From ‘Viable’ to ‘Awesome’: The Thinking Behind eBay’s Product Development
By: Claire Xiao, Product Manager for Ads, & Parin Jogani, Software Engineer for Consumer Selling
Hear from two technologists on how we’re evolving our approach to building new products.
Each product we build is sort of like starting a small business. It starts with an idea and a hypothesis of consumer need or desire. Then, we keep building upon this, testing and learning along the way.
At eBay, our technology teams work in an Agile model where we focus on building what’s called a minimum viable product (MVP) to test ideas for new products and features. MVPs are essentially prototypes—or the most basic versions of a product that allows us to collect consumer learnings before launching at scale. During this stage, the main focus is on functionality, knowing the experience can be optimized in the future as long as the MVP helps us validate consumer demand.
In the product development process, this can lead to less-than-ideal (and sometimes, outright confusing) user experiences. Through an MVP approach, what might actually be the result of a poor user experience—for example, where users cannot discover or navigate through—might be interpreted as a lack of customer need for the feature or product we are testing.
Given this possible outcome, we’ve recently shifted our approach to building products.
Shifting from 'Viable' to 'Awesome'
We don’t just want to build a product that is viable at minimum—we want to build a product that is awesome. When it comes to building products in the MVP model, we seldom talk about creating minimum awesome experiences, but this is precisely what we need to do in order to truly create the best experiences for our customers.
An awesome experience can mean different things at various stages of the product life cycle. For a mature product, it means a fluent experience and robust features that foster great user engagement and retention. For a fast growing product that strives to penetrate deeper into the user base, it means a series of moments that delight customers and create viral effects. For an MVP, however, it means a product that only allows us to precisely test product market fit.
Let’s consider an example. If you were opening a restaurant with an exotic menu, you might want to test the waters and start a food truck as an MVP for a few months before deciding to invest further in the restaurant. But a food truck with fantastic food in the wrong location won’t get many customers, while a food truck in a great location with a confusing menu won’t pull through many orders. MVPs can fail sometimes simply because they are confusing, not because they aren’t valuable.
Putting the New Model to Work: Promoted Listings Lite
Our team employed the MAP approach with Promoted Listings, an advertising product for business sellers that allows them to boost the visibility of their items with premium placements on eBay and to help enhance their sales velocity.
To really push ourselves in how we approached product development, we asked: By optimizing for the user experience, could we quickly validate user demand for a new promoted listings feature? This led us to Promoted Listings Lite.
We had a hypothesis that the full Promoted Listings offering had too many options and was too robust for the average consumer seller. Contrary to business sellers who want comprehensive levers to fully control their advertising strategy and grow their business, consumer sellers are less interested in advertising strategy and more focused on getting things sold without hassle. So, we chose to build an MAP that was as simple as possible for our consumer sellers: Promoted Listings Lite.
How We Built an MAP
1) MAP = MVP (functionality) + user experience
While building our MAP, we made sure to not only focus on functionality but also the user experience.
The original Promoted Listings experience for our small business users has full functionality, including manual bidding and campaign creation for advertising, which suits these users’ preferences for more control. Through early user research, however, we found consumer sellers sit on the other extreme of the spectrum: they need a simple, clear and hassle-free experience.
To satisfy these needs, we simplified the experience via a one-click campaign creation. To deliver this “advertising easy button,” we recommend a bid percentage and removed the concept of a campaign setting for this segment. In this “lite” version of Promoted Listings, users only need one click to complete the journey. The simplicity of the solution drove wide adoption.
2) Stay relentlessly focused
Great companies define their mission and principles to give employees a direction. When it comes to developing a new product, we also aimed to clearly state the questions we sought answers to. This guided our team to optimize for the core functionality and experience while making tradeoffs wisely. As a result, we were able to move faster without sacrificing quality.
The key to success for our Promoted Listings pilot was having clear objectives and a relentless focus.
To give you one example, there were a lot of debates around adding opt-out and reporting features to our MAP. It was really tempting to slide to the extreme—perfectionism and trying to solve all problems in the pilot. However, in a test-and-learn environment, perfect is the enemy of good. We reminded ourselves that our goal for building the MAP was to validate whether consumer sellers wanted to use a product like this. This enabled us to quickly re-focus our energy on solving only for the critical questions we were seeking answers to.
3) Form a cross-functional tiger team from the beginning
MAPs benefit from cross-functional expertise, so it’s a good practice to form a tiger team comprised of engineers, designers, user researchers, marketing experts and data analysts from the beginning of your project planning, and sync on regular basis.
Each expert will help your MAP progress in different ways:
Engineering partners will quicken product discovery since they can help design solutions that balance agility and delivery time. Engineering partners can also help nail down lean and modular solutions while keeping long-term scalability in mind.
User researchers and marketers can help build the value proposition for the feature and carry them over to go-to-market activities.
- A member from the legal team can also contribute to the successful messaging of the product positioning to end customers.
Working with this tiger team from the beginning not only helped us launch the new product, but also champion it, making it easier to scale in the future.
Shifting our product development approach allowed us to build Promoted Listings Lite quickly and created new opportunities for us to further improve this product line. Promoted Listings Lite is currently available in the U.S., U.K., Germany and Australia.