This article was originally published by Jensen Loke at https://www.linkedin.com/pulse/product-management-festival-reflections-jensen-loke/.
Nikila Srinivasan (Facebook)
Rahjiv Bhuta (Walmart Labs)
Silvia Thom (Zalora).
- Worse Beta of All Time by Helen Xue (Atlassian)
- ML from Product Perspective by Inga Chen (Spotify)
- Product Management & Organizational Transform by Maish Nichani(Pebble Road)
- Scaling The Product Manager by Sharon Anne Keen (Seek)
- Overcoming Cognitive Biases in Product Management by Srinivas Somayajula (Mail Chimp)
- Blurry Edges Make Better Products by Randy Hunt (Grab)
- Working Backwards: Amazon’s Culture of Innovation by Daria Tarawneh(AWS)
- Measuring the Right Thing: A Framework for Nailing Product Metrics by Isha Mehta (Atlassian)
1. Worse Beta of All Time by Helen Xue (Atlassian)
Helen’s story about launching Atlassian’s Identity Management Tool was a stark reminder that we always need to stay Customer Obsessed. As PM’s we always have to balance and make compromises where appropriate guided by shared principles. In summary, her sharing was about how her team had forced launched their product disguised as a beta but resulted in customers paying for a feature that they perceived to be free. And all this was done without positioning clearly the value the new product brought to IT admins and users alike.
Platform product management vs product product management are inherently very different things. Platform product managers need to think of the various product PMs that sit on top of their platforms as their own customers, and both PM groups cannot lead the build of feature sets or enhancements in their own silos. We have to consider the whole picture from the perspective of customers, as they feel the impact of our delivery end to end.
This is a very real problem we face at work. Are our firm employees overwhelmed by what we are delivering? We also can’t make too big shifts at once as users will have problems adjusting. Do the products integrate well with each other and add value to our firm employees? For that, senior leaders have to work with PMs together to craft the vision so that the various technical projects can be aligned to a shared goal.
2. ML from Product Perspective by Inga Chen (Spotify)
The best talk for me at the event on the PM’s approach to building a recommendation feature! But as a short disclaimer, I could be biased. Personally, building recommendations are a personal fascination of mine as getting it right is very challenging. It is the embodiment of getting the right data, deep-tech and context to deliver a meaningful solution to the problems customers face.
Inga started by sharing that personalization is the solution, but what is the problem we are trying to solve? Recommendations solve for information overload/decision paralysis. The Many Choices Problem plagues a large variety of different fields where data is rich.
For example, different consumer fields like Spotify has 40M songs, media with infinite content, social with graphs and networks, influencers to followers, e-commerce with various products, healthcare with doctors and prescriptions, fintech with stocks, funds
A cool approach is stay customer obsessed by always asking check questions. For example: “does it matter when people choose different things?”, “What inputs matter?”, “Do the differences between users matter?”
Perceived Personalization vs Under The Hood
I experienced the power of simple models when I worked with the Amazon Sponsored Products Ads Team. Invent and simplify, so they say. The simplest algorithms are explainable, serving out to customers at scale and extensible for their various purposes. All these are valuable lessons when it comes to building our own recommendation engines in the future.
As product managers, we also need to always learn and be curious, The way we embed this principle into our product is to account for echo chambers where data points become self enforcing in ML products. This can be done by injecting randomness into the model and I believe there is still a core benefit of having a product managed by a human vs machine.
What Can We Do Today?
- What’s your data wish list?
- Use the data you have — Tagging, Proxies.
- Bootstrap — be scrappy, manual tagging, consider 3rd party data, start collecting today: use to validate further statement
- Ask users! Give them something in return…
The checklist above shared by Inga triggered thoughts of an article I read before by Simon O’Reagan’s Designing Data Products. I highly encourage this read as a great framework to think about and ponder over how to solve the ‘many choices’ problem.
3. Product Management and Organizational Transform by Maish Nichani (Pebbleroad)
Maish discussed how an important factor that guided the success of digital transformation projects is alignment of technology projects by senior leaders. This allowed for coherent technology projects that made sense to drive visible value delivery to the business. We have to first be self aware of which state we are in today as an organization, and come up with strategies to move towards the north east corners of that chart to become true digital masters.
As PMs, we need to drive Business Transformation as part of our launches. More often than not, the people are the hardest part of the equation. In our context, changing employees are not enough, and changing the product owners themselves will also prove to be a challenging one and we cannot miscalculate the arduous effort for internal product marketing aka ‘change management’.
For this to be successful, product managers play a key role in
- driving this from the top down — convincing senior leaders and product/system owners that this is something that is worth doing so that service quality improves and we can build scalable products.
- driving this from the bottom up — identify early adopters, build self help capabilities, reduce user friction, and own customer success journey maps from a platform or product perspective.
Multi–sided markets bring together two or more interdependent groups who need each other in some way. We can find a striking resemblance to how marketplace product management mechanisms can work akin to how we implement our transformation operating model.
As PMs, we have to consistently balance stake holder needs, align on terminologies and definitions of what it means to drive change, and inculcate new ways of working. It takes a very coordinated effort and willingness to change but effective product managers can orchestrate this. PMs need to have a good grasp of how the organization, individuals, technology, and business, works in a harmonious ecosystem and leverage that.
A Thank You To the Speakers
Lastly, a thank you to all the speakers. I learnt a great deal from all of you and it was humbling for me to be in the presence of so many great PMs, especially at SEA so far away from the US. Conferences like these reinforce my earlier decision to come home and make an impact here.
All of your speeches were awesome, and I took way more notes than what I am sharing now. I’ll leave some suspense to the folks who want to find out more to approach me directly for it. I look forward to joining next year’s conference and be more involved in the community! 🙂
About the Author
Jensen Loke is a seasoned technical product manager and former Amazon senior product manager. He has previous experience from financial services, mobile payments, advertising and eCommerce. He is currently an Associate Director, Digital Technology at Temasek, Singapore. Jensen and his team attended the Product Management Festival APAC 2019 on the 17-18 June in Singapore.