Latest / Shiny New Object - a Marketing Podcast / Episode 294 / Victoria Lorenzana / Tala / Senior Product Marketing Manager
Transcript
- 0:00Victoria Lorenzana: It's okay to make mistakes, ask questions.
- 0:03You don't need to be perfect the first time, we all here are
- 0:06learning.
- 0:13Speaker: are you a fitness loving foodie? Then defy snacks
- 0:15is for you. Defy offers real high quality chocolate you love,
- 0:18combined with functional ingredients, giving you as much
- 0:21as 26 grammes of protein per bag. So you can be both a little
- 0:24bit bad and a little bit good at the same time as a female owned
- 0:28brand, defy dedicates a portion of its profits to other women
- 0:31owned businesses currently sold in the US. You can find us on
- 0:35defy with an eye snacks.com and follow us on social at defy
- 0:39snacks. Thanks.
- 0:46Tom Ollerton: Hello, and welcome to the shiny new object podcast.
- 0:49My name is Tom Ollerton. I'm the founder of automated creative,
- 0:52the creative effectiveness ad tech platform, and this is a
- 0:55podcast about the future of data driven marketing. Every week or
- 0:58so I have the pleasure and the privilege of interviewing one of
- 1:01our industry's leaders about where they think the future of
- 1:03data driven marketing is going. So if that's your kind of show,
- 1:06stick around, because I am on a call with Victoria Lorenzana,
- 1:11who is in Mexico, and she works with Tala, and she is a Senior
- 1:15Product Marketing Manager. So Victoria, for anyone who doesn't
- 1:19know who you are and what you do. Could you give us a bit of
- 1:22background?
- 1:23Victoria Lorenzana: Sure. Thank you so much. Tom for having me
- 1:26on this podcast. And yes, let me introduce myself. I studied
- 1:30social psychology in college, and something that has defined
- 1:33my professional career is the change. So I start working as
- 1:37market researcher, then I jump into the UX research world, then
- 1:43into the UX design world, and then into product manager. So
- 1:47currently, as you mentioned, I'm a product marketing manager. I
- 1:51have a lot of experience on analysing user behaviours,
- 1:55understanding market needs and marketing positioning, and I
- 2:00will say that one of my strengths is this ability to
- 2:04effectively switch into different careers, because my
- 2:07main interest is to make sure that I'm making the most impact
- 2:11into the organisation that I am collaborating. So that's all.
- 2:16That's me. I live in Mexico City. I'm Mexican. I'm huge fan
- 2:21of cats and horror movies, and that's me.
- 2:25Tom Ollerton: Is there any horror movies that features cats
- 2:28of note?
- 2:31Victoria Lorenzana: actually, but not any, any horror movie
- 2:34that involve involves cats? Comes to my mind only one about
- 2:39this kind of elf, which is like, uh, evil creature, and the cat
- 2:45is the hero, because he saves the little girl from this elf,
- 2:50or gnome or something like that...
- 2:52Tom Ollerton: Right. I don't work, yeah, we'll put that in
- 2:54the show notes. Uh, but maybe there's a gap in the market for
- 2:57cat based horror films, you know, maybe that, maybe that
- 2:59could be another career switch for you anyway, back to back to
- 3:03marketing. So if you were going to advise a student who was
- 3:07doing all the right things, like they were motivated, building
- 3:10their network, making content, all that stuff, what would you
- 3:13advise them to get started in this industry and make a career
- 3:16in data driven marketing?
- 3:17Victoria Lorenzana: I will say that this is a great question,
- 3:20actually, because I have been mentoring young students across
- 3:24the globe, and my biggest advice is to be fearless. I know that
- 3:31it may sound challenging, but when we are recently graduate
- 3:35and when we are looking to introduce ourselves in an
- 3:38industry or area, we have these huge expectations on ourselves.
