Designing an AI-driven data mobility solution from the ground up for seamless management and cost efficiency.

Context

Objective based management and data mobility across enterprise storage.

Problem

Lack of customer trust in the product leading to no deal conversion.

After launching the first version, Primary Data struggled to convert beta users into paying customers. Lack of trust emerged as a key issue during early customer conversations. While automation and data mobility were appealing, the shift in mental model was significant. Users found it unsettling that they couldn't track file movements, and the system provided little insight into why or how files were relocated, making it feel like "dark magic." Customers frequently questioned why specific files moved, and even the sales team lacked clear answers. Our goal was to build user trust to drive adoption in real-world environments.

Approach

Draw parallels from human-human trust to build the trust on product.

We began by examining the established model of human trust to identify parallels with our product, aiming to define guiding principles that would foster user trust in the digital experience. These principles would shape features, workflows, and cues while aligning engineering and product teams around trust. We categorized the key factors into four areas: control, visibility & validation, and troubleshooting. Also, just like human trust, building user trust takes time and repeated interactions. 

With this in mind, my approach was to focus on the following areas:

Build trust in the automation

Define a trust journey

Offer control, visibility, and troubleshooting capabilities for data placement.

After initial customer visits, the design team worked with technical sales to observe issues. Analyzing these results led to a week-long design spike with product managers to explore trust-building designs.

The first release gave users control over their data stores, changing their interactions. To maintain familiarity and credibility, we gradually shifted their mental model while keeping them in control. We initially highlighted potential savings to demonstrate mobility's value. Users could select non-production applications to move, retaining control over important data.

We combined visibility and validation, ensuring users understood the system's decisions. They needed to know how many files moved and if it affected bandwidth. Concerns arose from moving important files, so we allowed users to review those files. Validation focused on critical files.

Mistakes are acceptable if users can easily correct them. In policy-based systems, troubleshooting can be complex due to interlinked policies. Minor changes can create unwanted effects, making troubleshooting decentralized. In the first release, we provided issue detection, while remediation required manual adjustments.

Gradually introduce automation as users become more comfortable with it.

Building trust, much like in human relationships, requires time and consistent interactions. Our strategy focused on keeping users within their comfort zones while gradually guiding them toward full automation. This meant avoiding abrupt changes to their established workflows and providing clear motivation for users to transition to the next stage.

An important challenge was ensuring that both product management and the founding members understood that customers would perceive a value proposition at each phase of the transition. This understanding was crucial for maintaining customer engagement and driving sales. To address this, we planned to introduce features progressively. However, users would have the option to enable these features sooner, as their confidence in the system grew. By doing so, we aimed to create a seamless experience that fosters trust while empowering users to embrace new capabilities at their own pace.

Outcomes