Latest / Shiny New Object - a Marketing Podcast / You can't fake taste - with Charlotte Tilbury's Marukh Budhraja
Transcript
- 0:00Marukh Budhraja: With the kind of dashboards and analytics
- 0:02available to anyone working in social media, it can all feel
- 0:06quite overwhelming. I think that there's a huge role that you
- 0:10need to give to understanding the larger context. Don't forget
- 0:14that everything that you are seeing in numbers is actually to
- 0:18do with human behaviour.
- 0:25Speaker: Are you a fitness loving foodie? Then defy snacks
- 0:28is for you. Defy offers real high quality chocolate you love,
- 0:31combined with functional ingredients, giving you as much
- 0:33as 26 grammes of protein per bag. So you can be both a little
- 0:37bit bad and a little bit good at the same time as a female owned
- 0:40brand, defy dedicates a portion of its profits to other women
- 0:44owned businesses currently sold in the US. You can find us on
- 0:47defy with an eye snacks.com and follow us on social at defy
- 0:52snacks.
- 0:58Tom Ollerton: Hello and welcome to the shiny new object podcast.
- 1:01My name is Tom Ollerton. I'm the founder of automated creative,
- 1:04the creative effectiveness ad tech platform, and this is a
- 1:07podcast about the future of data driven marketing. Every week or
- 1:10so I have the pleasure and the privilege of interviewing one of
- 1:13our industry's leaders, and this week is no different. I'm on a
- 1:17call with Marukh Budhraja, who is Global Head of social media
- 1:21marketing, Charlotte Tilbury, so Marukh, it is great to have you
- 1:25on the show, but some people won't know who you are and what
- 1:27you do. So could give everyone a little bit of background on what
- 1:30you do.
- 1:31Marukh Budhraja: Yeah, for the few people who don't know what I
- 1:33do, I am currently head of social at Charlotte Tilbury.
- 1:38I've been I've been in my role for about two years now, and
- 1:42before that, I've actually had the chance to work with some
- 1:45really, really amazing global brands. So I was at Burberry
- 1:49before this, and I've worked with Kiko Milano and Max Mara.
- 1:53So my experience is a bit in both fashion and beauty, ranging
- 1:58from all levels of retail.
- 2:00Tom Ollerton: Right, fashion and beauty in social media for some
- 2:04massive brands, that it's got to be a moving target, I would have
- 2:08thought on a daily basis. So, I mean, I'm really, really looking
- 2:12forward to this podcast. So, so you've had great titles, great
- 2:15jobs. What's been the best investment of your own time and
- 2:21your own money in your career.
- 2:23Marukh Budhraja: Yeah, this one, it took me a moment to think of
- 2:26and and I realised it's quite left field for for most people
- 2:30say this, but I went to do my master's in Italy, and for me,
- 2:35that investment was learning Italian. And I did that when I
- 2:40was doing my master's, and it is the reason that kind of opened
- 2:45all of those, a lot of doors for me to then go on to working in
- 2:49Italy, which I had the privilege of doing for five to seven
- 2:53years, which was absolutely amazing. And I think that that
- 2:57was fundamental in bringing me to where I am today. So it was
- 3:00learning a new language at the age of 20, but it really opened
- 3:05a lot of doors for me.
- 3:06Tom Ollerton: So you learned it at 20, and then you were working
- 3:09speaking Italian within how long?
- 3:11Marukh Budhraja: within a year and a half.
- 3:14Tom Ollerton: No,
- 3:14Marukh Budhraja: yeah, it was, it was phenomenal. I do have to
- 3:17say, I realise I have an edge in learning languages. I do love
- 3:22that, but it is, it is what gave me so,
- 3:26Tom Ollerton: what is your edge in learning languages? Sorry,
- 3:29like it
- 3:30Marukh Budhraja: really quickly. I love it. It's probably like
- 3:34one of my favourite things to do is, like, learn languages. So
- 3:38Italian was one that I decided to learn when I was quite old.
- 3:41And yeah, it was
- 3:43Tom Ollerton: and what else could go on? There must be, you
- 3:45must have a list of languages you can speak.
