Auto Overclocking

CPU Overclocking helps to increase the performance of your PC by changing specific settings through your BIOS. 10 years ago, users had to manually go into their BIOS to make these changes for better performance. This was a tedious process as there were over a hundred different settings a user could change, so Intel created an interface that helped users make these changes without having to go into the BIOS.

To be competitive in the gaming industry, our stakeholders (Product Managers and Leadership) wanted to build an app foundation with feature parity to our competitors' products before differentiating ourselves from their offerings. Most of our competitors already had an Overclocking feature where users could manually tweak specific settings to increase PC performance. With our stakeholders' direction, we sought to craft a similar but more intuitive interface for our users.

After building that foundation, our stakeholders wanted to differentiate HP from our competitors and so they came up with a feature called Automatic Overclocking, which would be an evolution of what any of our competitors offered. Using a database of common, optimized settings for different models of PC’s, our competitors choose settings they think will work best for your PC, but even within the same model there could be huge discrepancies from one PC to another. To differentiate ourselves, our stakeholders wanted to create an algorithm that would run different settings on your PC and find the best settings for your specific configuration-without the user ever having to tweak any settings themselves.

Though our stakeholders thought this was a great feature to add into our application, we wanted to make sure our users found this feature useful as well.

We conducted some User Research and got a lot of great feedback on Automatic Overclocking. While there was some hesitation with Automatic Overclocking, most users thought that this could be very useful.

Fig2. Automatic Overclocking Quotes

Currently, the app includes a manual Overclocking tab: a simplified interface where the user can make changes to Overclocking characteristics instead of having to make these changes within the BIOS. Looking at the current structure of the application, it made sense to nest this new feature within the same tab.

Fig3. Information Architecture structure

Using our participants' gaming history and the knowledge we’ve obtained from two years of research and usability testing, we categorized our users into a list of seven different personas.

For Automatic Overclocking, we went with two different personas: Rich Son Steve and Social Angie. Rich Son Steve typically is not cash-conscious, but he is particularly conscious towards the brands he supports. He likes bragging to his friends about his high-end components, making this feature well-suited for him.

Social Angie is more of a casual gamer and typically gets her gaming PC as a hand-me-down from either her brother or boyfriend. Because her system will be old, she will have a need to overclock her PC to get better performance.

Fig4. Automatic Overclocking Personas

Based off of our Personas, and our past research, we created a Customer Journey Map to better understand what some of our user’s needs, goals, and thinking are. With this method, we are able to dive deep into our users’ problems and identify opportunities to develop features that will resonate with them.

Fig5. Customer Journey Whiteboarding

Fig6. Customer Journey Map

Using the Customer Journey Map, we went into creating Storyboards to figure out what the common scenarios were for users while using this feature.

Fig7. Automatic Overclocking storyboard

While we were able to think through high-level scenarios with great detail, it was difficult to get a full picture of the situation without input from the Backend Development team. We worked in parallel with them, and there were certain parts of the design that would be affected by technical constraints.

At this time, the Development team was working on creating the algorithm that would be used to find the optimal overclocking settings. To do this, our development team created a script to collect data on our different machines. The data they were collecting would give us insight into which parameters made the biggest difference to a user’s Benchmark score.

One particularly interesting insight was that the process to optimize your system settings could take anywhere from thirty minutes to two hours. Without clarity on how long it’d take to overclock their system, it was difficult to create an experience as most users wouldn’t be willing to wait for up to two hours. Given this, we started the Task Analysis process assuming the worst case scenario of two hours to get your overclocking settings.

Fig8. Initial task flow

With the first iteration of our Task Flow, we started to go into more detail and started mapping out the experience for each screen. Since there’s pretty high level of information density for users to consume, we wanted to give users enough information upfront before starting the overclocking process. The tutorial experience is going to be very important here, so designing this experience was one of our biggest priorities.

Fig9. Initial Wireframes

With our initial wireframes done, we wanted to test the design early and make sure that it was an intuitive flow for the users. Overclocking is a very hardware dependent feature, and since we were still early in our design process, we knew that when we conducted testing, some of the insights would have to be taken with a grain of salt.

We ended up doing 2 sets of testing with an Invision prototype. In the first set, we had 6 participants and in the second set, we had 10 participants. After the first set, we made some big changes to the layout based off of users’ feedback.

Fig10. Second Interation of wireframes

After the second set of testing, we were extremely happy with the changes we had made and for the most part, the flow was very intuitive to users. After giving the Product Manager the research data, we were told to try to simplify the experience even more. Stakeholders were expecting just a 1-click button, and while research proved that we needed more information than just one button, management pushed us to simplify the experience even more. Taking their feedback into account, we came up with the following wireframes.

Fig11. Final iteration of wireframes