Webinar
Marketing Automation: The Process Revolution

Marketing Automation: The Process Revolution

Discover, in this video, how marketing automation is redefining the way marketing teams operate, communicate and generate results. Through concrete cases and practical examples, we explain how to free up time from repetitive tasks, ensure consistent follow-ups and create personalized experiences at scale. This is the silent revolution transforming marketing, and your brand can, and should, be part of it.

In this webinar, discover:

✔ How to use automation to free up time and enhance creative work;

✔ How brands can maintain closeness and personalization, even at scale;

✔ Practical strategies to integrate automation without losing the brand's DNA.

(00:00) [Music] who will explain the topic of marketing automation, the process revolution, and the use of AI to automate campaigns from segmentation through to results analysis. Tiago, whenever you are ready. Fantastic. Thank you very much, Nuno. And brilliant, Hugo. It was also fantastic to learn from you.

(00:32) This is very interesting because these topics, and what each one of us will share, are completely interconnected and interdependent. There is even some overlap between each of these subjects. Mine will be no different. Let me just share my screen and please confirm that everything is working properly.

(00:54) You can see a slide on my side, correct? Fantastic. Thank you. About me, I will bring you a little into this world and continue the principles we have just heard about personalization, which Hugo addressed. My world is the world of Liminal. We are a team of management, marketing, communication and sales professionals working with software to manage customer relationships.

(01:21) What we do is implement CRM, automation and business intelligence platforms, as well as widely known players such as Salesforce, HubSpot, Zoho and others. For more than eight years, we have been working with brands you all know, which are very successful clients on our side.

(01:39) Recently, we compiled this knowledge into a book, which I will also promote here so that you can consult it. It is available on Amazon for anyone who wants to understand how to connect marketing strategy and operations to platforms such as CRM and automation, and how to use artificial intelligence to reach this level of hyper-personalization.

(01:59) This book addresses this world. We are going to talk about automation and the revolution we can achieve with artificial intelligence by using marketing automation. I want to start by telling you a story. And this connects so well with what you were just talking about. A little over 10 years ago, the father of a teenage girl in the United States called his supermarket chain, in this case Target, in a very bad mood.

(02:28) His complaint was very simple. He said, you are harassing my 16-year-old daughter with advertising and flyers that only talk about and highlight products for mothers, pregnant women or people who have just had children. Stop this, it is nonsense. I feel offended. Well, the manager who took note of the complaint did what he had to do and tried to gather information from headquarters.

(02:54) That is when he realized that Target's marketing department, this American retail chain, was developing, at the time, a consumer analytics and marketing automation project. The goal was to reach mothers or pregnant women who had declared that status, as well as women who were potentially pregnant. They were not certain, but there was a high probability.

(03:21) From there, they would personalize this type of communication and promotional offer. The manager did what he had to do, returned to the conversation with the dissatisfied customer, explained all of this, explained that it was a pioneering program and that there had probably been some mistake in the algorithm or something like that, and apologized repeatedly. At the end of the call, the dissatisfied customer said: well, in fact, my teenage daughter really was pregnant, and no one in our house knew.

(03:49) Just because we can does not mean we should. The New York Times uncovered this in considerable detail. It concluded that Target, even at that time, with only 25 purchased items, could predict with near-perfect accuracy the probability that someone was pregnant in the second trimester of gestation and, from there, automate ultra-personalized communication actions.

(04:17) Just because we can does not mean we should. Why does marketing have this obsession with hyper-personalization? Since the 1980s and 1990s, this idea has been precious to marketing and is perhaps the holy grail we are all looking for. We know that customers and consumers who have a dazzling, fantastic consumer or brand experience not only become happier and more loyal to the brand, but are also willing to spend more, at least 16% more.

(04:51) And what is a spectacular consumer experience? It is exactly what Hugo already talked about. It is a personalized experience. It is what we all want. We want brands to recognize us in our individuality, or at least to be relevant and not bother us with offers and communications that have nothing to do with us.

(05:10) If Steve Jobs' aspiration was this, to anticipate and even know more than the consumer, the truth is that most brands today are still far from this. But this is where we want to go because there is value in it. Now, more than 10 years after the Target example I shared, the truth is that this predictive and automation power is available to any company today, not only to large brands or marketing departments with large budgets or great IT teams. Any company today can start developing this type of application, and that is what I will show you with some live examples.

(05:46) But before jumping into the use of technology, we always, always, always have to think about strategy. I would even say business strategy, not just marketing strategy.

(06:07) What does this mean? It means that the importance of what we will do from here with technologies depends on strategic intent. If our intention is only to grow, if our intention is only to have more clicks, more likes, more conversions and more sales, then please, let us use all these tools to become the best and biggest spammer in the world.