- 3:44So I will say that the first advice, or the most important
- 3:49thing, is that we need to understand that we all, no
- 3:53matter the expertise level, we all are continuously learning,
- 3:57and it's okay to make mistakes, ask questions. You don't need to
- 4:02be perfect the first time. You don't need to execute perfectly
- 4:05everything. So it's okay to make mistakes. We all here are
- 4:09learning, and the other important thing that I want to
- 4:12mention is, as we keep learning, it's important to have guidance,
- 4:17right? So maybe find these figures that represents for you,
- 4:22like the ideal professional profile that you want to become,
- 4:27and ask, do these profiles to be your mentor or ask for advice,
- 4:34because you don't need to be alone on this career path. You
- 4:38can be also, you can have also the follow up and guidance from
- 4:46more experienced profiles, so feel free to ask for help and
- 4:49guidance. That's all my advices.
- 4:59Tom Ollerton: So that's fantastic advice for someone
- 5:01looking to make their first steps in the industry. But I'm
- 5:03also keen to know what is your advice? What's your best top bit
- 5:06of advice to become a data driven marketer, so when you're
- 5:09talking to your colleagues and people on your team, what is the
- 5:12silver bullet, golden rule, bit of advice that you find yourself
- 5:15sharing most often?
- 5:16Victoria Lorenzana: Okay, so I will say that make sure that you
- 5:20understand your audiences. Because in marketing, something
- 5:25that happens a lot is that the area the industry require us to
- 5:30be creatives, right? And it's fantastic. This is one of the
- 5:34things that I really appreciate about marketing that sometimes
- 5:38you need to rethink, or you need to think about how to tackle
- 5:42different challenge from a creative perspective, however
- 5:46you need to always keep in mind who is your customers, who are
- 5:50your audiences. By having this deeper understanding of them, of
- 5:54your audiences, their needs, and also by taking a look to the
- 5:58competitive set, you will find out the right answers. You will
- 6:02find what you need to communicate, and you will find
- 6:06the most effective ways to communicate these messages or
- 6:09position your product and things like that. So having this
- 6:12understanding is critical, and that means that you need to have
- 6:16these research skills at least high level to understand who are
- 6:21your customers.
- 6:22Tom Ollerton: And what are those research skills that you suggest
- 6:25that people have? So I totally buy into got to understand your
- 6:29audiences. Know who your customers are, but what do you
- 6:31find are the most successful ways? Are you heavy quant, or
- 6:34are you in person interviews? So what is your preferred approach
- 6:37to really understand the person that you're trying to speak to?
- 6:39Victoria Lorenzana: Yeah, this is a great question. I would
- 6:42like to think in terms of internal sources or current
- 6:48sources of information that in the organisation you may find
- 6:52like, for example, take a look to your users' demographics. A
- 6:56lot of companies that their user demographics for different
- 7:00purposes, and this is a nice starting point. By taking a look
- 7:04to the internal data that you may have around your customers,
- 7:09in order to have a nice picture about who are them. Also
- 7:13internally, you can take a look to previous research pieces,
- 7:16because a lot of companies have these user research areas,
- 7:20insights areas or marketing intelligence areas in which you
- 7:24can find these different information sources about your
- 7:27customers, their needs, their preferred channels, consumption
- 7:35behaviours and things like that. So just by having the ability to
- 7:39take a look and find out internal data sources or
- 7:43internal inside sources around your customers, you will be able
- 7:48to keep in mind or effectively understand what are your
- 7:51customers. But if you want to go ahead and develop by your own
- 7:57research, I will say that it's worth for people to have this
- 8:01basic knowledge about how to conduct a successful research
- 8:05project, and this is completely linked with the research
- 8:09objective. So for example, let's say that you don't have these
- 8:12internal data sources, and you want to understand, for first
- 8:16time, who are your customers, you will need to be able to
- 8:20design a survey and effectively reach out these customers,
- 8:25customers, to ask them about their demographics, behaviours,
- 8:30preferences, usage behaviours and things like that. But if you
- 8:36want to have a deeper understanding on a specific
- 8:39behaviours or needs, you need to think on qualitative terms. And
- 8:44maybe for these purposes, you may be it's more it's more
- 8:49beneficial for you to conduct this kind of qualitative
- 8:52methodologies, like, for example, in deep interviews, sit
- 8:57in front of your customer and ask this customer about their
- 9:04behaviours, daily life needs around the the product, the
- 9:09service that you are offering, and things like that. So it's
- 9:13all relate with the research objective, how you can apply
- 9:17different methodologies. But I think that there is a lot of
- 9:21information sources and a lot of guidance from the different
- 9:26organisations that you can take a look and use them as
- 9:32references. Because for external you want to to develop by your
- 9:38own research projects, if you want to execute you will find a
- 9:42lot of guidance on internet from from very, very accurate
- 9:46sources.
