 12:39multiple and different generative AI technologies
- 12:42together with a very specific purpose, which is why it's
- 12:46called agentic. It's because it's an agent. So you might
- 12:50have, for example, an agent tasked with campaign planning,
- 12:54and so this particular agent would not just come up with the
- 12:58copy so generate new text like chat GPT would, but it would
- 13:02also create the whole the whole Gantt chart with all the
- 13:06milestones feeding all the creative copywriting, and
- 13:10potentially even create all the media assets that go with it.
- 13:14And so you can see how existing roles will, at some point in the
- 13:19near future, I believe, be augmented, in some cases,
- 13:22potentially even replaced by the agentic workforce. And so the
- 13:27reason why I think this is so exciting is because we're really
- 13:31witnessing the dawn of a completely new phase in
- 13:34business. Of course, I talk specifically about marketing,
- 13:38but this really applies across across all functions. I think
- 13:43probably customer operations, if you think about contact centres
- 13:47and marketing, are probably the two sectors that are going to be
- 13:52the first one to benefit from these technologies, but
- 13:55eventually you will see this even in healthcare. There are
- 13:59already some experiments of generative AI applied in
- 14:03healthcare settings, and you can clearly see hybrid workforces
- 14:07emerge there as well. And so I think the challenge now, as we
- 14:12only are literally seeing the dawn of these technologies. The
- 14:16challenge for leaders, I think, is to think about their own
- 14:20teams. And how do you train how do you make sure that you bring
- 14:24in not just the technology, but also you think about the change
- 14:29management and you bring your people along the journey so they
- 14:33can effectively accommodate for, if you want the these new
- 14:37colleagues, you know, these new machine colleagues. And the
- 14:41final bit, I would say about agentic is also another big
- 14:45shift happening in marketing, is that now, obviously, we've been
- 14:50talking a lot about business to business, business to consumer,
- 14:54and now in some cases we're talking about business to
- 14:58business to consumer, which is sort of a little bit more
- 15:02complicated models, but I think now there is an emergence of
- 15:06also what I call sort of machine to machine marketing, where you
- 15:10are actually writing marketing and you are bringing to life
- 15:15marketing campaigns that are targeting machines. They are
- 15:19targeting other agents. Because, again, some decision in terms of
- 15:23buying decisions or considerations consider, the
- 15:27consideration journey will actually be driven by agent
- 15:30agents in the client companies systems and so you need, still,
- 15:35obviously, to write web pages for people that will read it
- 15:39from buyers who will read it human buyers, by the same time
- 15:43also be ready to cater to an emerging machine audience of
- 15:47agents going around your properties trying to find out
- 15:51the best products for their particular outcome. So this is a
- 15:55completely new, this is probably the main revolution that I
- 15:59think, you know, you cannot probably think of any like,
- 16:03anything like this since the birth of marketing. Maybe the
- 16:07internet was as disruptive as this, when we suddenly, sort of,
- 16:11you know, broke the barriers, and these intermediated new
- 16:15markets were opened and new ways of doing marketing. But I think
- 16:20agentic AI might be even more transformative, because also has
- 16:24a very, very profound cultural and societal element attached to
- 16:28it.
- 16:29Tom Ollerton: So a point of clarity here for anyone
- 16:31listening to this who isn't as experienced and doesn't have
- 16:34your perspective. So we even gave the example of the agent,
- 16:39the marketing agent that would come up with the all of the copy
- 16:42lines and the Gantt chart and generate all of the ads for
- 16:46argument's sake, I think those are the three examples that you
- 16:49gave. So in what sense is that an agent and not just an
- 16:52automation? Because I think that marketing is very good at coming
- 16:56up with acronyms and words to make itself sound more
- 16:58complicated than it is, and an agent is someone that acts with
- 17:02agency to your point where I can go and do stuff and can have
- 17:06critical thought and then act upon it and then bring back the
- 17:09benefit of that work to you as the employer of that agent.
