Many industries are affected right now with digitalization. In the media industry, the big change started already years ago. Transforming traditional newspapers or TV channels to successful digital services is not easy. Sanoma Media Finland is a media company who has succeeded in this transformation, with growing digital newspaper subscription base and attractive multi-channel advertising offering to marketers. One of the key success factors has been the ability to utilize data.
Iiris Lahti, former Development Director – Data Utilization at Sanoma Media Finland, currently Founding partner of AI Roots, is coming to the Data Innovation Summit 2020 to share this story. As someone who has experience with driving the data-driven transformation in a media company, Iiris can impart her knowledge with how to utilize data, analytics and machine learning in designing digital media services, content, marketing and better digital customer experiences.
Hyperight: Hello, Iiris, we are happy to welcome you as one of the speakers at the 5th edition of the Data Innovation Summit 2020. To start off, please tell us a bit about yourself.
Iiris Lahti: I am a data enthusiast with the background of driving transformations in various industries as a consultant and in leading positions. Most recently, I have worked in the media industry, supporting the transformation of Sanoma Media Finland with my team of analysts and data scientists. I am specialised in driving the utilisation of data, analytics and design thinking in digital product development, consumer sales, marketing and content creation.
I am also passionate about helping professionals find their strengths and roles where they can excel and grow. I have just recently jumped into a new phase in my career as a founder of AI Roots, a network of data freelancers. Our mission is to help companies create sustainable business value from data.
Hyperight: Your presentation will cover how to utilise data and machine learning in designing digital media services, content, marketing and better customer experiences. What are the lessons you’ve learnt using data and ML in digital media services?
Iiris Lahti: In traditional media, such as newspapers, there is the need to break bubbles and provide a holistic and independent view to the readers about what is happening in the world. Therefore, applying advanced ML/AI solutions to, for example, content recommendations or personalisation is not a very easy question. There is the challenge of finding a balance between ethics and ambition: How far is the media ready to go in exploring the ML/AI solutions to be able to achieve their business targets, but at the same time, being transparent to their readers and true to their ethical guidelines?
In other media, such as video streaming services or audio, utilisation of ML/AI is more common and nowadays even expected by the users. Netflix and Spotify have raised awareness of the service features enabled by ML/AI. Some customers might be so fond of their movie recommendations or personal playlists, that they don’t want their spouse or children to “mess up with their algorithm”. The same also applies to digital marketing, where people tend to get annoyed if remarketing campaigns are too late or the targeting of the advertising is not right.
During the years in media, I have learned not to underestimate the power or the customer. They ultimately decide how and when to use the content, what services they want to pay for. Looking back at historical data is not enough in industries such as media that are changing so fast and have so much uncertainty. Traditional CRM or digital analytics is good, but especially in digital marketing, content or product design, they are not enough to be able to understand what the customers really want, or would want in the future. Design thinking methods help to understand customer motives more deeply. Customer insight is needed for segmentation, brand research or analysing customer satisfaction. In the best case, of course, these methods would be used together.
Hyperight: As 2020 is the year in which the Data Innovation Summit turns 5, could you point out what have been the essential breakthroughs with data and advanced analytics in the last five years according to you, especially in the digital media industry?
Iiris Lahti: In the media industry, digitalisation and data-driven transformation have gone hand-in-hand. When I started in Sanoma, it was still unclear how the company or the whole industry will go through the transformation, whether it would come from entirely new digital products or transforming the existing ones. Sanoma has been very successful, particularly in re-inventing and transforming existing strong brands to the new digital era. The trends in consumer preferences have also been positive. The customers, even the younger generation, are more willing to pay for high-quality digital content than before. Customers want the media to provide them independent, reliable and entertaining content.
Also, the technology landscape has changed during the past years in digital marketing, such as cloud-based data lakes, marketing automation solutions, data management platforms (DMP), digital advertising ecosystems, search engine algorithms or browser privacy settings. The changes will continue also in the future in these areas, forcing the industry to be continuously innovating new solutions and being ready to change course when needed.
Hyperight: One of the challenges that stand in the way of data-driven transformation is poor data literacy. What are some other challenges that could inhibit the potential data unlocks in an organisation, and how can they be handled?
Iiris Lahti: Data governance models and catalogues are definitely hot topics right now. Companies want to understand what kind of data they have, what restrictions there are in gathering, integrating and using it, and what kind of potential it could provide for future growth. Also, many companies have learned that it is not enough that the IT department knows the data or, in the worst case, the capability has been outsourced to an external partner with limited visibility in the business. The classical saying “data has no value unless it is used” still applies. Companies, who have been able to generate sustainable value from the data assets, have been able to transform their entire business towards data-driven, including things such as strategy, organisation structure, roles, processes and practices. Also, in this kind of organisations, the roles get mixed. Suddenly being an analyst is almost everybody’s job and the data scientist becomes an equal member in every team. Companies are looking forward to democratising their data and breaking the data silos. However, data democratisation does not necessarily mean aiming for full self-service analytics, but making the basic data available and shifting the focus towards advanced analytics solutions and ML/AI. It might also mean reducing the data, finding the most relevant data for the users or learning how to create better insights and more concrete actions from the data.
Hyperight: Some sources warn of the risk of “over-datafication” in digital marketing which may result in excessive dependence on big data and a tendency of sacrificing creativity and originality. Is there a real risk of “over-datafication” in digital media services?
Iiris Lahti: Without an innovative, creative and unique marketing or content strategy and deep understanding of the market trends and customer motives, there is a risk the company will get lost in between marketing automation technologies, data-based targeting opportunities and reaching short-term tactical targets, such as the conversion rate. In tactical sales, being obsessive about data-driven marketing is not a bad idea, if it means that the customers will receive relevant offers at the right time, and at the same time the marketing spend will be more effective. Modern marketing automation tools enable testing tens of variations of a campaign instead of the traditional human-centric A/B-testing. But still, everything starts with understanding the basics, setting clear targets and coming up with that unique idea of a new brand slogan, visual look or a campaign that will be something original and new. It is not choosing between using data and being creative. It is more about helping those creative minds, designers and content producers work together with the data enthusiasts, and finding the balance from both sides. I am personally a big fan of multi-skilled teamwork. Succeeding in a digitalised world does not require unicorns who can do it all, but a team that has a variety of skills and can effectively work together.
Hyperight: And lastly, what are your predictions for data utilisation and data-driven transformation in the next several years?
Iiris Lahti: My prediction is that the data, analytics and even ML/AI competence moves closer to the everyday business operations, instead of managing it as a traditional IT function. The legislation and ethical principles of gathering and using the data will continue to be an important topic in the EU and hopefully, it will also differentiate us from other big markets, such as China or USA. As many industries have started to invest more and more in their data capabilities, the lack of talent in the market will grow in the future. At the same time, the data industry will be affected by increasing the level of automation and emerging technologies. AI and ML can also be used for managing data pipelines and quality. Cloud-based environments make algorithms more accessible even to companies who don’t yet have their own data science team. These trends will change the role of a data scientist in many ways.