Personalization used to be a novelty. Now is rapidly becoming the norm in consumer marketing today. We simply expect marketing emails to use our first names, for Amazon to show us recommended products based on purchase history, and for Netflix to suggest similar shows after we finish our latest binge-watch.
Arguably, personalization is most impactful in the music industry. There’s a great deal of complexity involved. Music is incredibly personal. For example, identical twins growing up together might have drastically different tastes in music.
This presents both a challenge and an opportunity for companies in the music industry. If they get their personalization efforts right, the payoffs can be huge.
We had a chance to chat with Ahmed Kandil, SoundCloud’s Discovery Product Manager. Ahmed is revealing how he applies the power of personalization to every project he works on and the benefits from it.
Ahmed Kandil Career Development
Ahmed is a self-confessed data enthusiast and experimentation evangelist. After graduating from Cairo University, he worked at Hewlett Packard Enterprise as a Quality Assurance Engineer. Next, he moved over to Vodafone as their Customer Insights Specialist where he becomes their Competitive Intelligence Team Leader.
A stint as Regional Planner at Philip Morris International follows before Ahmed moves to home24, a furniture startup going through an initial public offering (IPO). After a brief term as Growth Product Manager for N26, a mobile bank, Ahmed moves into his current position in June 2020 as the Discovery Product Manager for SoundCloud.
Can you start by telling us a bit more about your background?
Currently, I work on a lot of things but my focus is on computer science. Information technology has always been my primary niche. I consider myself a product manager with a business-heavy background focused on leveraging data and analytics.
I’ve always been very attracted to the customer side of things. I didn’t just want to be in the background, but to understand how to ultimately help the end-user. That’s basically why I left my first job at Hewlett Packard.
As fascinating as it was, I wanted to focus more on the consumer. I ended up joining Vodafone on their consumer insights team. This involved gathering lots of data, gleaning insights from it and connecting this to the business and marketing teams.
That’s when I began to love my work. I could understand what the consumer is going through, identify what they want, and relay that information to the wider team.
But I quickly recognized there was still something missing. Despite providing the impetus for consumer-facing projects, I was mostly in the background. I wanted to be part of the implementation team and use these data-driven insights to improve the end product.
That’s why I joined home24, a furniture company based out of Berlin. They were going through an IPO and everything was pretty hectic. It was basically excitement and chaos in equal measure. As product manager of MarTech, my role was to make their marketing efforts as efficient as possible. It was great fun.
It sounds like quite a large project. Where did you even begin?
The first step was to track our customers in as much detail as possible. We wanted to understand their behavior beyond simply looking at what they clicked on. Once we started to get familiar with how they behaved, we acted on these insights by personalizing the experience for every individual consumer.
Once we realized what the data was telling us and how this should inform our personalization efforts, we started to put these insights into action. It took a while to get right, but the more you finetune, the better results you will get.
How did you manage to improve home24’s marketing?
I was put in charge of the discovery with a task to help consumers find our products more easily.
This mainly fell into 3 categories: search, recommendations, and ranking.
We moved from fact-finder to elastic search, customizing and localizing it for several different countries. This was an incredibly technical, data-science-heavy process. You need to understand the queries people make and the correlation to conversion and retention.
This also involves multiple iterations of testing to get it right. We started by using our searches and queries as basic inputs for the models before tweaking and testing them.
We wanted to upsell and cross-sell by saying to each consumer: “Here are products that are relevant to you and you might be interested in.”
These insights were based on their personal history and preferences. We wanted to make our recommendations hyper-targeted so when they pop up as banners will immediately pique the customer’s interest.
We aspired to rank items and results based on two things: relevance and experience. Imagine you’re searching for a white sofa. Yes, we need our white sofas to rank highly, but if we see a customer looking at our cream sofas, then we include these too.
Does your background in personalization come in handy now that you’re at SoundCloud?
It really does. I’ve learned to truly believe in the power of personalization. More importantly, I’ve learned how to get results quickly.
People sometimes believe that personalization is a long-term game and it takes a while to produce tangible results. This isn’t true. There are plenty of things you can do quickly and scale at a pace.
I focus on the listener experience to help users have great music discovery moments easily. We do this through personalized recommendations based on what we know about them and their listening habits. We balance this with serving new and hot emerging artists.
It requires a lot of personalization efforts. We’re trying to match users who like this style of music with emerging creators who don’t have a fanbase yet. When we do this successfully, it’s like magic. Creators quickly start to gain an incredible amount of momentum in no time at all.
So it’s essentially two sides of the same coin. You’re helping both listeners and musicians at the same time?
Exactly. We don’t have a business without listeners or musicians. They’re both equally and important to us. For creators, we provide an all-inclusive dashboard that helps them filter out the noise and begin to truly understand their audience.
If we give them relevant data on what people are listening to, which tracks people play on repeat, and at what point listeners drop off, they can start honing their craft based on what their fanbase responds positively to.
It’s almost like reverse-engineering music. Some musicians might prefer to just do what they enjoy and hope they will eventually attract a fanbase. That’s totally fine. Musicians are ultimately allowed to create whatever they want to create.
As an alternative, with the data available, they can take stock of what their audience likes and put these insights into action by fine-tuning their music accordingly.
Personalization is the way forward, no matter the company or the industry. We found that a personalized landing page generated 32% more conversions than a non-personalized landing page.Ahmed Kandil, Soundcloud’s Discovery Product Manager