The future of customer acquisition for media companies
During the past few years, I’ve sold Frosmo products and services to a whole range of companies. From our national lottery and casino, Veikkaus, to big international retailers like Clas Ohlson. However, as far as ecommerce is concerned, their digital challenges are all pretty similar. The major exception to that is the media industry.
And that’s because media companies face a unique problem; they have to charge money for something that people expect to be free. As a result, these companies have been relying on technical innovations even more than others in a bid to change the business model and stay relevant. It really has become adapt or die.
To be part of that reinvention of an industry by providing the software and solutions has been a great ride so far, but really one that’s only just getting started.
The points I’ll discuss in this article:
- Competing with social media
- Editorial content
- Natural Language Processing
- Targeting vs. testing
Intro to Frosmo and Media
Before I go into the trends that we’re helping media companies with right now, I think it’s good to look back at what we’ve done previously. And at this point, it’s also good to recognize the effort put in by our own developers and project managers to crack a market that was on a losing streak. It’s what has turned our experts into experts and Frosmo into a go-to place for the media.
Some of Finland’s top digital newspapers and streaming services have been improved with Frosmo. They’ve used the Platform to add and improve content recommendations, make layout changes, gamify reading experiences, optimize websites for mobile, add personalized paywalls, and much, much more.
The Frosmo Platform makes it fast and easy for front-end developers to change websites without having to access the back end. This concept enables endless development flexibility. Combine this with data, market trends, and the ideas of your digital team, and together with the Frosmo Platform, you can turn any website into a superior browsing experience, personalized for every occasion.
Implementing solutions of the future
And that brings us to the solutions of the future. Web development is a continuous process and so is the quest for additional revenue and customer acquisition. An increasingly hot topic is microtransactions. You may know these from games and now online newspapers and magazines are looking to make use of them, too.
For media, microtransactions means that Frosmo sets up a sort of mini paywall for readers to consume just a single article or get unlimited access for a week or two. The setup is a win-win for the consumer and for the newspaper, as the consumer doesn’t have to buy an entire subscription just to read something they’re particularly interested in and at the same time, the person gets used to spending money to consume news.
While it’s too soon to share any figures just yet, I think that microtransactions are here to stay and perhaps even become their own source of revenue.
Using the Frosmo platform, it’s pretty easy to create well-converting mini paywalls by taking into consideration the reader’s preferences. Once these preferences are established by tracking behavior and identifying the categories that the reader is most interested in, relevant articles can be gated off and the existing paywall can be used guide people to the right purchase funnel.
Bringing the conversation back to the website
Another challenge that media companies are facing is that the conversation has shifted to social media. On one hand, it’s good for these companies that their content gets shared on social media, but it also leads to a decrease in sessions and number of articles read, which in return negatively impacts ad revenue. In general, people who click on an article on Facebook are still in the mindset of “browsing Facebook” rather than “reading the news”. As a result, they read one article and then go back to cat pictures.
In essence, you’re now competing with social media over someone’s time. And to win that battle, it’s important to offer more relevant and more engaging content. You really, really need to get your layout and recommendations right, as you only get one shot to grab the reader’s attention.
Frosmo has a few tricks up its sleeve to help with both the layout, which is the next point of discussion and by using machine learning methods such as Natural Language Processing, further discussed at point 4.
Layout changes, especially editorial
One of the main reasons why people are loyal to certain news sites and magazines is because the tone and standpoints match theirs. Opinion pieces play a huge role in showing an outlet’s identity and they are great conversation starters. The same can be said for thorough investigative pieces. And in both cases, people love sharing them.
However, shorter-than-ever concentration spans rarely go well with long-form articles since there are tons of distractions on a website. Normally, clicks are great because you want users to read and explore, but a lot of resources have gone into these often important topics.
To really get people engrossed in the story, and to make people visually show that people are getting their money’s worth, the article has to look premium. This means blocking out distractions and finding novel ways to add visuals.
While changing page layouts and rearranging content is the majority of the work done with the Frosmo Platform, it’s still not that often used for these one-off occasions. And that’s a missed opportunity that I think will change the coming year as companies are more and more willing to invest in differentiating factors to add value and at the same time people are more willing to pay for quality.
Improving recommendations through Natural Language Processing
Like pretty much every self-respecting tech company these days, we’re investing heavily in machine learning. We’re all very excited about a model that we’re building that’s using Natural Language Processing or NLP in short.
NLP matches content to content with greater precision for example by analyzing the words in an article and assigning a similarity score. Displaying articles with higher similarity scores then leads to more accurate and higher converting recommendations. These recommendations can then be refined further by excluding already read articles. And better recommendations translate into more clicks and conversions.
One thing that we’re experimenting with is the way we can transparently show the machine learning outcomes. Showing the generated tags alongside each article in the Frosmo Platform would be one way to avoid black box problems. At the same time, we’re looking to make the outcomes manipulable so that anyone in the ecommerce team can make tweaks, rather than just data analysts or scientists.
And this is extremely important because understanding the data and outcomes within the ecommerce team is hugely beneficial for business development. It enables data-driven development and helps teams understand what should be changed or tested in order to reach the intended business goals.
This leads to another topic which I’ll cover in my next blog post: attribution modeling. My experience says that those companies which are doing well have C-level management that cares about the attribution modeling of your AI. The core of business development is to manage your customer journeys and this can go horribly wrong if business goals and usage of data are not talking to each other.
Separating targeting and testing
The last trend that I want to highlight is more of a workflow or methodology change than a technical or usability improvement. And that is the need to stop developing everything at once. Specifically, to separate targeting from testing, at least in the setup stage.
User journeys are so complex that even the slightest change can have an (adverse) effect on test results. As such, it’s important -perhaps vital?- to figure out your initial targeting criteria and make layout changes accordingly prior to running tests. Think of targeting as a long term effort and testing as an additional short term way to improve targeting results.
A great way to improve results for certain targeted audiences is multi-armed bandit (MAB) optimization, a form of “testing” where an algorithm decides which variation gets shown to whom. While it can be used to find winning variations in a short amount of time or on pages with less traffic, where it really shines is when you let MAB experiments run continuously.
For example, if you have a carousel with news recommendations, MAB could be used to rearrange the order of which article gets shown when based on conversions. Then, when a new article is published, it can automatically replace the worst performing content with this fresh piece. It guarantees the most amount of clicks, indefinitely.
Innovation never ends
Media companies have started figuring out what works, but that comes with its own risks. It seems that most media companies have now found themselves after the initial shock of the digitalization wave has worn off. But there are still many challenges and opportunities like the few seen above.
Staying fresh and relevant through continuous development and innovation is key to the consumers’ hearts. Never forget to innovate, the faster and easier the better. And fast and easy, isn’t this what we all are looking for, regardless of whether we’re developing or consuming?
Let’s chat about your media sites. You can find me on LinkedIn.
About the writer:
Timo Vuori is a sales leader with eight solid years of experience in solution sales and business development in ecommerce and SaaS. He has created dozens of digital solutions for international clients in sectors including retail, travel, online services, banking, gaming, and telecommunications. Working in a groundbreaking industry with cutting edge technology, he keeps himself always up to date on the latest technologies and recent development in AI, voice search functionalities, and the evolution of internet buyer journeys.
Results-focused, Timo’s passion is to create lasting relationships with his customers and find a “win-win” in any situation. He believes that the route to achieving competitive advantage is through an understanding of global markets and industry trends and creating a winning alignment of business strategy and technology.