- 3:47Marukh Budhraja: I mean, I grew up in India, so we grew up
- 3:50bilingual. So Hindi and English are our core languages. But then
- 3:55I learned a bit of French in school, which I have to say,
- 3:59probably really grounded me for learning Italian later, because
- 4:03they do have similar sort of language structures. And then I
- 4:07picked up a little little bit of Spanish while I was living in
- 4:10Italy. But they just come to me, I'd say, easier than a lot of
- 4:14people I'd see,
- 4:16Tom Ollerton: right? You lucky thing, I'm not one of those
- 4:18people. So if you could learn one language. What would be?
- 4:21Marukh Budhraja: Ooh, maybe Japanese.
- 4:23Tom Ollerton: Nice, nice. Well, I hope next time I speak to you
- 4:25on this podcast, in three, four years time, be like, yeah, now
- 4:28living in Tokyo, doing all that. So brilliant. Okay, cool. So
- 4:32that is, we've never had learning a language before. I
- 4:34don't think so. That's an excellent tip. And I guess what?
- 4:38Yeah, you just opened up yourself to a whole new market
- 4:40by speaking.
- 4:41Marukh Budhraja: It was I in the in the little interview that I
- 4:45did with a headhunter who was at the university that I was doing
- 4:48my masters at, they did the interview in Italian, and
- 4:53thought I was Italian, and so they offered me the job, and at
- 4:56that point they didn't realise they couldn't sponsor my visa.
- 5:00And they liked me enough to then have to do it.
- 5:03Tom Ollerton: So, right, you've impressed us enough. Gosh. So
- 5:09all right, that's a great life career tips. But what about
- 5:13specifically data driven marketing? What advice would you
- 5:17give, or do you give to colleagues and friends to become
- 5:21better data driven marketers?
- 5:23Marukh Budhraja: Yeah, I think with that, I have to say, with
- 5:26the kind of dashboards and analytics available to anyone
- 5:30working in social media, it can all feel quite overwhelming. So
- 5:35I think that there's a there's a huge sort of role that you need
- 5:40to give to understanding the larger context. So always look
- 5:44at the data, but within larger human context, don't forget that
- 5:49everything that you are seeing in numbers is actually to do
- 5:52with human behaviour. And so an example would be for in social
- 5:59we see reach goes down every year around. This is organic
- 6:04reach. Reach goes down every year in the period around
- 6:07November and December. And this tends to kind of worry brands
- 6:11saying, you know, oh, our content isn't performing. It's
- 6:13our holiday content. You know, it should be seeing all of the
- 6:17audience. And then if you think about it in a wider context, it
- 6:22isn't because of the content you're posting being sort of
- 6:25were or not of high quality. It is the case that you're fighting
- 6:29in such a difficult environment because every brand is putting
- 6:35out their paid content at the time because of cyber month or
- 6:40holiday or whatever. So it's to do with understanding the human
- 6:44context, and it's to do with understanding events that could
- 6:47be happening outside of what your numbers are telling you,
- 6:51and always looking at those two things together.
- 6:53Tom Ollerton: So where do you go for your human context? Then? Is
- 6:55that you rely on the media agency for that, or research
- 6:58partner or internal
- 7:00Marukh Budhraja: we rely quite a bit on both social listening
- 7:03tools, but also just within the team. We're very open to like,
- 7:08Why do you think this is happening? And we pose the
- 7:10question to the team and and we think quite anecdotally as well,
- 7:14going like, oh, maybe we posted this too late on a weekend. You
- 7:18know, it's Christmas, and no one's looking at their phones
- 7:21because everyone's at Christmas dinner. So so there is some sort
- 7:25of like anecdotal evidence that comes into it, but it feels
- 7:30quite sort of true, because you see that trend across various
- 7:36Brand Channels. It's not just your brand that's maybe seeing
- 7:39low reach at that time. So So you, if you look at it within
- 7:42the larger picture, it's not only your brand that's seen
- 7:45lower reach, for example, in November and December, because
- 7:49it's it's happening to anyone posting organic content.