- 9:47Tom Ollerton: This episode of the shiny new object podcast is
- 9:56brought to you in partnership with Madfest, whether it's live
- 9:59in London. Or streamed online to the global marketing community,
- 10:03you can always expect a distinctive and daring blend of
- 10:06fast paced content, startup innovation pitches and
- 10:09unconventional entertainment from madfest events. You'll find
- 10:12me causing trouble on stage, recording live versions of this
- 10:15podcast and sharing a beer with the nicest and most influential
- 10:18people in marketing. Check it out at www.madfestlondon.com.
- 10:29So now we're going to move on to your shiny new object, which is
- 10:33the use of AI to analyse customer feedback. So this feels
- 10:39incredibly cutting edge and futuristic. I think I know what
- 10:42you mean by that, so but for someone who's listened to this,
- 10:45and I'm not sure what that means, could you explain what
- 10:47using AI to analyse customer feedback is, and how do you do
- 10:50it, and why it's a good thing?
- 10:51Victoria Lorenzana: Well, sure, first of all, I want to start by
- 10:55sharing why it's important to listen the customer feedback,
- 10:59and it's because we design products to effectively meet
- 11:02customer needs, right? This is our main objective from all
- 11:07companies, from all services. But how we can effectively meet
- 11:12these consumer needs, we need to understand that we are in a
- 11:15continuously change environment, and consumer needs may change
- 11:20quickly according to different lifestyles or adoption or for
- 11:26according to different competitor softwares and things
- 11:29like that. So this is why it's important for us to keep an eye
- 11:34on the customer feedback, but also in order to be able to
- 11:38optimise our product service or offer to effectively meet
- 11:42customer needs. So listen our customers. Is the key for a
- 11:45successful business. It's also completely related with how we
- 11:51can strategically meet our business objectives. So having
- 11:56said that, that customer feedback matters and it's
- 12:01important for us to keep an eye on that. I will say that why
- 12:05this is new AI. We have listened a lot of AI, and I think that
- 12:12from different perspectives, at the beginning, it sounds like
- 12:16scary, right? Like AI has the capability to effectively design
- 12:22product experiences, or also to design copy and made some
- 12:27straightening in terms of specific roles or tasks that we
- 12:32are continuously developing, particularly for example, in
- 12:36marketing. So in marketing, one of the most critical elements is
- 12:39to be able to design a right marketing copy for different
- 12:43purposes, a tight line value propositions or things like
- 12:46that, but leaving aside the scary part of the AI and this
- 12:52idea of replacement for different professional profiles,
- 12:56I think that AI, it's super beneficial if we understand how
- 13:01to take advantage of it and how we can how identity and it's
- 13:06critical in terms of identifying what are the most challenging
- 13:10tasks that we are continuously developing on our daily basis,
- 13:15and how we can take advantage of the AI capabilities. So by
- 13:19merging these two ideas of how relevant is to listen the
- 13:22customer feedback and how effective is the AI to simplify
- 13:27tasks. This is why I am bringing this proposal to effectively
- 13:31using these tools to analyse customer feedback. In the past,
- 13:36for example, to analyse customer feedback, and this is coming
- 13:40from my experience as researcher, we need to review
- 13:46each of each one of the customers' comments. For
- 13:50example, let's think on Google Play Store reviews, or let's
- 13:54think on social media comments. Or, for example, when we have
- 14:00this kind of customer surveys, and we have open end questions.
- 14:04We needed to go and review each one of the customer responses.
