Platform team

Powering the #1 road trip app

Roadtrippers is the ultimate road trip companion. It will help you route from A to B and discover what you love in between.

Since 2012 it has helped plan more than 38 million trips, covering more than 42 billion miles.

The app has a smart route planner that helps you find the most interesting places to explore within a set distance from the route to your destination.

It also has editorial travel guides and itineraries to follow, turn-by-turn navigation, and shared planning with your co-conspirators.

The complexities of place data

Powering Roadtrippers is a collection of 7M+ places of interest.

Each pin on the map has hundreds of possible data points whose veracity needs to be continually evaluated, and which needs to be stored and retrieved without issue.

But there is no single source of truth about places and their data. In fact, we relied on many:

• What Roadtrippers users said
• What partner location data sources said
• What open source map data said
• What the public internet said

Each pin on the map can have many “places” from many "sources" from which the data is sourced. And each of those will have varying degrees of completeness, accuracy, and timeliness.

How do you do that?

When I joined, there was a complex data ingestion pipeline, and the engineers working on it were being carved out into their own platform team.

There was also only a rudimentary internal interface for manually editing place data. We needed a new app to support this new growing team.

The goals for the app were three-fold:

1. Create a more robust internal editor for anyone in the organization to be able to audit, suggest, or correct information about a place
2. Have it support our growing machine learning efforts for making determinations about place data
3. Preserve optionality that we might commercialize the platform team, it’s data, and this app, in the future

Editor

A place data editor for curation and AI training

I designed a new web app for internal correction and training of our place data.

For the sake of speed, I designed it exclusively for desktop, and we built it with React and Rails, previous engineering choices.

The root view was a Yelp-like list and map combination with a prominent search and a series of filter options.

Each listing was broken into a dozen or so buckets: identity, contact information, location, hours, etc.

I decided to have edit buttons on each individual section. It was counterintuitive, but with some quick user testing I found folks were able to move faster through their workflow overall by limiting the scope of each individual saved update.

That way, you could be 100% confident nothing else could be changed but the section you opened up.

The app needed to support an exhaustive list of attributes about a place (with yes/unknown/no options).

It also needed an interface to support the existing, super detailed categorization.

And since there were multiple types of roles within the organization with different levels of authority over the place data, we needed to represent that in the app.

Some roles didn’t make direct edits, but instead were able to submit suggestions for other roles to approve or ignore.

Boone

The travel data brand that almost was

As Roadtrippers sales folks went out to sell partnership opportunities, they kept hearing interest from prospects in our place data and consumer interfaces for it.

The company grew interested in commercializing the place data platform—for tourism organizations, consumer brands, and even local businesses.

Our thought was it could generate revenue in a way which wouldn’t cannibalize Roadtrippers, could amortize our costs associated with data engineering and machine learning, and would create opportunities to help travelers across emerging interfaces—car touch screens, voice-based devices, etc.

We knew our value proposition centered around the question:

Imagine a future where your customers and audience look to you first to answer their travel questions.

What would that future look like for you?

So we needed a new brand, go-to-market messaging, and a claiming process for local businesses to work within our editor to manage their listings (think Yelp or Yext).

That was something I took on.

Name

We needed a name for the new brand. It had to be something which could stand alone and not require the context of Roadtrippers or our new parent company. But it also needed to fit thematically into the brand portfolio. It also had to look right sitting next to Roadtrippers.

After I ran multiple rounds of the naming process with key folks in the company, we were arriving at an answer. Here’s what I wrote in Notion at the time presenting the choice to the organization:

A great many aspects of our business are the same as they were when we started, but some important things are now more clear.

We are a place data service in 2018. We explore, we learn, we store, we share. We're modern day pioneers in the digital world.

Many brands today relying on artificial intelligence or automation use familiar, personal names: Siri, Oscar, Riley, Benny, Albert. It creates a singular character for the audience to picture in their mind's eye.

And we're now focused on North America. We're embedded in RVing, camping, road traveling, and the outdoors. A digital pioneer. A character with a real name. A subtle, modern Americana.

Boone.

Branding

Pulling from the same themes which led to the name Boone, I explored a number of color palettes related to the outdoors and Americana.

We arrived at an amalgam of those two ideas: a dusty, vintage Americana with support from secondary colors found along classic road trips.

For the wordmark and headline type, I decided on ITC Souvenir. It’s a classic Americana typeface from 1914, once described as “like Times New Roman dipped in chocolate”, and also harkens back to road trips of the 1970s (because of it’s ubiquity that decade).

With that out of the way, I designed and built the marketing site to focus on three specific markets:

1. Destination tourism organizations (DMOs) interested in using place data to promote their region
2. Travel and travel-adjacent brands looking to use place data for personalized marketing throughout their sales funnels
3. Local businesses who want to claim, maintain, and promote their business

Then we built a way for local businesses to find their listing on Roadtrippers (or another parent company app), claim it, subscribe, and manage it themselves.

Explorer

During our initial customer development, we found significant interest in idea of Boone, but the scope of a full implementation meant we couldn't move quickly.

As both a sales demo tool and a quick win to show off the data, we set out to build an interface for our place data which could be easily embedded on brand sites.

I designed a javascript widget called Explorer that travel brands could incorporate into their sites and landing pages as a way to both engage their audience with travel content, but also as a cheap test for a larger Boone implementation in the future.

It was a single line of code added to the page, and with simple modifications to the data properties being passed, it was incredibly customizable.

Sites could display places near the user and let them filter by category. It would display the results as either a list, a Pinterest-style grid, or a Roadtrippers-like map. And you could tap a place to see info, hours, and actionalable detail.

Business Stuff

As the general manager, I also handled business and team stuff.

Along with company leadership, I put together a forecasting model for Boone for budgeting and resource allocation at the parent company level.

And since I designed and managed the product, I also focused on team process and dynamics to keep velocity as high as possible

Results

Sigh. Boone never had a chance to get off the ground.

We faced challenges related to corporate ownership, strategy, and resources. So we got caught in no man’s land: we weren’t scoped like an internal team with a limited focus on one vertical app, and we weren’t going to be capitalized correctly like a new horizontal-business startup needed to be.

Leadership turned over, priorities changed, and budgets shifted, the plug got pulled.

Boone became the platform team once more, returning to support the suite of apps in what is now Roadpass.