LAX and Mastering the Final Mile

  5 m, 49 s

Few acronyms generate as visceral a reaction as “LAX.” It seems everyone passing through these pearly gates and into the City of Angels has an opinion, and they’re not afraid to share it. Consider it this way: there are 10,000 more words in our LAX data set than in James Fenimore Cooper’s epic novel The Last of the Mohicans. That’s around 400 pages of just Facebook reviews about Los Angeles International Airport. LAX’s topics report broke evenly across the lines normally seen in airport data sets; a lot of talk about staff attitude, wayfinding difficulties, and facility cleanliness; but a new topic popped into a top spot: traffic.

Spot the pattern

One way airport companies generate money is from an umbrella of sources known as “off-terminal concession revenue.” Falling beneath this umbrella is money from parking and ride sharing companies, like Uber and Lyft.

Using signals in natural language optimizes how these revenue streams are managed, enhancing the customer experience. In this way, the customer becomes the airport’s business advisor. When social media channels are structured and mined using NLP technology, airport stakeholders gain access to decisive feedback. Take this recent comment as an example:

“The procedure to get to terminal three (3) is very unorganized. Signs leading to the bus are not very visible. Once coming off the elevator there is a cluster and too many agents in the area. I think it can be controlled by 1-2 personnel, however there were about 4-5 directing traffic. People waiting to get on the bus were preceded by people getting off the elevator. Needs more organization.”

This review, isolated in seconds from within a massive data set using Lexalytics’ Semantria Storage & Visualization (SSV), informs decision makers on three areas: wayfinding, staffing, and crowd management. These sentiments are echoed again and again throughout the data set, pointing not just toward a problem, but a solution. This context is made clear by the text mining technology. In this way angry customers in a negative situation become reconnaissance agents in the field.

Ask the audience

We took a standard CSV of LAX’s Facebook comments and uploaded that to SSV, Lexalytics’ web based dashboard. After clicking “Start Analysis,” SSV returns a visual analysis of the data set. Taking a birds eye to the topic of traffic, we see an expected trend. 42% of Facebook reviews are negative. Another 42% bear neutral or ambiguous sentiment, and a small 15% bear some positive sentiment.

How can anybody be positive about LA’s traffic? Well, there are all sorts of people commenting on the LAX Facebook page. Some favorable reviews allude to complex thoughts, like an uber driver discussing how returning to LAX again and again allows you to “appreciate [the] traffic rhythm and placement of each terminal.”

Negativity with traffic is expected, but signals within this sentiment help stakeholders make a structural difference to airport facilities. The best comments point to specific and achievable goals:

“Took me 50 minutes from the time I left the rent car return to get to my terminal drop off. Horrible traffic! This was at 11:00pm on a Monday night. This airport is completely unorganized in the load and unloading zones!”

This comment illustrates three key facts: congestion around LAX does not correlate with peak LA traffic patterns; Airport Boulevard and Aviation Boulevard, where most rental car agencies are located, may be optimized for traffic; loading and unloading zones are unintuitive and may cause problems for some motorists.

Finding the perfect spot

Terminal traffic congestion happens when Person 1 asks Person 2 for a ride to the airport. However, in almost all cases, locals going through an airport prefer to retain autonomy by bringing their own car. Not surprisingly, long term parking is a function of two things: price and convenience. Going back to traffic, LAX still has a long way to go. Says one guest: “I’ve never seen a worse airport for traffic. Gridlock just to park at terminal parking. This place is a total nightmare.”

While long term parking is within a mile of the airport, motorists still need to travel down a congested eight lane boulevard, traversing several intersections. Introducing innovative traffic reducing mechanisms, like roundabouts, might save the LAX commuter headache. But the problems don’t subside once the motorist enters a parking structure.

“LAX is one of the worst airports to get in and out of. Lot C parking is a nightmare!” Says one irate traveler. “I have been waiting 30 minutes for Lot C bus on a Friday night. Ridiculous!” Complains another. Ire over parking is plentiful. Another commenter adds some specificity:

Lot C service is horrible. Bus service is spotty. Parking lot has not been paved in years. Wait on average 20+ minutes for service that is billed as ‘every 7 to 10 minutes.’ What a joke.” From overhauling boulevards to improving parking bus service, LAX needs to make radical changes based on consumer feedback. Otherwise, they’ll continue to lose valuable revenue from parking.

 

Taxonomy maps organize thousands of text documents into easy to navigate nodes. Drilling down into a node tells you more about the underlying story, while sentiment is retained in the color of the node. Here we see that there are many issues underlying LAX’s mass transit.

LAX and Sentiment Analysis

It’s easy to feel despair after dipping a toe into the lake of fire that is LAX’s social media reviews. But there’s hope on the horizon. Chief Experience Officer Barbara Yamamoto spoke candidly about the airport’s use of text analytics in a recent interview.

“LAX piloted a big data research tool that analyzed guest feedback utilizing Artificial Intelligence to assess Emotional Intelligence by looking for words and phrases that correlate with important guest experience metrics… LAX listened to the voices of nearly 17,000 guests and has benefited from insights that defined root causes or offered evidence that led to root causes when supported by other internal metrics and research.”

Since this interview, LAX has started construction on an airport train originating at the heart of the landside experience. It will move passengers to and from a consolidated rental car facility, parking structure, and the Los Angeles Metro, alleviating many of the traffic concerns present in our data set.

“LAX is listening to its employees and guests and using improved customer feedback tools to help make better decisions that will improve the guest experience.” says Yamamoto. “Sometimes it’s the little things that really add up to a memorable experience.”

From WiFi connectivity in Las Vegas to wayfinding at SFO, NLP social listening success stories are multiplying through the airport space. And it looks as though LAX is well on its way to joining that powerful AI powered future.

We do have one word of advice for our Angeleno friends as they blaze forward into this bright future: maybe tune your configuration for bus NEAR/5 “LOT C” — you’ve got some pretty grumpy people waiting for a very late bus!

 

Categories: Airport

Leave a Reply

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>