- 14:09And I'm a little old school in terms of research, what we used
- 14:15to do is that we codify each response based on the answer
- 14:21based on this topic, but it was a manual process, so you need to
- 14:25go through each one of these answers, each one of these
- 14:28comments, codify them, and after that, being able to effectively
- 14:33categorise them, and also to present a kind of quantitative
- 14:37landscape in order to highlight what are the most frequent
- 14:40comments or topics across customer answers. It was a very
- 14:46hard task. And for example, let's say that when you had a
- 14:50sample of 1000 responses, that means that you needed to
- 14:54manually review 1000 comments. And usually what happens. Is
- 15:00that customer responses, customer comments, not only
- 15:03refer to one single topic when a customer is providing feedback,
- 15:08customer may address one, two or three or more topics within one
- 15:13answer. So the complexity of analysing qualitative
- 15:16information is huge, but currently, with AI, we can
- 15:21simplify this, this task, and I will explain how a Tom. I just
- 15:29want to make a pause to make sure that we are good.
- 15:32Tom Ollerton: Yeah, this is fantastic. I love that you're
- 15:34going into a very practical demonstration of how to do this.
- 15:37Yes, I'm all ears. Thank you.
- 15:39Victoria Lorenzana: Okay, perfect. So. But also, let me,
- 15:42let me also mention another way in which we used to analyse
- 15:45qualitative comments, and it was by programming on SQL or even
- 15:50using Excel capabilities. But what happens? What used to
- 15:54happen with this kind of analysis is that it was hard for
- 15:57the different softwares, even for Excel or for SQL to
- 16:02effectively find out multiple responses within a single answer
- 16:08from a customer. So at the end of the day, you miss part of the
- 16:12information. And also with this manual coding that I was
- 16:15mentioning, the risk was that, since was a purely human
- 16:19process, it was, for sure, sensitive to human mistakes. So
- 16:24what happens if you are very exhaust and you are not properly
- 16:29doing the coding or identifying the different topics? It was a
- 16:34challenge, right? So now what we can do is by having the right
- 16:40prompt to any AI tool to well, particularly, I have experience
- 16:46by using two AI tools, ChatGPT, and the other one is Google
- 16:51Gemini. I have a successful experience on implementing this
- 16:56tool and using these tools to effectively coding customer
- 17:00qualitative customer responses by prompting correctly. So what
- 17:06you need to do is to select a significant sample of these
- 17:09qualitative responses. And for these, I need to refer to how to
- 17:14effectively identify a significant sample. There are
- 17:17some online calculators that help you to estimate the right
- 17:22sample. So for example, for 1000 of comments, the right sample or
- 17:29the accurate sample, you don't need to analyse 1000 comments.
- 17:33So a statistical level, you need to find the right number of
- 17:39answers. And by finding this right number, you will find the
- 17:44right representativeness. Cross the World sample. So for
- 17:49example, for 1000 I will say that with a nice statistical
- 17:53level, is like 200 responses, just one example. So on
- 17:59satisfying this sample, this significant sample, and after
- 18:04entering the right prompt to ChatGPT or Google Gemini, you
- 18:09will find these qualitative analysis within seconds. And you
- 18:16can ask to these AI tools to provide you not only the list of
- 18:22the most recurring topics, but also you can ask to these tools
- 18:26to give you also the percentage the number of time that this
- 18:32topic was mentioned across the customer comments, for sure that
- 18:37my understanding is that This kind of AI tools are still in
- 18:41progress to keep optimising their models, their analysis
- 18:46models, and sometimes the tool can crash out a little bit. So
- 18:52what you need to do is just retry, re ask, and see and see
- 18:59if the tool is working fine and providing you the right data,
- 19:05but in my experience, I will say that I have run at least 20
- 19:09analyses by asking these tools to analyse Google Play Store
- 19:13comments or open end responses from very complex
- 19:17questionnaires, and it has been very successful, and it has been
- 19:22super efficient for me, because in the past, for example, with
- 19:25the manual coding, it took like one week of purely analysis of
- 19:32pure analysis of the open end responses. But now, with these
- 19:36tools, ChatGPT and Google Gemini, you can have this
- 19:41analysis of open end customer feedback responses within
- 19:46minutes in one day. So it's super efficient. And I think
- 19:50that in the past, for example, several organisations were
- 19:54scared to analyse their customer feedback sources, right? Because
- 19:59the complexity of the analysis, but now, but using these tools,
- 20:03you can have a recurrent reporting on the customer
- 20:07feedback, on the open end comments, without requesting too
- 20:12much energy on the analysis. And you will be aware and keep keep
- 20:18yourself up to date on what customers are saying. So I will
- 20:23say that just a quick recap, you can use this methodology for
- 20:27social media listening purposes, in order to analyse customer
- 20:31feedback, sources like CSAT surveys, NPS, the open ended
- 20:36responses, or to analyse when you launch a questionnaire and
- 20:40you have open ended responses, and you can also use this
- 20:44methodology to analyse Google Play Store or App Store reviews.