- 7:58Tom Ollerton: So you've got a really unusual, shiny new
- 8:01object, right? We've not had this one before, but your shiny
- 8:04new object is taste, right? Which I argue is probably not
- 8:09really very shiny. And youth taste has been around for a
- 8:12while, so you have to explain yourself. Why is taste your
- 8:14shiny object?
- 8:15Marukh Budhraja: Well, I think, like you said, it's definitely
- 8:19not a new thing. I didn't invent the word or the taste, or, you
- 8:23know, existence of the word, but for me, it has to do with
- 8:28essentially cutting through the crowd. We're reaching a point
- 8:33where most brands now have, including small businesses, now
- 8:38have AI tools to be able to create content, and it's very
- 8:43easy to use tools to create content inspired by other
- 8:48content online or or that looks very similar to other things you
- 8:53see online, that can look very polished and elevated. But I
- 8:57think what really differentiates a brand then or content then is
- 9:01taste and the uniqueness of it. And so for me, it is, how do you
- 9:06maintain that in in sort of a world where it's really easy to
- 9:12look very, very elevated and polished without having to, you
- 9:15know, necessarily use the resources that were only
- 9:18available to, you know, brands with a lot of budget,
- 9:22Tom Ollerton: right? So, so to play that back to you, in a
- 9:25world of AI where you can just crank out something that looks,
- 9:29looks like the real thing, you think there's an opportunity for
- 9:33the real thing to to actually do something that has art and
- 9:37craft?
- 9:38Marukh Budhraja: Yeah, absolutely. But I don't want to
- 9:41negate the you know, I am all for using the AI tools. I do
- 9:46think they simplify life a lot. But even for content creation,
- 9:49it's more how do you use it to create content that still looks
- 9:56and feels original and not copied? So, um. So for example,
- 10:01we have explored shooting stills of content and then animating it
- 10:07with AI, because video is usually quite complex to create
- 10:11and shoot. So it is about using it intelligently, but then
- 10:16maintaining a sense of authenticity and uniqueness,
- 10:19because that will be the differentiator.
- 10:21Tom Ollerton: So, so you're seeing is you still do the human
- 10:25bit, you still do the creation. You still do a shoot, you know,
- 10:28product shoot, and then you go, right, you know, just, just make
- 10:31this work in make this a two second video of four second
- 10:34video for the PDP or something like that. Okay, yeah, so, so
- 10:38why have you arrived at that opinion?
- 10:40Marukh Budhraja: I think it's what I see as cutting through.
- 10:43And I even see like the data, if I had to bring it in, does back
- 10:48that. So when you do something that feels truly creative and
- 10:51original and different, it is what performs on social. And
- 10:56that's where we see engagement. That's where the community comes
- 10:59in and, you know, responds, whereas if you do create
- 11:02something that a is maybe already all over their feeds, is
- 11:07flooding everyone's feed already, it just doesn't
- 11:11resonate, or people just treat it as the seventh version of the
- 11:15post they've already seen. So so definitely it is now you know
- 11:20something you can back by data. But, yeah, I just arrived,
- 11:25right?
- 11:25Tom Ollerton: So that is your so looking at your own data, your
- 11:28own content, when it's algorithmic, it's aping
- 11:31something that's already there. But when you take the time to
- 11:34make an idea that's original, it works. And that thing that's
- 11:38gonna be very reassuring to a lot of people in the industry,
- 11:41yeah, because, I think some someone said to me, I think it
- 11:44was on this podcast, they said, the way you need to think about
- 11:46AI is, if it was, if the whatever the bot produced was
- 11:52slow and expensive, would you, would you be happy with it?