- 17:14Because what you're talking about there sounds like
- 17:16automation. So I'm, I'm trying to get to where it's like when,
- 17:20when someone says agentic, do they mean an actual agent like
- 17:24or is it really just a process?
- 17:26Francesco Federico: I think it's a very good point and to answer
- 17:29your question, it is indeed automation. It is indeed a novel
- 17:34form of automation. I think the reason why we call it agentic
- 17:38and we come up with a new label is because it's completely new.
- 17:42So previously, automation was very much rule based, and so you
- 17:46would have a defined set of rules, you know, if this
- 17:49happened, then do this, etc. And then came advanced automation,
- 17:55where you could connect various processes together and sort of
- 17:59automate end to end. And to some extent, the agentic AI works
- 18:04using sort of the same paradigm of sort of chaining one action
- 18:08after another to achieve a certain outcome. What is new
- 18:11here is the technology that underlies this automation, which
- 18:15is not just rule based anymore, but is based on generative AI.
- 18:20And again, this brings me to my previous point, that the
- 18:24technology will actually now be able, not just to predict, but
- 18:27also to recommend what to do next, and in some extent, in
- 18:30some circumstances, even do it without human intervention. So
- 18:35for example, if you look at some tools around sales, you had this
- 18:40kind of automation for years where, for example, the tool
- 18:43would tell you in a marketing funnel, would identify some
- 18:47leads that are not progressing as quickly as they should down
- 18:51the funnel, and would flag it to the demand management team for
- 18:56action. But this is just literally some basic sort of
- 19:00statistic models, linear regression or standard
- 19:03deviation. What agents can now do is take action and really
- 19:08understand deeply why that particular lead is not
- 19:12progressing and recommend a series of next steps that are
- 19:16specific to that individual lead and not just a recommendation
- 19:21based on X amount of previous data accumulated in the
- 19:25technology. And so I think this is the shift, is really moving
- 19:29from rule based to tools that can actually, in a way, think,
- 19:35although obviously not how we think as humans, but really sort
- 19:39of profoundly analyse data and generate options and and
- 19:45multiple scenarios and and act on those.
- 19:48Tom Ollerton: So really it's the proactive element, the
- 19:51generative element, that makes it not just robotic process
- 19:55automation or something, and making it actually a different
- 19:58thing.
- 19:58Francesco Federico: Yes, and also I would say the
- 20:00sophistication So really, the ability of understanding what is
- 20:05that they're working on. Again, back to the example of creative
- 20:09you might have someone, a creative designer, coming up
- 20:12with a hero creative for a campaign, and then the agent
- 20:16would automatically create 50 assets, starting from a hero
- 20:21one, and that requires not just, you know, resizing and moving
- 20:25text around, but really understanding what the asset is
- 20:29trying to achieve from a rational and emotional
- 20:32standpoint. And I think this level of sophistication is
- 20:36certainly something new that we didn't have before.
- 20:39Tom Ollerton: And so going back to the Marc Benioff quote, which
- 20:48was, this is the last generation of CEOs that will just order
- 20:51people around, but then they will all also order agents
- 20:54around. So what I'm having trouble with is connecting
- 20:57those, your description and that statement, because, to be blunt,
- 21:01and my limited understanding, because everyone's got limited
- 21:04understanding, because it's so new, is that an agent, an AI
- 21:11marketing agent, in this case, is a tool, right? So the
- 21:14marketer, say, yourself, as a CMO, would go, right, we're not
- 21:18going to use this agency anymore. We're going to use this
- 21:21one person who comes up with an idea, and they will command an
- 21:25agent to produce the 50 ads that you mentioned. However, that is
- 21:29not a that is not a CEO ordering agents around. That's a CEO
- 21:33speaking to a marketing person to use a bit of software. So I'm
- 21:38doing, I'm I'm just trying to get my head around the
- 21:40difference between those things, I think there's a bit of AI spin
- 21:44going on, certainly from some of the players, whereas in reality,
- 21:47it's just going to be a marketing person using a tool,
- 21:49which is what they've been doing for about 20 years or so.