(06:34) That does not contribute to a healthy future for our brand in the medium term. If, on the other hand, our intention is to develop our brand, our presence, and development is different from growth, because development assumes that yes, we grow, but the rest of what is around us also grows, namely the relationship and closeness with our customers.

(06:56) If this is the strategic intention, then we have what we need to dive into the technological world with the goal of strengthening our relationship with our customers, rather than simply exploring ways to influence better and more intensely. With this in mind, where should we start a technology strategy that seeks to achieve these objectives? There are three deeply fundamental pieces here, and most companies often start building their marketing technology stack from the surface, forgetting the foundations.

(07:29) These foundations are, in essence, the central nervous system of the customer relationship. These three technological pieces are CRM, automation and business intelligence. I will give you some examples of how these three solutions can and should work in an integrated way, including how artificial intelligence should support us in this.

(07:53) Starting with CRM, I will show you a few different environments, from HubSpot to Salesforce and others. Let us start with HubSpot. The CRM concept is not new. The challenge here is that it cannot be just a police-control tool for sales and customer-facing teams. How many calls did you make? How many proposals did you win or not win? No.

(08:15) CRM has to be a strategic tool and it has to be a tool that enables a first step towards omnichannel. What does this mean? It means that CRM must be connected, at minimum, to all our digital channels. The website, for example. We must be able to understand that this person in our database, whether a customer or a potential customer, reacted to an email that was sent to them, watched a video from one of our salespeople, but also visited our website, consulted this, that and other information, spent a certain amount of time browsing the website or reacted to a paid campaign, and so on.

(08:43) Omnichannel connected to the website. On the other hand, there are strategic elements for marketing that the CRM should also be a source for. There is nothing more strategic in marketing than segmentation.

(09:04) One thing we should demand from our CRM is precisely this: that it tells us where each person in our database, customer or not, is in the decision journey, the customer journey. It is very different to speak to a customer who is returning, who had left us and is now returning, than it is to speak to a lead, or to a lead that is already in the hands of sales. It is also important to understand which segment or persona each person in our database belongs to.

(09:37) It is from a segmentation exercise that we can understand our reach capacity in segment A, B, C or D, and how our entire marketing mix, which is not just communication, but also price, distribution and so on, is performing for each of the segments we target.

(09:57) This is elementary, yet most companies do not get here. Let us bring in artificial intelligence. Why? Because with predictive artificial intelligence, whether within the CRM suites themselves or through business intelligence platforms, what we should be able to do is work on segmentation predictively and often with clustering algorithms.

(10:18) For what purpose? For example, to have customer or potential customer segments grouped by churn probability. What is the probability that this group of people in my database will leave me or switch to a competitor? Predictive segments by customer lifetime value probability.

(10:42) Should I invest in these leads or in the others? What is the expected CLV for one group versus another? By bringing predictive artificial intelligence into our system, we can do this. But none of this guarantees that it is the best time to talk to these people. So what do we want, and what can we also bring into CRM? Intent signals solutions.

(11:05) I will try to explain this very quickly. For our business at Liminal, we are interested in understanding whenever someone in Portugal, or within our target market, is looking for solutions related to Salesforce, HubSpot, CRM consulting or automations. So what do we do? We bring into our CRM a solution that allows us to search the internet.

(11:29) I will show you here, looking for Salesforce consulting, Salesforce solutions. We add this term to the intent technologies library. What these technologies reveal to us is exactly that: all the companies that are online across several portals, several publishers and different sources.

(11:59) In a moment, I will show you another platform that also does this and is more transparent regarding the sources. But it shows us this. I know at this moment that someone in a company called Start Engine is searching online and researching the term CRM. But that alone does not interest me.

(12:18) What interests me? I am interested in people in my target market, and that is what I do. For example, all companies in Portugal looking for these terms, or all companies in the business-to-business segment. CRM equipped with artificial intelligence to complement the entire view we have of each customer and make our timing and our message relevant and useful for each person.

(12:43) We have not done it yet. This is where marketing automation comes in. When we talk about marketing automation, let us cover the basic concepts. We are talking about the ability to create and orchestrate an entire multichannel customer contact campaign in a journey builder, that is, on a screen like this.

(13:10) And it will be activated customer by customer, at the right moment. I will give you an example while still covering the basics. We will bring artificial intelligence into this shortly. The classic example is this: we have an e-commerce portal. Someone goes to our portal, adds a product to the shopping cart and ends up not completing the purchase. We have an abandoned cart scenario. What do we do with these automation platforms? By the way, many platforms are excellent at this, not only HubSpot or Salesforce or any other I have mentioned so far.

(13:37) We can do this. There is an abandoned cart, we wait one day. What do we do next? We interact with these people and remind them. You have an abandoned cart. Perhaps it was just forgotten, I do not know. Go back there. It is a very classic flow. But let us do something different. If they are current customers, perhaps it is better to send them an SMS.

