Tuesday, July 23, 2013

Travel Meets Technology: A Weekend in Portland, Oregon

When Facebook Graph Search came out, skeptics bemoaned the end of personal privacy. Now that people can perform targeted queries over your social media history, there are few places to left hide. Facebook Graph Search will make it all too easy for your mother to find out that you have been drinking and for your boss to find out that you are a Republican.

The unveiling of Graph Search was exciting to me for more than voyeuristic reasons. For years, I have appreciated how Facebook has helped better connect me with people and enrich my world view. I have also liked how social media democratizes knowledge and allows for everyone to express their views in ways only experts and journalists previously could. Graph Search seemed like a useful way to learn even more about the people in my social network and also the rest of the world.

To explore my stance, I decided to see if Graph Search was useful for researching topics other than my friends’ personal lives. I came up with the goal of researching a lifestyle piece using solely Graph Search. As the target destination I chose Portland, Oregon because I had never been there and did not have many friends posting about it on Facebook. I assumed having few contacts in a location would be the most common experience of someone using Graph Search to learn about something new, as most people’s social networks tend to be fairly limited. I then planned a trip to Portland to compare the Graph Search itinerary to the New York Times’s 36 Hours in Portland, which represents an expert-curated itinerary for the same length of time.

This project became not just an evaluation of the effectiveness of Facebook Graph Search, but a study of the changing relationship between people, technology, and experts. On the one hand, technology provides us tools to scour the internet’s fares and opinions to potentially provide us with succinct summaries of the world’s information. On the other hand, it is not clear whether someone familiar with the domain in question can outperform technology. There is also the question of how we can use technology to enhance interactions with experts or to democratize the availability of expert knowledge. To explore this, I used other tools such as AirBnB, Bing Travel, and non-Graph Search capabilities of Facebook. The final itinerary is a result of cross-referencing the Graph Search and Times itineraries with sites like Yelp and with Portland locals, combined with serendipitous events.

I found my ad-hoc internet travel agent to be immensely useful for giving me fast and easy access to large quantities of information. Fare search and predictors give even amateur trip planners a good idea of times and prices at which to buy. Graph Search allowed me to see a collage of what my trip could look like visually through a simple search of “Photos taken in Portland, Oregon.” Sites like Graph Search, Yelp, and AirBnB provided up-do-date information about where people where spending their time and money. Advances in search technology, combined with innovations in peer-to-peer models for communication, allow us to learn about the world in previously impossible ways. With all this information at hand, it is tempting to think that the masses provide us with all of the opinions we need.

During this project, however, I learned to appreciate the curation of experts. Someone local or known to have good taste is more likely to make good recommendations than a random sampling of people from the internet. In general, unless you know something about the people posting about a place, it is difficult to determine how much to listen to the opinions presented. I realized that with sites such as Yelp and AirBnB, I close-read the writing style and content of reviews to form my own opinion about how “expert” I deem a reviewer about the relevant domain. It is, at present, difficult to determine the taste “footprint” of Facebook users posting about a place. Especially for researching travel and entertainment, it would be useful to be able to identify and weigh more highly the contributions of certain people, for instance relative experts or those with similar taste.

I describe the results in the form of a timeline for both the planning and trip periods.

Some night, months before.
12am.
Use only Facebook Graph Search to research a travel itinerary for Portland, Oregan (see According to Facebook Graph Search: 36 Hours in Portland, Oregon and Facebook Graph Search as a Journalistic Tool).

The next day.
Post to Facebook, Twitter, and your Gmail chat status about these posts. Chat online with a friend on the West Coast until he suggests a trip to Portland to explore these itineraries. Use Bing travel's fare predictor to decide the best time to book plane tickets. (In my case, it told me to wait.)
Tools: social media; online messaging; fare search.

Some other night, a month before.

11pm.
Plot geographic locations corresponding to both the Graph Search and the New York Times (NYT: 36 Hours in Portland, Oregon) itineraries. Discover that the Graph Search itinerary seems to be concentrated around downtown (West) Portland, while the NYT itinerary has locations on both sides of the Wilamette River. Discover that many of the places mentioned in the NYT article, written in August 2011, have already closed.
Tools: Google Maps; Google Places.

12am.
Examine the availability of people renting available apartments, rooms, and guest houses on AirBnB. Discover that the more interesting, affordable, and highly rated locations seem to be in East Portland. Make a booking in the Mt. Tabor neighborhood.
Tools: AirBnB.

Yet another night before the trip.
Make a proposed itinerary for the weekend by combining activities from the two itineraries. Cross-reference with Yelp; discover that many of the Graph Search locations seem to be reviewed less favorably as being "touristy." Cross-reference with TripAdvisor; find that the recommended activities seem to be less urban and more outdoorsy.

Use Facebook to contact Joe, a friend of a friend who lives in Portland. Make plans to meet up.
Tools: Yelp; TripAdvisor; messaging.

Friday of the trip.

11am.
Brunch at Cafe Broder, whose Scandinavian brunch comes highly recommended by the Times. Broder is worth every minute of the potentially long wait. Try the lefke (potato pancake) and a baked egg scramble.
Tools: New York Times.

2:30pm.
Walk off brunch by shopping in the vintage and curiosity shops. Walk past Pok Pok, one of the most popular Asian restaurants in Portland.
Tools: serendipity.

