KPIs for digital marketing: how to measure ROI
Measuring your campaign performance correctly doesn't depend on the amount of data you collect, but on your ability to focus on the KPIs that truly move...

Sure, I am Xavi Puig, a graduate in Advertising and Public Relations from Pompeu Fabra. Above all, I identify as a deeply curious person — I am literally interested in everything! That has led me to work across different sectors throughout my professional career: branding consultancy, innovation consultancy, technology companies… I currently work as Head of Insights at Telefónica, where I am responsible for identifying innovation opportunities for the company. My work involves staying highly informed: closely observing users to understand their needs, analysing market trends, exploring new technologies, etc. All of this with the aim of developing new ideas that add value to the company.
It is a combination of science and art. I am constantly consulting specialist media, interacting with other companies in the sector and keeping up with the latest trends. It is a systematic process, but it also requires a lot of intuition to identify seeds that are still in very early stages and know when to pull on a thread. They are not usually obvious, at least not at first — especially amid all the noise in the innovation space, where value competes with public relations.
Yes, we have a laboratory in Madrid called the Human Experience Lab, where we conduct research with users to better understand their needs and behaviours. We carry out all kinds of activities there: from in-depth interviews to testing prototypes of products we have not yet launched to market. We adapt research methods based on the objectives and the phase of research we are in. Internally, we tend to divide research into 3 major types: exploratory, generative or evaluative. Although, as you well know, in practice these types of activities always end up somewhat mixed.
I am the Head of Insights within the Video Innovation area, which is the area responsible for working on the innovation of our entertainment platform. It is a global product we develop at Telefónica (the holding company) and then deploy across the operators we have in different countries, adapted to their brand and geographical context. In Spain and Hispanic America our operator is Movistar, in Europe we use O2 and in Brazil the brand is Vivo. As I say, we work on a global product, which means many of our innovations are piloted or launched in other parts of the world outside Spain.
"If you only look at what your competitors are doing, the most you can aspire to is being an imperfect copy of the best practices of the moment."
Exactly, our goal is to improve the consumption experience at every level. And that involves, among other things, looking beyond the boundaries of the sector. If you only look at what your competitors are doing, the most you can aspire to is being an imperfect copy of the best practices of the moment. That is why, personally, I believe the value lies in connecting less obvious dots from other places.
In class I always use the Tesla example. Tesla's true innovation was conceiving cars as if they were not cars. In their case, drawing inspiration from the technology industry and reframing the car as a technological device. Suddenly the car is hardware running on software that makes it operational. This subtle reframing of the product changes absolutely everything. All of a sudden, the brand can fix faults in a model without having to go back through the factory, vehicles can improve over time (the opposite of what happens in the conventional industry) or even allows for standardised production processes (all vehicles are the same and the radical difference is in the software). With this approach of connecting dots from two different industries, Tesla has managed to position itself as a benchmark in the automotive industry. Now it is the legacy brands that are looking at how Tesla does it. Had they only looked at industry best practices, they would only have managed to be a poor copy of Ford and company. Hence the importance of looking outward.
The other thing we do a lot is observe people very closely. Both at the micro level (behaviours, habits, tastes…) and the macro level (social changes). The way people build relationships and meaning has a direct impact on consumer products. Following the evolution of people closely is vital.
"Tesla's true innovation was conceiving cars as if they were not cars. (...) Hence the importance of looking outward."
I do not think there is a theoretical limit to innovation. It is true that there may be limits in terms of specific technologies, as in the case of a wheel, for example. However, a product is rarely built on a single technology. Rather, a product tends to be an assembly of several, which means that theoretical limit we were talking about would already be much harder to reach due to the combinatorial variable. And that is only talking about "hardware", which is saying a lot.
Innovation can manifest in many other forms beyond the artefact. For example, through the business model, the customer relationship, work processes, etc. On the other hand, innovation is strictly linked to creativity, which by definition is inexhaustible. There will always be opportunities to innovate.
Both activities are fundamental, and to a certain extent, inseparable. Quantitative analysis gives us WHATs. For example, how many users connect to the platform, how long they stay active, where they drop off… This type of analysis has been enormously amplified by the rise of digital products, which offer the possibility of monitoring literally any parameter in real time.
