ezCater / Fall 2019 – Spring 2020

Closing the mobile gap

Looking at our metrics across the order experience for top of the funnel anonymous visitors, my team at ezCater noticed a gap between mobile and desktop conversions. To tackle this we set out to run specific experiments tied to quantitative and qualitative data for direct results. My role was Lead Growth Designer and I worked on a team with a Product Manager, Engineering Manager, Design Researcher, Customer Acquisition Manager, plus several Software Engineers, and SEM Specialists (internal at ezCater and at Google).

Investigate

We identified mixed levels of intent in our quantitative metrics— were mobile visitors more likely to simply explore menus without intending to convert? From a human-centered perspective we were curious how to get to know these anonymous visitors so we could serve up more relevant search results and confront the paradox of choice dilemma. We hypothesized that conversions will increase if we collect more event information upfront because of a better downstream UX for the visitor.
Moderated user study revelation
The majority of potential customers are willing to answer at least one question (especially headcount, budget, food type, dietary restrictions) as long as their results reflect those preferences.

Ideate

Next I led a group of the team through a one-day Design Sprint Workshop at Google in NYC to tackle these pain points and open up our thinking. We decided to focus on facilitating conversions for high-intent mobile visitors.

Design

After our collaborative brainstorm I took our ideas and created a mockup to present in under 2 hours. 🤯 Our mockup was a blue-sky version with big ideas across growth initiatives and best UX practices:
  • Progress bar for setting expectations
  • Ability to add a full catering order into cart in 2 clicks (original: 6+)
  • Potential to land a user deeper by enabling add-to-cart in 1 click
  • Thumb accessible with all CTAs above the fold and a headcount slider
  • Search results showing priority of urgent timing and 5 star options

Test

Following the sprint we trimmed down our design so we could build and test on our live site in an A/B experiment guiding visitors from our How It Works page into a flow with a new questionnaire. Visitors would submit a delivery address, event date, and time, then on an interstitial page answer questions about their food preferences and event details, then see personalized search results.

​Our results showed a dip in order start rate but improvement across checkout start, order complete, and new customer rates.
Analytics
​Multiple experiments across our landing pages yielded similar results: all showing diverging impacts between click-thru/start rates and conversion rates. Since those initial rates fell, we believe that the visitors who weren’t going to convert anyway dropped, making the visitors who entered the store more likely to engage.

Adding a whole new page with 6 questions to the funnel surprisingly improved conversion, even while on mobile.

Build

Through our iterative process, we landed on a design proven to lift conversions built into our landing experience. The positive downstream impact is likely because of an improved experience with more relevant search results.
Growth benefits
✅ Visitors in the store flow down the funnel are now more likely to engage & convert
✅ Less likely for catering partners to reject an order with more info upfront
✅ Reduced burden of choice by highlighting options that fit visitors' needs
✅ Showing dietary accommodations, cuisine variety, and "last minute" language conveys that we can handle those cases
☠️ ​Higher cognitive load on visitor upfront

​What's next? Now that we’ve confirmed anonymous visitors are willing to answer questions about their event AND it improves their order experience, we can work to personalize to reduce their cognitive load.
Next case study: Building trust →
Abby Luke / 2021 / abby@abbyluke.com