4pm.
Get a late lunch at Por Que No? Taqueria, recommended by the Times itinerary. The Times warns of a long wait, but during off-peak hours the line is fine. The ceviche is delicious and there is the option of getting it on cucumber slices rather than with chips.
Tools: New York Times.

7pm.
Jog up Mt. Tabor, recommended by fellow reviewers of your lodging on AirBnB. Mt. Tabor once an active volcano but is now merely a hill. Watch the gentlemen of Portland ride low bikes and skateboards down the hill as the sun sets.
Tools: AirBnB; serendipity.

10pm.
Try in vain for thirty minutes to call a cab from numbers off Google Search. (We still don't know if Mr. Taxi is real.) Default to dining at Sapphire Hotel, recommended by both your AirBnB host and a friend, a former seedy hotel that now probably has one of the better cocktail menus you have ever seen. Enjoy bacon-wrapped figs and perhaps a burger while your friend drinks the "You're not my real dad," a bourbon cocktail that comes with a cigarette.
Tools: word of mouth.

12am.
Take a walk down Hawthorne Street, recommended by AirBnB reviewers as being close to shopping and dining locales of interest. Consider stopping in and playing pool or drinking a beer at one of the bars. Walk past various closed shops and a group of 20-somethings sitting on an awning and drinking. Take the scenic route back along residential streets. Take some time to smell the flowers. Especially if it is summer, they will smell great.
Tools: AirBnB; serendipity.

Saturday.

11am.
Propose going to the Portland Saturday Market, recommended by Graph Search. Wait for Joe to veto this suggestion, saying that it is full of people from the suburbs. Have Joe instead take you for Vietnamese food at Ha VL in Southeast Portland. Be impressed by the fact that it is in a shopping complex full of Asian restaurants and that the other patrons are largely Asian. Did you know that the Vietnamese have pho for breakfast?
Tools: Graph Search; word of mouth.

1pm.
Take a walk around the Japanese garden, recommended by Graph Search, the Times, and most other travel itineraries for Portland. After you achieve a sense of peace, take in the beauty of the rose gardens, founded in 1917 and the oldest continuous operating rose garden in the United States.
Tools: Graph Search; New York Times; word of mouth.

3pm.
Get all natural, hand-crafted ice cream at Ruby Jewel, recommended by Joe. If you have enough appetite, try the salted caramel apple pie a la mode.
Tools: word of mouth.

4pm.
Do some shopping downtown at stores you serendipitously discover by wandering around. Tanner, Polar, and Yo! Vintage are all on the same block.
Tools: serendipity.

5pm.
Wander around Powell's Books, discovered by Graph Search as a highly recommended destination. The largest independent new and used bookstore in the world, Powell’s takes up a city block and has multiple sections, including ones for comics and rare books.
Tools: Graph Search.

7pm.
Discover Cacao Drink Chocolate while walking and take a hot chocolate break. Sample the hot chocolate espresso shot or a cup of melted single-origin chocolate.
Tools: serendipity.

8pm.
Dine on French cuisine at the Little Bird Bistro, recommended by the NYT as the more accessible alternative to the popular Le Pigeon, flagship of chef Gabriel Rucker.  Spend a leisurely couple of hours enjoying the food and cocktails.
Tools: New York Times.

11pm.
Imbibe local beers at  Eastburn, recommended by Joe for its proximity to Little Bird. Enjoy the wide variety selection of beers on tap, perhaps on the patio.
Tools: word of mouth; serendipity.

Sunday.

11am.

Have brunch at Woodsman Tavern, which you discovered on your way back from Broder the other day that a local called her favorite restaurant in Portland. On your way out, browse the snacks and sodas at the adjoining store.
Tools: serendipity; word of mouth.

2pm.
Ditch the original plan of going to Voodoo Donut, discovered via Graph Search, after a local tells you that it is "touristy" and a "last result." After failing to hunt down the donut truck that is supposedly the best source of donuts, pick up a snack at Blue Star Donuts before they run out for the day. (This usually happens in the early afternoon!)
Tools: Graph Search; word of mouth.

3pm.
Wander around the boutiques of East Burnside. There is Redux, recommended by the Times as an analog Etsy housing the work of over 300 artists. There are also fun surrounding vintage shops. In the way of designer boutiques there is Machus for mens's high fashion and Lille for lingerie.
Tools: New York Times; serendipity.

5pm.
Continue exploring downtown Portland. Walk around Pioneer Courthouse Square, recommended by Graph Search. Peer into some boutiques you passed by earlier but did not enter, such as Frances May.
Tools: Graph Search; serendipity.

7pm.
Dinner at Biwa, a homestyle Japanese restaurant recommended by the Times. Enjoy the yakitori (grilled chicken), handmade ramen and udon, rice balls, and sake.
Tools: New York Times.

Discussion.
It turned out that while Graph Search provided a nice initial preview of what Portland would be like, the New York Times and locals suggested higher-quality activities: those that were more highly reviewed on internet sites such as Yelp and by other locals. Serendipity is also a useful tool: finding one thing you like can help you find other similar things, either by walking around the neighborhood or by asking people there for suggestions. For all technology can do to provide new ways for people to interact, for purpose of travel it would do well to start by replicating these processes of consulting experts and exploring clusters of similar options. Since we already have sites like Yelp and TripAdvisor to allow people to do this with strangers, it will be interesting to see how Facebook Graph Search can allow us to leverage the social graph to improve upon this.