Bar-Raising Product Management: Navigating the product life cycle with Aaron Bonko
Welcome to our Bar-Raising Product Management series. In this collection of articles, we dive deep into the intricacies of technical product management to share insights, best practices, and real-world experiences that help Technical Product Managers (called ‘Product Management-Tech’ or PMTs at Amazon) elevate their skills, understand how we work at Amazon Ads, and learn more about the work that excites us.
For this first installment, we hear from Aaron Bonko, Senior Principal Technical Program Manager (TPM) on the Measurement and Data Science team. With almost nine years of experience at Amazon, Aaron has a role that places him at the intersection of technology and business strategy. Here he discusses how we typically manage the product life cycle at Amazon.
The product life cycle at Amazon Ads
Our product life cycle at Amazon Ads isn't drastically different from other tech companies, but over the years, we've refined our approach to align with Amazon's culture of ownership and customer obsession. We typically break it down into five stages:
- Pre-development: This is where it all begins. At Amazon, we always talk about “working backwards,” and we start by deeply understanding our customers and their needs. The Amazonian process is to contextualize the product within our broader portfolio and develop a data-driven PR/FAQ (press release/frequently asked questions) that imagines the product on launch date and the benefits it will give customers. This vision statement is the basis for all work on the product; we don’t start building until everyone is happy with the PR/FAQ.
- Development and testing: Here we start bringing ideas to life—preparing designs, building prototypes, and conducting thorough testing. Unlike a traditional waterfall approach, we don't wait for a complete, exhaustive requirements document. Instead, we start with a high-level overview and engage our technical teams early. This allows us to get valuable feedback on complexity and potential simplifications right from the start.
At Amazon, we always work to create a minimum lovable product versus a minimum viable product. “Customer Obsession” is our first Leadership Principle, and even the simplest version of a product we release needs to delight. Testing is integrated throughout our development process, not just at the end, and we employ a combination of unit, integration, and user acceptance testing. One practice that has served us well is involving actual customers in our testing phase whenever possible. We often run beta programs in which select customers can try out new features and provide feedback. This real-world testing proves invaluable for identifying usability issues and refining our products before full launch. We also pay close attention to non-functional requirements during this phase. Performance, scalability, and security are not afterthoughts but integral parts of our development and testing process. We conduct load testing to ensure that products can handle the expected volume of traffic and security audits to ensure that our customers' data is protected, which is a priority for us. - Launch: Execution is key during this phase. When we implement our go-to-market plan, it’s important to think through what’s going to influence our release approach and how that may impact our technical scope. During the launch phase, we also work closely with marketing and sales teams and monitor launch metrics rigorously—often in real time—to ensure that we're meeting our goals and to quickly identify any issues that need to be addressed. One key lesson I've learned over the years is the importance of flexibility during this phase. No matter how well you plan, there will always be surprises. The ability to pivot quickly based on real-world data and feedback is crucial.
- Post-launch: The work doesn't stop after launch. We continue monitoring long-term success metrics and look out for “Day 2” signals that might indicate that we're losing sight of what truly matters with our products. These signals could include teams becoming overly focused on hitting specific metrics or goals at the expense of what truly matters for customers or the business. Another example is when a team overemphasizes creating processes and mechanisms, leading to creativity and agility being stifled. While mechanisms are important and an essential part of how we operate at Amazon, they should enable rather than hinder progress.
- Iteration and evolution: In this final stage, we use post-launch data to validate our original hypothesis or to assess if we need to pivot and adjust our vision. For example, maybe the customer demand for our product isn’t what we thought, or they don’t understand its value proposition. We incorporate these learnings into future versions of the product as well as other products we’re creating in our portfolio. We connect with customers and analyze usage data to get feedback on the product and to ensure that we’re addressing their concerns and product gaps as we plan future enhancements
Learnings on the product life cycle
As we've refined our approach to these five stages over the years, we’ve gained some valuable insights about the broader product life cycle:
- The life cycle is rarely linear: While we present these stages sequentially, in reality they often overlap and iterate. We might be in post-launch for one feature while simultaneously in pre-development for the next.
- Customer needs evolve rapidly: What was cutting-edge last year might be table stakes today. We've learned to build flexibility into our long-term visions to accommodate shifting market dynamics.
- Data is crucial, but context is king: While we're data-driven, we've learned that numbers alone don't tell the whole story. Understanding the context behind the data is often what leads to breakthrough innovations.
- Avoiding Day 2 thinking starts on Day 1: We've found that planning for long-term success and potential pitfalls from the very beginning leads to more sustainable products.
- Speed matters but not at the expense of quality: While we aim to move fast, we've learned the hard way that rushing to market with an unpolished product can damage customer trust.
- The most impactful ideas often come from unexpected places: Some of our best product improvements have originated from customer support tickets or comments in user testing sessions.
These insights have not only shaped how Amazon approaches each stage of the product life cycle but have also reinforced the importance of cross-functional collaboration throughout the process. This collaborative spirit isn't just a nice-to-have; it's fundamental to how we operate and innovate. At Amazon Ads, we don't adhere to a formal RACI (responsible, accountable, consulted, and informed) model. Instead we have guidelines that promote ownership and accountability:
- Product managers own the "what."
- Engineers own the "how."
- TPMs own the "when."
However, these aren't rigid rules. Our culture encourages everyone to contribute beyond their primary responsibilities, fostering innovation and rapid problem-solving.
Evolution continues
Building products at Amazon Ads is an ever-evolving process as we continue to learn from our customers and their expectations. We're constantly pushing ourselves to innovate and, in the process, to deliver experiences that delight.
In the next Bar-raising Product Management article, we’ll hear from one of our PMT directors on influencing without authority.