(13:57) Before it sells out, go there now. But if we are talking about people who have never bought anything from us, a lead, we probably do not even have their phone number, only their email. We want to be less intrusive, so we use email. In the email message, we might even say: here is a discount voucher.

(14:14) Classic. We have all seen this work many times in our lives, right? We continue from there and then say: if this person did not buy, try a WhatsApp message, try reaching them with Facebook Ads, Google Ads and so on. If the person did buy, what can we do next? Send a satisfaction survey asking how happy they are with this purchase.

(14:36) Nothing extraordinary. This is just to align expectations about what automation means in this journey builder context. How can artificial intelligence support us here? Three examples: generative artificial intelligence, predictive intelligence and AI agents.

(15:11) Artificial intelligence. Example number one. Imagine the person answered this survey and is satisfied, loved the product. Fine, we do nothing. But if, on the contrary, they are dissatisfied and complained, we are interested in understanding the content of the complaint. We do not have the capacity to analyze each comment one by one.

(15:28) So what do we do? We call generative artificial intelligence into the process to help us interpret the comments in the satisfaction survey comment boxes and classify at least three use cases that matter to us.

(15:45) One is: I want a full return of my product. Another is: this was a shipping or delivery problem. And another case, it could be any other, is that the person probably made a mistake when giving the rating. So we simply ask artificial intelligence to analyze it and give us the result.

(16:04) We can already run a test with the specific contact. This prompt is classifying this contact's response. This is all in a dummy environment. And it says: this gentleman probably made a mistake and gave you a bad rating, but notice that his comment is: these guys are the best, even with a spelling mistake in English.

(16:29) Can you see it? So what can we do next? We have classified each of these negative responses and, for this specific one, we can call generative artificial intelligence again to generate the right email to interact with this person. From there, we continue the conversation. Predictive artificial intelligence, and now we are moving to another environment, in this case Bloomreach.

(16:57) A moment ago, we made an arbitrary decision. What is the best channel to interact with the customer? Is it SMS or email marketing? Well, what we should do is something else. We send SMS to everyone who has a higher probability of interacting with SMS. But if the probability of interacting with email is higher, then email is what we send, or a push notification, and so on.

(17:22) Also, not everyone will have a high purchase propensity. Some have a high purchase propensity if we send a message and even a discount voucher. Others would already buy with just a reminder or a smaller voucher. This is where predictive artificial intelligence can help us amplify automation capacity and then, of course, measure the results at each of the touchpoints.

(17:47) Can you see it? For example, in this first touchpoint, we sent a first email. How many purchases resulted from it? How much revenue resulted from this first touchpoint? Let us go to the last example. Generative artificial intelligence, predictive artificial intelligence, and now AI agents.

(18:07) When we talk about AI agents, we are talking about artificial intelligence that creates all this orchestration autonomously. That will still be the future. What we have today still has some limitations, but it is already very powerful. What I will show you is an agent from another platform we work with and implement for our clients, called Amplemarket. The agent is specialized in business-to-business prospecting and is called Duo.

(18:33) What does this agent do? Do you remember intent-based signals? It is automatically looking for all the people it finds online, particularly on LinkedIn, who fit the ideal customer profile for my business. More than that, it is looking for these people and for signals that show interest in my product or service category, whatever that may be.

(18:53) Signals such as comments on portals like G2, comments in Slack communities, profile visits, comments on competitors' channels and so on. From there, this agent will automatically, as you can see on the right, create the entire orchestration.

(19:16) It creates a first email that connects the signal with the reason why we are speaking to this person. We noticed this. And because we noticed this, perhaps what we have to offer is relevant. From there, it continues the conversation, seeking to book a meeting, for example. It visits that person's LinkedIn profile automatically, creates a connection request and creates a new email. Notice this.

(19:39) Hi Natalie, I noticed your review on G2 about HubSpot. The points about video in emails and dashboard flexibility really stood out. It sends voice messages. This is my voice, as you can tell, synthesized. I did not record this, and I did not even write this text. So it creates an entire multichannel orchestration without me having to do anything.

(20:05) Artificial intelligence working in typical if-then workflows. Three examples of how we can use generative intelligence, predictive intelligence and AI agents in this technology stack that every company should consider when thinking about technology to manage customer relationships.

(20:31) This is what we have been doing with many of our clients, such as CCB and GoContact, with considerable success. However, the challenge is always this. Do we want to dive into this world only to grow and become the biggest spammer in our neighborhood, or to genuinely build close, relevant relationships with our customers that ensure the future and relevance of our brand? Very well, thank you very much.

(20:31) Thank you very much, Tiago. Fantastic presentation and very enlightening on the impact of AI on campaign automation, from segmentation through to results analysis. [Music]