- 20:56Tom Ollerton: So what I'm hearing is that it was a very
- 20:58laborious process, even with SQL or with Excel skills, and this
- 21:04has fundamentally saved you a lot of time. So you've got open
- 21:07ended responses. It's all messy. One of the AI providers, GPT or
- 21:13Gemini, can just instantly go through that and work out
- 21:16percentages. So is it purely a time saving thing here, or have
- 21:19you used it to help spot things that you wouldn't have spot
- 21:23before?
- 21:24Victoria Lorenzana: Yeah, I will say that the advantages, let me
- 21:28list some of the advantages that I have seen so far. For sure
- 21:32that time efficiency is the most relevant. The other one is, for
- 21:36example, for SQL or for Excel, you need specific skills, but
- 21:41now with GPT and Gemini, everyone can prompt these tools
- 21:46and run the analysis. So I will say that is accessible for
- 21:51everybody, without the need to having a special training. And
- 21:54the other advantage, for example, in comparison with the
- 21:57analysis that you can run with SQL, is that with SQL, you have
- 22:02certain limitations, and maybe you can skip certain topics. So
- 22:07in my experience, with these AI tools, you have more accurate
- 22:12analysis, and even over, for example, the manual coding, the
- 22:16manual review, because on this methodology, with the manual
- 22:20reviewing one by one. You are sensitive to make mistakes, but
- 22:27with GPT and Gemini, the accuracy is impressive. So this
- 22:33is a nice way for you to have a very, very accurate landscape
- 22:38about customer feedback topics without losing information.
- 22:43Tom Ollerton: So unfortunately, we've come to the end of the
- 22:45podcast, and I really appreciate you helping me understand and
- 22:48the listeners understand more about your approach. It's very
- 22:51rare that someone lays out exactly how they do what they
- 22:54do, so I'm very grateful for that. So if someone wanted to
- 22:57continue this conversation with you, how would you like them to
- 23:00reach out to you? Where do you want them to reach out to you?
- 23:03And what makes a message that you will respond to?
- 23:06Victoria Lorenzana: Oh, awesome, sure. So you can find me on Ali
- 23:09B list as mentor or on LinkedIn, Victoria Lorenzana, and the only
- 23:14thing that you need to do is reach me out and let me know if
- 23:19you want to learn more about the usage of AI for analysed
- 23:23customer feedback, or if you want to learn a little more
- 23:28about my experience switching careers, I'm super happy and
- 23:32excited to provide mentoring to everybody or support.
- 23:36Tom Ollerton: So yeah, fantastic, Victoria. Thank you
- 23:39so much for your time.
- 23:40Victoria Lorenzana: Sure, thank you so much for having me.
- 23:48Tom Ollerton: Hi. Just before you go, I'd really appreciate it
- 23:52if you could take the time to write a review of the shiny new
- 23:55object podcast on Apple podcasts or iTunes, or whatever it's
- 23:59called these days, or whichever podcast provider you use, we're
- 24:03in any podcast, so it would go a long way for us. If you could
- 24:06just share the Word and give us a bit of a support on those
- 24:09channels. That'd just be fantastic. If you haven't got
- 24:11time. That's also cool. And yeah, if you could tell your
- 24:14colleagues about the podcast, and also, if possible, don't
- 24:17forget to subscribe. And I'd love to hear your feedback if
- 24:21you'd like to speak on the podcast or be a guest, or you
- 24:24think I'm asking the wrong questions, anything I'd be super
- 24:26interested to hear what you think. So please email me. Tom
- 24:31at, automated creative.net that's Tom at, I'm not going to
- 24:36bother spelling it anyway. You'll work it out. Thanks so
- 24:39much.