- 11:55Marukh Budhraja: Yeah,
- 11:56Tom Ollerton: you know, design, design, a, you know, summer look
- 11:58for whatever, and being Oh, well, that was quick. All right,
- 12:01like, we'll see 25 grand, and took a month. Would you be
- 12:04writing back to the supply? Going, Oh, it's amazing. Unless
- 12:07you are, then it's not good enough, right? I think that's,
- 12:09Marukh Budhraja: yeah, yeah. And, I think we are at a point
- 12:12where, you know, at least, at least, where I work, we are
- 12:16testing AI. But it is, it takes a while to, you know, to reach a
- 12:21point where you can, you can call your output good enough to
- 12:24put online, we are, you know, there's a lot of testing and
- 12:28learning involved. So it is something that you know doesn't
- 12:32happen overnight. It's not quick, but I think it is so
- 12:35essential in in, you know, making things efficient for the
- 12:40future,
- 12:46Tom Ollerton: this episode of the shiny new object podcast is
- 12:50brought to you in partnership with Madfest, whether it's live
- 12:53in London or streamed online to the global marketing community,
- 12:56you can always expect a distinctive and daring blend of
- 12:59fast paced content, startup innovation pitches and
- 13:02unconventional entertainment from Mad fest events. You'll
- 13:05find me causing trouble on stage, recording live versions
- 13:08of this podcast and sharing a beer with the nicest and most
- 13:11influential people in marketing. Check it out at www dot mad fest
- 13:16london.com.
- 13:23So how are you hiring for taste, right? So that's a, you know,
- 13:27you've come on a podcast and you've you talked 100 things you
- 13:30could have talked about, but you've chosen taste as being
- 13:32like the the one thing that's going to make data driven
- 13:34marketing work, right? And that's then, that's a powerful
- 13:38choice. So you must be looking for that in other people. So am
- 13:42I making the assumption that if someone shows up and goes, well,
- 13:44I don't know anything about because my ex So, but it's, I've
- 13:48got this chatbot, and like, you're, would you say no to that
- 13:52person over Ah, well, this person's been worked in the
- 13:55category. They've, you know, the clearly interested in this
- 13:57topic, and they've evidenced that in other channels and
- 14:00stuff, or, I guess the answer would be, you do a bit of both.
- 14:03But I'm curious to know if taste is so important, how do you spot
- 14:06that in people?
- 14:07Marukh Budhraja: Yeah, thank you. You look for creativity and
- 14:09you look for experience that people have had. So how, when
- 14:15you pose them a problem, are they thinking outside the box to
- 14:17solve it, or are the solutions, like very common coming from
- 14:22them, but also just, I do think that having, you know, worked
- 14:27with certain brands, there is a sort of way of working that that
- 14:31comes into your into your sort of daily and the way you
- 14:37approach questions at Burberry, for example, we we would use
- 14:41sort of inspiration moments, not only from the fashion world, but
- 14:46anything cultural as well. So it could be in the music industry.
- 14:51It could be coming in from hair, hair and, you know, barbers and
- 14:58artists and stuff like that. And all of that would come to then
- 15:02go like, how can I use that later in an upcoming project? So
- 15:06it's, it's about looking beyond sort of your conventional
- 15:11competitors as well, and just and just kind of taking that in
- 15:14and being quite a sponge, I think so experience and
- 15:18creativity.
- 15:19Tom Ollerton: So how do you think the whole AI thing's going
- 15:21to play out. So as you say, you're doing experiments, we're
- 15:25working with clients using generative AI to produce an
- 15:29increasingly large amount of their work. And the way that
- 15:33we're looking at it, I'd love to know your perspective on this,
- 15:36but what we always say is, don't just use it to save time and
- 15:40money, use it to do the thing that you couldn't do before. So
- 15:45if you, you're launching a product and and you're going to
- 15:49do a bunch of ads, like, don't just make the same one ads that
- 15:52you're going to make before in a 10th at a time, and, you know,
- 15:55high five the CFO, because you've saved some money. Think,
- 15:58right, okay, I literally have the power to generate anything.
- 16:02So if I could trade anything, I could test everything. So we're
- 16:06that's way the way we're using it. So you test a really broad
- 16:08range of different formats and styles and content and copy and
- 16:12the looks and feels and all the rest of it in the ads, and then
- 16:14you get the data back on what works. So then you're in this
- 16:16you're building every campaign on the the intelligence of the
- 16:19last whereas the other brand over there has gone, haha, we
- 16:22did our one ad, and we save, you know, 15 grand, or something
- 16:25like that. So that's how we're doing it. Like, how can, how can
- 16:28we make sure that brands are putting the very best bit of
- 16:33creative behind every media dollar that they have?
