Taxi Advertising Concept/Smart Display

Rooftop Advertisement on NYC Taxi

Getting to sleep some nights is hard for me. My brain will just not turn off. I'm constantly thinking and analysing the world around me asking myself, "can that be done better?" or "Why did they make that design choice? Was it motivated by business or user need?" Since moving to New York City in May of 2015 this relentless thought process went into overdrive.

Given that I'd' never lived in a large city before I seemed to look at everything with a fresh pair of eyes. One night while standing on my balcony I saw a taxi drive by. As others do it had a static print sign on the rooftop.

Normally you'd dismiss it or just absorb it as it went by but something caught my eyre. The ad was for an event that had occurred 2 weeks prior. I began to dwell on this fact. Many questions came to mind including: 

- Why didn't the driver change the ad?
- Who is in charge of orchestrating the thousands of ads on all the taxis? 
- What does it cost to advertise?
- Is it based on time? 
- Why don't they show more relevant information other than a phone number or date? 
- Can't the design be more attractive and aesthetically pleasing? 
- Is there an SDK like platform or style guide dictating the design and information architecture of the ad? 


Inspiration / Foundation 

Pebble Time Animation 

At the beginning of the year a california company called Pebble released a smartwatch. Given the trend of wearable technology you'd be mistaken if you thought this was just like all the other smartwatches. One of the key difference was that it used a color e-paper display. The images were "static" but still were able to execute simple but clever transitions between apps. 

When i began to think about the taxi advertisements you naturally think to use an LED display, but I quickly realized this wouldn't work. When i conducted research I counted that a person only has about 4 seconds to view, consume, and understand the advertisement as the vehicle drives by.Four seconds isn't long enough when you figure that the average commercial on tv is about 30 seconds. 

Cost was also an issue. Granted LED displays are coming down in price they are still expensive when you think about the size of the display on the roof is equivalent to a 60" TV. I had noticed there is a hardware model out there that's an LED display but i've only seen a few. 

Then I had the thought to expand the color e-paper display to a large screen. The power consumption should be low enough to be driven by the vehicle's battery or even in device batteries. But what about the ad itself? If I used this type of display I could show real-time data alongside advertisements. Or even multiple advertisements on a single display.


Targeted Advertising 

Whiteboard sketch of process

When i started to think about the ads themselves something else came to mind. Would a company want to advertise in manhattan as well as queens? Would that make sense for their brand? With the existing static printed ads the same ad was shown to millennials in brooklyn as the wealthy in meatpacking district. But their user personas and what motivates them to purchase items is very different. Wouldn't they want to tailor their ads to fit the demographics viewing them? How could I get the ad display to understand where it was in world and change the advertisement accordingly? The easiest cheapest way seemed to construct a companion smartphone app for android and iOS. This would be the workflow. 

  1. Advertisement is currently being displayed. 
  2. If the taxi changes locations, and using the driver's smartphone, understand the geolocation of the vehicle 
  3. Using predetermined demographic and geographic zones, transmit targeted ads to the driver's phone
  4. If the vehicles crosses a zone, the driver's smartphone, using BTLE, would connect to the display
  5. Once connected, the app would transmit an updated targeted ad to display


Leveraging pre mapped zones and targeted advertising smaller businesses who previously couldn't afford to advertise on this medium would now have the ability to gain impressions hyper locally. Because advertisements would display dynamically based on various categories including location, time, and cost.  Much in the same way that Facebook and Google cant chargeback to advertisers based on clickthroughs, smaller businesses could display an ad during target times (4-7 during dinner) and locations (4 block radius around business) it's cost can be dramatically reduced. Pricing tiers would breakdown accordingly:


- $500 and under
- $501 - $2,000
- $2.000 - $10,000
- $10,000+


- Morning
- Afternoon
- Evening
- Late Night


- City-wide
- Burough
- Neighborhood
- Block


- Single Block
- Multiple Blocks
- Entire Display



All advertisements will have to follow a somewhat strict SDK to guarantee continuity and uniformity. That said brands will still be able to use their key identifiers. Using the Pebble style guide as a base I've created a few samples to show how companies can better serve the consumer with necessary and nice to have information. 

Bronx Version, Select image to view larger

Brooklyn Version, Select image to view larger

Manhattan Ad, Select image to view larger

Blocks The display is arranged is interchangeable blocks. Advertisers can purchase by the block or the entire screen. Each block is its own element with its own animation and information. 

Ethnographic Information With the main advertisement for a basketball game, I've included

- weather forecast on event date
- Date/Time of the event
- Location of event
- Localized public transportation information
- Real-time ticket prices and availability

Vector-Based Using Pebble's vector based graphics it's very simple to read, clear typography, and straightforward. Making is vector based means its easy to create pixel perfect ads that while appear static are actually changing and updating using real time data. 



Hardware Design

The Two Existing Advertisement Display Mounts 

I've started to sketch out what the hardware/display might look like. I imagine there would be many iterations on this but I do have preset pillars for the requirements including: 

- Universality (Fit top 4 vehicle models)
- Non-Destructive Mount
- Low Power Consumption
- Resilient Hardware (weather proof)
- Stable Software (99.9% uptime)
- Light Weight
- Repairable
- Low Cost
- Aerodynamic