Designing for growth
A case study for Fast, one click checkout. Constructing a design process to achieve specific growth metrics. I completed a four step process to target a problem, identify leading metrics to change, design an experiment, and prototype a solution.
Optimizing the one click checkout experience 🚀
My role: research, user testing, optimization, behavioral insight, prototyping
Software used: Figma, Photoshop
Fast is the world's fastest checkout (pun intended). I selected Fast for this study because I believe Fast is solving the most frustrating problem user experience on the internet: long checkout processes and forms.
A special thank you to Pantelis Korovilas, head of design at Fast, who was kind enough to shed a light on one of the growth opportunities and UX challenges that Fast has encountered in their data.
Step 1: Create an objective
After identifying Fast as the brand I am designing growth for, I began researching KPI's and leading indicators in the checkout process. I determined the leading indicator I wanted to develop a hypothesis and experiment for was customer acquisition and checkout conversion.
Step 2: Develop a hypothesis
I begun researching to lay a foundation for a successful hypothesis. I conducted user interviews and walk throughs of the Fast checkout experience with first-time and returning users.
My research goals:
+ Gain insight into a customers feelings and hesitations when experiencing the Fast checkout journey.
+ Feel empathy for first-time users who are unfamiliar with a one-click checkout.
+ Discover core checkout functions bring the most familiarity and trust to users.
Here's what I learned:
+ Users are trained to see several screens during a checkout process, with the final screen being a confirmation page. Users felt confused that buying something in one-click could really be "that easy".
+ Returning users (who have a Fast account and access the full benefits of Fast checkout) lacked confidence that their checkout process was complete and successful after using the one-click checkout.
+ Users enjoyed the simplicity of the checkout process and the "aha" moment when they realized they really could complete a purchase in one-click.
To create a great hypothesis I knew I needed to: define the problem, define the solution (how will I solve this problem and what rationale I will use), and target specific results. Using this formula, along with insights from my research, I developed the following hypothesis:
Step 3: Design an experiment
Growth Mindset: "Embrace the uncertainty, messiness and challenges of design thinking. Approach experimentation with a goal of learning, not immediate success. Don’t lost faith in your ability to influence change."
To design a successful experiment I created a comparison, with a control group, to put before users. I designed two additional checkout screens and an additional product page based on my hypothesis. Quality insights rely on quality data, I tried to remove any bias and test with a diverse audience.
I gave users instructions for the experiment outlining the task and context:
Task: Choose which checkout screen gives you confidence in your purchase and trust in the checkout process.
Context: You're buying an item using a one-click checkout, with no routine time to review your cart or see a confirmation page.
Below are the three visuals given to users, image B is the control.
I wanted to test several components of the confirmation screen:
Does adding a lock icon leading to a checkmark icon help users feel more secure and confident in their purchase?
Does "return to store" instead of "keep shopping" make a difference to the user?
Which items are most important to keep within the prominent green box for convenience and understanding?
Second image presented to users. Image 1 is the control. I wanted to test whether adding "view your order" to the order complete button would increase a users trust toward Fast and the checkout process.
The design experiment confirmed image B (the control and current Fast screen) was most preferred by users. I wasn't discouraged by these results as I believe Fast has provided a clean and optimized user experience to the checkout process. My experiment was still a success because it gave me valuable feedback from users as to the reason behind their choices.
The user feedback provided me with the insight that users like being told what's next but prefer the simplicity and easiness screen B provides.
Back to the (Figma) drawing board.
With the insights from the first experiment, I conducted a second experiment comparing two checkout screen prototypes with users. Screen A was the control.
Screen A: the control group. Screen B: Tested the addition of the "what's next" dropdown, containing helpful information to first-time users, and "view your order" inlay to the order confirmed button.
Results from the second test
With the results from my second experiment I moved to the next step of creating a working prototype.
Step 4: Prototype a solution
The final step in my process was to prototype the solution and put it before users. I prototyped the full Fast store in Figma. (The Fast store prototype (based off the real Fast store), included a skydiving product theme as an ode to the Fast-est ceo/co-founder.)
Want to play around with the prototype yourself? Click here.
The prototype received valuable feedback showing an increase in user trust with Fast and willingness to purchase using Fast again. I found the addition of the what's next dropdown especially helped people in an older age demographic, who had limited online shopping experience, understand the benefits of Fast and know what was expected of them as a customer.
Here's what one user said after walking through the prototype:
"I appreciated being given more information on what to expect after the first purchase click."
Ideal Final Step
If I worked at Fast my next steps would be to conduct A/B testing to track the growth metrics I identified with my proposed solution.
However, since I don't have that capability I had to replicate an experiment and find a way to substitute tracking growth metrics.
I decided to conduct a final user test with my current Fast store prototype.
Final testing to confirm increase of trust
In this final user test I observed eight users purchasing a product, four of these users used a control Fast store with their current checkout system and four users purchased using my updated Fast store prototype.
Task provided to users: Purchase a Fast skydiving helmet. Once a purchase is complete, rate your trust in your purchase and a successful checkout experience.
I provided all users the following scale to rate their trust.
My findings showed results of increased trust in the success of the checkout process. Based on my hypothesis, I believe my prototyped solution would lead to an increase of checkout conversion and active buyers across online retailers that use Fast checkout.
What I learned through the designing for growth process
The greatest challenge I faced in this project was trying to design for growth with a product that I believe is already optimized and focused on the user. I have great respect for the product team at Fast and the clean, easy checkout experience they crafted.
I learned to trust the design process and the outcomes from testing. Even if testing outcomes don't support my work or hypothesis, it still provides valuable feedback. Having a growth mindset means not losing faith in my ability to influence change, and that's what I am taking away from this project.
Thank you for reading!