- 16:35Marukh Budhraja: Yeah, I think that's that's a really good way
- 16:37of looking at it. I think in the long run, people are going to
- 16:42use it for both. I think at the moment, the focus is in in
- 16:46average businesses is very much just on making it efficient. So
- 16:50how can we, you know, save time and resource. But I think that
- 16:54the ideal way of using AI in the future is very much, how do we,
- 16:59how do we remain efficient, but also, how do we achieve
- 17:02something that we wouldn't be able to, you know, with the
- 17:06resource that we currently have and and that sort of is the
- 17:10potential that they have, and that's that's where they can
- 17:14really bring in the creativity and authenticity. And it doesn't
- 17:17like you could focus on doing something extremely surreal with
- 17:21your content that you know would take maybe CGI, and you know the
- 17:27kind of creators that you cannot afford right now at all or any
- 17:32day, yeah,
- 17:34Tom Ollerton: and if you are, you get... I'm aware it's
- 17:36probably not your end of the business. But are you getting
- 17:39any I want to say heat, but I mean influence from the team
- 17:45that are responsible for generative optimization. Sorry,
- 17:48generative engine optimization. So Tom Telford on the podcast,
- 17:52it's not out yet, and he was talking about how the way to do
- 17:56Geo, LM search or AI search really well is to is to have a
- 18:02really cohesive, broad strategy where every time you show up in
- 18:05the world, it's good, so therefore that's something that
- 18:07an LLM can scrape. You know, it's not about like somebody,
- 18:10it's not about gaming that actual algorithm. It's about
- 18:13just showing up in in the right way in the world, across PR and
- 18:16press and social. So I was just curious to know, are you feeling
- 18:19that pressure to think, right? Well, okay, yes, good, great job
- 18:23in social but actually, it's got to be done in a way that shows
- 18:26up in in that kind of search.
- 18:28Marukh Budhraja: Yeah, we have had a few of those
- 18:30conversations. Interestingly, Charlotte Tilbury was one of the
- 18:33most recommended brands on an AI forgive me. If I get this wrong,
- 18:39I think it was chat JBT searched over the holiday season, and I
- 18:44think it's a holistic like you said, it's a holistic job across
- 18:48PR and the website, but also I know, like Reddit is a big
- 18:54source for where your llms Pick up A lot of that information. So
- 19:02what are your fans saying about your product on Reddit? I know a
- 19:05lot of brands don't actually, you know, have profiles on
- 19:09Reddit, but it's a huge community there. So what is
- 19:12being said about your product there on social as well? I think
- 19:16there is not, not as much, because we do want to keep our
- 19:20copy quite like, relevant to to our audience, that that is an
- 19:24effort to like, include the words that we would want to show
- 19:29up if someone was using, for example, Tiktok, like a search
- 19:32engine.
- 19:33Tom Ollerton: Yeah, that's such a result over the holiday
- 19:35period. Number one, there was some some high fives going
- 19:38around there. Brilliant. And, right, okay, because, yeah, I've
- 19:42heard, I've heard Reddit described, sorry, chat GPT
- 19:46described as a wrapper for Reddit, or a search engine for
- 19:49Reddit, which is quite and it's always interesting to see the
- 19:52way that, like x or Twitter was so important. 10 years ago, it
- 19:55was, it was huge, and now it's this nowhere, and all of that,
- 19:58all that money. Probably it's gone, gone over to Reddit and
- 20:02with, with the llms, you know, make, making that a very
- 20:05understandable source of data for brands. Because I think
- 20:08beforehand, it was terrifying you like Reddit,
- 20:11Marukh Budhraja: It's years and years of text, which is, which
- 20:14is what you know the LLMs need.
