Cookie Jar

Product management best practices

September 26, 2016
By Young Blom

As a product owner in a fast-paced environment where the only constant is change, I’m always thinking about how to compress time to deliver features. The primary objective is to build customer usable product that actualizes key benefits in the in short, medium and long term.

Here is how I approached a product launch that allowed me to ship quickly without jeopardizing user experience as well as delight my customers. The primary challenge I faced was how to balance developing a clean user interface with a complicated back-end system moving at different development velocities.

Our goal was to allow the demand partners to use the nToggle self-service platform to create rules to shape exchange traffic. The system was intuitive when managing fewer than twenty rules. However, we knew that manual toggles were a temporary solution until we introduced our machine learning platform AutoToggle. I had to architect a product plan that built out a user interface that could provide immediate user benefits that I could phase AutoToggle in as it matured.

The first phase was to allow the customer to build and set their own rules and filters. This gave the customer the control of their traffic and understanding of the product’s power. Even in the first phase, I needed to make tradeoffs from everything I spec’d out. I had to decide whether to deliver in three months or a month earlier with the most critical features needed to allow customers to build and set their rules and filters.

This meant defining which features were absolutely needed in my minimal viable product (MVP) and how those that weren’t included could affect the user experience. A rule builder was required but could function without all attributes and forecasting capability. Ability to allocate percentage of traffic, basic metrics and publishing was critical.

Knowing my MVP helped me to construct the UI in a way that the features left out could be seamlessly integrated later without a major rework of the UI. I always had to think ahead – so I could deliver framework that allows me to break out of the UI with minimal redesign work, enabling me to easily plug in new components later on.

The second phase was to build out a semi-automated AutoToggle. I could have waited for full automation, but would have lost out on improving customer benefits and new business opportunities. There were a few steps that required manual changes behind the scenes to enable configuration changes to put a rule in auto mode. Since some manual process was involved, the platform was guided by an account manager to enable AutoToggle until it was fully automated. Even though manual steps were required, it allowed us to exercise the product, get feedback, iterate and have immediate positive customer impact.

The third phase will allow the customer to enable/disable AutoToggle and trigger all the back end components to happen automatically. Since I planned ahead in the second phase, I built the manual / auto labels as disabled toggles. The third phase simply requires the team to activate the existing toggles to allow the customer to switch to manual or auto with minimal UI work and focus on plugging in all the new components to automate AutoToggle in the back end.

When building any product, your grand idea will require multiple phases based on various facets being delivered at different time frames. You must anticipate and plan for multiple phases and define your MVP for each phase to avoid broken experiences. At the end of each phase, I collected feedback and identified any development challenges to understand if there were any business or market changes before I defined the next MVP phase.

Allowing yourself to break down your design and development into smaller MVP phases helps you to focus and gives the team a sense of accomplishment. It also keeps the momentum going for sales and telling your story in the marketplace. Most of all, features are released faster and you can gather quicker feedback which is crucial because you might need to course correct.

The positive outcome of using this method expedited AutoToggle to enter in the marketplace and help multiple customers optimize their QPS efficiency between 60% to 90%. As well as for some it showed revenue uplift. Most of all, it led to Demand partners adding more exchanges and acquiring new customers.

Next time you’re planning a major rollout, stop and think about whether you really need this feature right now or is it a nice to have? Does it hinder the customer from completing a task or conducting their business? Are the customers screaming for a feature? How can I build for now but minimize reworking the product later? Does it tie back to your goals?


Tags: , ,