When I joined Gartner in 2019 one of the products I was brought on to support was a pay-per-click lead generation auction system for their Digital Markets unit. Prior to my joining the decision was made to update the auction functionality. Gartner had hired a consulting firm to do generative UX research to understand user needs. Users of this product can have hundreds of auctions based on keyword categories, several different site placements, and finally their geographic target markets. For larger companies, this could mean potentially hundreds of auctions to manage.

The legacy auction system tried to solve this complexity by having all auction bids be a percentage adjustment of one set bid price. While this attempted to make the adjustment of bids easier some of the larger companies with higher spends wanted more granular control of actual costs.

Problem Statement

Higher spend users want granular control of bid prices so that they can run a more robust marketing program where they can regularly test price adjustments within a set budget.

UX Research and Testing

Although the high spend users had this need there was uncertainty about how this would impact lower spend users. Those users were proportionally a much larger majority of our users. The lower spend level might suggest fewer resources to monitor a more complex auction system. We ran UX research interviews with several different user segments to better understand how more granular controls would impact different user groups. We captured some additional user needs including the ability to apply bulk updates to potentially alleviate the extra effort that granular controls might require.

At this point, we identified several user pain points that needed additional testing:

• Granular bid controls would require additional effort and resources for most users. A subset of those users had not identified a need for granular controls.
• To help users make highly informed decisions at each auction point we would need to provide them with much more data than they previously had access to. This could create a new pain point if it leads to analysis paralysis for users who don’t have the experience or resources to manage this data analysis.

We went through several rounds of usability testing on concepts like bulk bid adjustments, the ability to work outside of our system with third-party tools and upload bid adjustments via CSV, and user flow adjustments. Through UX research we were able to demonstrate that product hypotheses the team had made, such as bulk bidding features, would not solve the actual user needs and ultimately would be a wasteful feature to build.

Final Design and Release

At the culmination of UX research my team held several stakeholder presentations to share research analysis and secure buy-in on several product recommendations. The first release of the product update was completed about eight months after my in-house team took over the project. While we risked negatively impacting the major revenue generating product we saw no negative impacts after launch. The product today generates millions per month for the business unit. Post-launch research with users saw an increase in NPS scores, from 6.0 to 7.4, that we could attribute directly to our auction product enhancements.

As a final of this iteration of the auction system, I presented to corporate stakeholders as part of our annual Usability Day speaker series. It was important to demonstrate the value of the UX design and research process as well as how UX can show measurable success in support of our business unit’s OKRs.