Situation:
Unless you’ve been living under a rock for the past 10-20-30 years, you know that the world is transitioning to digital. Specific to media houses this means a gradual decline in printed media. With this transformation comes new opportunities but as with any new opportunity, there is also a pinch of risk involved.
In the past (if we go far enough back) people were relying on printed media to get their news. Back then it was a whole other level of loyalty game that had to be played. The regime has changed massively as we entered the age of digital and now the attention economy.
With a total of 17 newspapers and 4 websites,our client is one of the bigger Media houses in Denmark. As with most, if not all, Media Houses they’ve had their eyes on personalization, aka. individualization, for many years.
Complication:
In the attention economy, every company is involved in the combat zone. A combat zone where you win or lose customers if you are not perceived as being relevant enough.
So far, the solution in the industry was to base the recommendations on either “Most Read” or “Newest Content” with the expiration date set by each journalist.
If you think, logically, it is a good way to solve the challenge. You will probably reach the most people with the right content, however, it is with a lot of uncertainty and the expiration date will not show a united front.
As an example, imagine an article about an animal escaping a zoo, say a green chimpanzee or pink kangaroo (again; imagine an article). That article might be relevant for say, 1-2 weeks. If you on the other hand have a tax reform article, that content might be relevant for 6-12 months even though tax reform is generally seen as slightly more boring than a pink kangaroo.
When each journalist has to estimate the expiration date, the results will vary depending on their attachment to the specific subject. This means that there is a lot of change management involved to keep the “dead by” date, up to date.
Opportunity
The Media House was facing these 2 conundrums and was considering their options. Should they hire a team of Data Scientists and wait 1-2 years with no promise of reaching the end goal?
That sounded like a risky and expensive approach – not without reason, other known players in this field have gone with this approach, and ended up having spent large amounts of resources for a disappointing result.
Without giving the full number (because it depends on the compute power needed) that number covers more than 20 years of software subscription (not even considering the running maintenance costs).
The Media House needed a solution if they were to keep and grow market share long-term. Since you are reading this use case, you know that they didn’t go for the in-house-built solution.
If you are in the media space you know that this is by no means an easy task to solve. There are so many variables with so much different content coming out every single day.
Luckily, we built Allyy.io for recommendations on deep content, exactly like news articles, and although it took some extra time as this was the first time working with a Media House, once we got it done, it could easily be adopted by any other Media House.
Once we got all the integrations and models set up, our client decided to test the approach, by feeding our recommendations into a “recommended for you” carousel for a test group. We would then go ahead and compare the performance to a regular group with the old setup (Most read, Newest content).
Resolution:
And to be honest, the numbers fully speak for themselves and are quite astounding, so I will keep this section short.
By individualizing every person in the test group compared to the regular group, our client reached an amazing:
- +19% sign-ups to paying subscription (article behind paywall)
- +24% renewal of subscription
- +20% quicker revisit
- +21% more clicks
I think we can agree that for an already established business such as a Media House, growing signups and renewal at this level after only 3 months (including development and integrations) is quite amazing.
Ping us if you are curious about how many years of Allyy.io subscription you can get for the equivalent of having your own team of data scientists.