- 21:52Francesco Federico: I think this is very true. I mean, at the end
- 21:54of the day, we talk a lot about AI, but AI, exactly as you say,
- 21:58it is a tool. I mean, it's a technology, and like all
- 22:00technologies, it's a tool. I think the reason why we feel
- 22:05compelled sometimes to come up with new labels and new names is
- 22:09because this is really very new. It's very different from what we
- 22:13have seen in the past. And I can clearly see examples where the
- 22:19CEO themselves will be interacting with agents in a way
- 22:24that is new. For example, there will be... previously you
- 22:28mentioned dashboards. I mean, in the very near future, you would
- 22:31be able just to query dashboards in natural language using your
- 22:35voice. And so you would sit at your desk, your big CEO desk,
- 22:39and ask a question, like, you know, how's the pipeline today?
- 22:42Or tell me what people are told about our brand the latest NPS
- 22:47survey, and the tool would reply in natural language. And so to
- 22:52this extent, you might even say some CEOs will have sort of an
- 22:58agentic Chief of Staff, you know, that will provide them
- 23:01information that would otherwise historically have been given to
- 23:04them by by a human collating all this information. So I think all
- 23:08roles across organisations, from the operational ones up to the
- 23:12CEO, will see some sort of agentic functionality creeping.
- 23:17But then at the end of the day you are right, these are tools.
- 23:21These are technologies that all these people will use. But I
- 23:25think again, given the level of interaction, they can which,
- 23:28which is really novel because you interacted the same way you
- 23:32would interact with another human, this is what I think
- 23:35prompts the speculation as to whether we'll have we can call
- 23:40this a hybrid workforce, because, yes, these are tools,
- 23:43but you interact with them in a way that really feels and sounds
- 23:47incredibly natural. And this is even before you talk about
- 23:51robotics. This we're only talking about voice interfaces
- 23:55or computer interfaces. Imagine when you start adding into this
- 24:00equation, the physical dimension as well. And then you would
- 24:03actually have robots walking around the office performing
- 24:07certain tasks in full, fully integrated with the human
- 24:11counterparts. So are they tools? Yes, of course, this would be
- 24:15still tools, but integrated in the in the fibre of the of the
- 24:22of the office environment in a way that is certainly
- 24:26unprecedented.
- 24:27Tom Ollerton: Yeah, it's great to hear your perspective, and
- 24:29we've offered our clients the ability to be able to talk to
- 24:32their data as well, right? So we obviously produce a ginormous
- 24:35amount of paid advertising for clients, and each one of those
- 24:39ads is tagged, so we have this incredible amount of metadata
- 24:41associated with all of the work that we do, and then that's we
- 24:44Tom Ollerton: whatever it's called these
- 24:45found clients just talk to it: which market's performing best,
- 24:49which ad's performing best, which message's performing best,
- 24:51which psychological theme is driving conversion, all that
- 24:52on the podcast or be a guest, or you think I'm asking the wrong
- 24:54stuff. So we've absolutely evidenced it in our work. But I
- 24:59think that. It just must be incredibly confusing for people
- 25:02going right, the future... the future is in an agentic
- 25:04workforce and and then how do people even get started with
- 25:08that? I mean, we we are building that future as a business, and
- 25:11very excited about this, but I'm just a bit loath to call it
- 25:14agentic, because I don't think that clears it up. I think it
- 25:18confuses it. And I think someone said to me once, where there's
- 25:21mystery, there's margin, I think there's a lot of people making
- 25:24margin out of it by making it seem more complicated than maybe
- 25:27it is. But anyway, that's a bit cynical today on this, on this
- 25:30Monday, the 19th, 19th of May. But Francesco, if anyone wanted
- 25:33to get in touch with you about this, where's a good place to do
- 25:37this, and what makes a message that you'll actually respond to?
- 26:48questions, anything, I'll be super interested to hear what
- 26:50you think. So please email me at Tom at, automated, creative.net,
- 26:57that's Tom at, I'm not gonna bother spelling it anyway.
- 27:00You'll work it out. Thanks so much.