However, when we ask ourselves about the reason behind that data (why the user drops off on that screen, for example), the quantitative approach gives us no answer. This is where qualitative analysis is needed — the activity that provides us with WHYs. Through methods such as in-depth interviews or ethnography, we try to capture subjectivities that allow us to understand the underlying reasons behind a given reality. In summary, both types of research are indispensable to obtain a complete picture and make informed decisions.
Artificial intelligence is a complex yet fascinating topic. Not only at a technological level, but also ethically, socially, politically… In fact, history can largely be told through the technological innovations that appeared at each moment. That is why the emergence of new technologies (especially those with global impact) conditions absolutely everything.
In the case of generative artificial intelligence, there is an additional nuance. For the first time in history, we see machines entering a territory that has historically been exclusively human: creativity. We have always seen machines as brute force, but never as creative entities. Hence the threat.
Personally, it is a topic that touches me very closely through my work. Today, it is hard not to think it will change everything. However, as people often say: "the soldiers in a war are the least equipped to explain what is happening in that war because they are too immersed in it". I think that is what is happening to us right now with generative artificial intelligence: we are so deep in it that it is difficult to predict what the long-term impact will be.
"Technological changes are like a train: no matter how much you want to stop them, it is impossible because the momentum they carry is too great."
It is difficult to define it with a single word. If I had to choose just one, I would say I am optimistic, but with my own connotations. Let me explain: I accept that technological progress is unstoppable, and optimism is my proactive way of approaching that reality.
In a certain sense, technological changes are like a train: no matter how much you want to stop them, it is impossible because the momentum they carry is too great. If you try, you are most likely to get run over. That is why it is better to work from a place of acceptance. This is even more imperative in my case, working at a technology company where a large part of my company's success depends on leveraging the latest available technologies.
That said, this does not prevent one from simultaneously maintaining a critical and reflective mindset. In fact, I believe these two things necessarily go hand in hand, both at an individual and collective level.
"We are immersed in an era of accelerated change, with numerous companies competing to lead the technological race."
Of course, the train analogy refers to the inevitability of technological progress. Imagine you are standing on the tracks trying to stop a train that is approaching. No matter how hard you try, the train will continue on its course, and if you do not get out of the way, it will run you over. My idea of optimism lies precisely in accepting this fact and working from there. For example, by asking what we can do from inside that train.
We are immersed in an era of accelerated change, with numerous companies competing to lead the technological race. It is a challenging situation that we cannot stop, but we can adapt and find better ways to navigate this changing environment.
"Tools like ChatGPT can be useful for speeding up certain tasks, but I believe the cognitive process must remain in the hands of people."
From that critical optimism I was telling you about. Accepting and leveraging the changes that come, while at the same time analysing (and not losing sight of) what we are losing along the way.
A few weeks ago, I was teaching qualitative research classes at Universitat Pompeu Fabra. When the time came to talk about preparing materials before fieldwork, I told them they could lean on ChatGPT without a problem. They looked at me strangely. Why did I say that? Because the technology is already there at their disposal, and whether you tell them or not, they are going to use it. You cannot put doors on an open field.
However, not all uses of ChatGPT are equally valid. In my case, I suggested they use it as raw processing power, and that they themselves remained in charge of the cognitive part. That means, among other things: writing a detailed prompt, reading the result the machine generates in detail, questioning all the aspects that could be improved, asking the machine to make as many changes as deemed necessary…
In this way, the machine has worked for you, but the cognitive process is still carried out by you, as a human. This nuance is essential for me. Because in reality, part of the purpose of doing that work is not so much that the interview guide gets formalised, but that during the process of elaborating that guide, the person is preparing themselves to conduct a good interview.
In the way I was suggesting they use ChatGPT, they were still preparing the interview themselves, even if the machine writes the text. Now imagine how their interview would have gone if they had shown up with the first script that ChatGPT had spat out. Badly.
In summary, tools like ChatGPT can be useful for speeding up certain tasks, but I believe the cognitive process must remain in the hands of people. Among other things because, ultimately, the measuring stick we use is human judgement, and that, today, remains irreplaceable.
Measuring your campaign performance correctly doesn't depend on the amount of data you collect, but on your ability to focus on the KPIs that truly move...
Discover how to apply data-driven marketing decisions to optimise campaigns, reduce errors and improve results compared to intuition.
Every company has data. Simply by existing in a digital environment, information accumulates and digital analytics becomes a reality (whether you want...