- 20:15Tom Ollerton: So, yeah. So, so, last question, so one of the
- 20:25parts of AI that's really interested in me at the minute
- 20:27is how limited text is, right? You know, so, as you say,
- 20:32there's years and years of conversation about text, but
- 20:35actually, text isn't always great at carrying certain types
- 20:40of meaning, right? Where, for example, video is right. So I
- 20:44think the example, like I keep on using I read somewhere, is
- 20:47like, you have lots of text to say, if you knock a glass off a
- 20:49table, it will smash and it will kind of go everywhere, right?
- 20:52You know that that's true. But if you watch a like, 10,000
- 20:56videos of a glass falling off a table and an AI would have a far
- 21:01more nuanced understanding and understanding of that. But,
- 21:05yeah, all these algorithms are mostly trained on text. So I
- 21:11wonder, where are those kind of natural nuances? And I wonder if
- 21:13your industry would actually, you know, like, how, like, such
- 21:18a visual industry that really, how accurate context be, in some
- 21:22senses, versus a video, a tutorial or a demonstration, or
- 21:26those kind of things. So curious, curious to know if
- 21:29you've seen any, any indication of that shortcoming of AI
- 21:35happen?
- 21:35Marukh Budhraja: Um, I think so, like maybe, maybe from a
- 21:38slightly different angle, that that is a challenge that comes
- 21:41up so much when you're prompting creatively, right? So when you
- 21:46when you want to create a video on Nano Banana or whatever you
- 21:50you realise the limitation and how specific your prompt needs
- 21:55to be, because the AI obviously doesn't know from its visual
- 22:01library, what might happen to a product if it goes from point A
- 22:05to point B or transforms in a certain way? So that limitation
- 22:09might have come up in that way where you you really have to
- 22:12take your time in making your prompts as detailed as possible
- 22:15at the moment, so that the text to video. You know, generation
- 22:20is is more accurate, and that's probably in the most, the most
- 22:25relevant to that.
- 22:28Tom Ollerton: Well, I think it just, just shows that, like
- 22:31we're all working out, and I really, I'm quite confident and
- 22:35enjoy saying, like someone's just, I don't understand, like
- 22:38it's, there's so many different ways to cut it and to do it. And
- 22:42it's interesting to hear that you guys are kind of winning in
- 22:45AI search already, but also clearly experimenting with AI.
- 22:49But at the core of it, you're, you're you are not going to give
- 22:53up on taste. And that's your shining the object. It's like,
- 22:56you know that the human, the human joyous love of the craft,
- 23:00and the detail is the thing will matter in data driven marketing.
- 23:04What a beautiful point. So thank you so much for coming on the
- 23:07podcast. If someone wants to get in touch with you, where can
- 23:10they do that and what makes a message that you'll actually
- 23:14respond to?
- 23:15Marukh Budhraja: I'm happy to connect on LinkedIn. So my
- 23:18LinkedIn is just my name, but I'd say, if you want a response
- 23:24from me, keep it short and make sure you've got a good opening
- 23:29hook. I work in social so the hook is what matters.
- 23:33Tom Ollerton: Well, brilliant. I love it. Thank you so much for
- 23:36your time.
- 23:37Unknown: Thank you.
- 23:44Tom Ollerton: Hi, just before you go, I'd really appreciate it
- 23:48if you could take the time to write a review of the shiny new
- 23:52object podcast on Apple podcasts or iTunes, whatever it's called
- 23:56these days, or whichever podcast provider you use. We're at any
- 24:00podcast. So it would go a long way for us if you could just
- 24:03share the Word and give us a bit of a support on those channels.
- 24:06That would just be fantastic. If you haven't got time. That's
- 24:08also cool. And yeah, if you could tell your colleagues about
- 24:11the podcast, and also, if possible, don't forget to
- 24:15subscribe. And I'd love to hear your feedback, if you'd like to
- 24:18speak on the podcast or be a guest, or you think I'm asking
- 24:21the wrong questions, anything, I'd be super interested to hear
- 24:24what you think. So please email me. Tom at, automated
- 24:29creative.net, that's Tom at, I'm not gonna bother spelling it
- 24:34anyway. You'll work it out. Thanks so much.