From: myfirstmillionpod
A/B testing, a method of comparing two versions of a webpage or app to see which one performs better, is a common practice for internet companies to optimize user experience and conversion rates. This approach can also be applied to restaurant menus, particularly with the rise of digital menus [00:01:57](00:01:57).
The Opportunity
The shift during the pandemic to online menus and QR codes presents a significant opportunity for restaurants to implement A/B testing [00:01:28](00:01:28) [00:01:38](00:01:38). Many internet marketers observe that restaurants are not currently A/B testing their menus [00:00:08](00:00:08) [00:01:42](00:01:42). This gap creates a demand for specialized A/B testing software that restaurants can easily integrate [00:01:49](00:01:49).
Traditional Menu Design
Even before digital menus, optimizing physical menus was a known strategy to increase revenue. Michelle Banesh, featured in a Hustle story, was a professional who redesigned physical menus based on human psychology and eye-tracking patterns (like the “Zed” or “Z effect”) [00:02:15](00:02:15) [00:02:32](00:02:32). Her work demonstrated that simply by redesigning a menu, a restaurant could increase its revenue by nine dollars more per customer, without changing any products, but rather the orientation and order of items [00:02:43](00:02:43) [00:02:50](00:02:50).
Insights from Behavioral Economics
A sushi restaurant startup, which initially operated as a delivery-only “cloud kitchen” more than 10 years ago, experimented with A/B testing their online menu [00:03:57](00:03:57) [00:04:08](00:04:08). They consulted Dan Ariely, author of Predictably Irrational, who advised them on pricing [00:04:33](00:04:33).
Ariely suggested that sushi prices were too low, despite the common perception that sushi is expensive [00:05:44](00:05:44). He argued that people associate higher prices with higher quality, similar to the wine industry [00:06:07](00:06:07). When they A/B tested a 50% higher price on one menu version, they observed a higher conversion rate, not just more net money [00:06:13](00:06:13) [00:06:17](00:06:17) [00:06:21](00:06:21). This significantly impacted their margins in an industry known for thin 10% margins [00:06:30](00:06:30) [00:06:35](00:06:35).
They also tested descriptive language, adding phrases like “Alaskan salmon hand caught” or “always fresh, never frozen” to item descriptions. These additions, though factually true, significantly increased purchases [00:06:41](00:06:41) [00:07:09](00:07:09).
Digital Menu A/B Testing
The digitization of menus via QR codes provides an “inflection point” for easier A/B testing [00:03:00](00:03:00) [00:08:00](00:08:00). Unlike physical menus where data collection is difficult, digital menus simplify the process [00:08:06](00:08:06).
When customers scan a QR code, they can be directed to different versions of a menu, similar to A/B testing a landing page [00:01:40](00:01:40) [00:02:20](00:22:20) [00:03:00](00:03:00). This allows restaurants to test various elements such as:
- Menu design and orientation [00:02:53](00:02:53) [00:03:18](00:03:18).
- Item order [00:02:53](00:02:53).
- Pricing strategies [00:06:11](00:06:11).
- Descriptive language for menu items [00:06:41](00:06:41).
- Visuals and photos [00:09:18](00:09:18).
Metrics that can be tracked include time spent on different menu versions, scrolling behavior, and ultimately, items sold and average customer value [00:08:39](00:08:39) [00:08:43](00:08:43) [00:03:33](00:03:33).
Business Opportunities
The need for A/B testing in restaurants opens up several business models:
- Software Development: Create user-friendly A/B testing software specifically for restaurants that can be easily integrated with existing Point-of-Sale (POS) systems, allowing for automated reporting and analytics [00:01:49](00:01:49) [00:03:12](00:03:12) [00:08:27](00:08:27).
- Consulting/Agency Model: Offer services to redesign and A/B test menus, charging a fee or a percentage of the increased revenue [00:07:18](00:07:18). This could involve providing two different physical menu versions and tracking sales to prove which design is more effective [00:07:25](00:07:25). A recurring monthly fee could be charged for continuous testing [00:07:51](00:07:51).
Given that the average restaurant owner may not know how to implement these strategies, a service that promises to increase conversions or average customer value by a specific percentage (e.g., “50% more conversions,” “average customer value by three dollars”) would be highly appealing [00:03:25](00:03:25) [00:03:30](00:03:30).
Extension to Delivery Platforms
The concept of menu optimization and A/B testing can also be extended to third-party delivery platforms like DoorDash and Uber Eats [00:09:14](00:09:14). These platforms often feature long menus, which can lead to paralysis by analysis for customers [00:09:42](00:09:42) [00:09:45](00:09:45).
A service could optimize restaurant listings on these platforms by:
- Improving photos and descriptions [00:09:18](00:09:18).
- Highlighting dietary information (e.g., vegan options) [00:10:40](00:10:40).
- Curating or trimming menu options to focus on popular items [00:11:07](00:11:07).
- Developing better copy and taking better photos [00:11:16](00:11:16).
One business model explored involved creating “shadow brands” or clones of existing local restaurants on delivery platforms (e.g., “Choo Choo Chinese” for a local Chinese restaurant). These cloned brands would have optimized menus, bright photos, and trimmed offerings, but orders would still be fulfilled by the original restaurant [00:10:49](00:10:49) [00:11:51](00:11:51). This approach could generate 30% extra revenue for the restaurants due to better marketing on these platforms [00:12:00](00:12:00).