customer profile
966 days ago
 

Using semantic technology to create better consumer profiles in travel

NB: This is a guest article by Larry Smith, a partner at US-based Thematix.

It took me about 15 years from when I started travelling for business and pleasure to appreciate the value of a professional travel agent.

In 1979 as a junior MadMan I relied upon secretaries and expense accounts for guidance, while personal travel was collaboration between an advertised special, a travel agency storefront, and my girlfriend.

By 1996 my prospering Internet development company was flying people around the globe and Ann Marie, our independent travel agent, made several dozen bookings a week seem trivial.

What’s more, she made personal travel an exciting purchase, something to anticipate enjoying for all 200+ of us when we needed to get away for a week or weekend. This was personal service with expertise and excellence.

While we worked to transform ecommerce and marketing on the internet for companies including American Express, Royal Caribbean Cruise Lines, Microsoft, AT&T and Martha Stewart, our objectives were known but the consequences continued to manifest as browser interfaces transformed customer relationships while click processes pinpointed purchase pain points.

To a marketing technologist, uncovering the consumer behavior and purchase process is critical to creating the benefit in a unique user interface with data presented just so.

Once structured so the servers can spin the data, much of the creative expression is forever cemented into a template in a consistent process. Some content is enriched through data feeds from news, weather and review sites, but mostly the template makes the asset uniform; seat, bed, wheels, ticket — at a place for a price on a date certain.

At the time (circa 2000) we thought this was a job amazingly well done: margins increased, salesmanship elevated, process simplified and customer satisfaction enhanced.

Now. Not so much. The pendulum has swung too far the other way.

Data has become the soul of the interaction, and a uniform template the face of the relationship, so the personality, knowledge and insights integral to our travel experience have become a casualty. Unfortunately there is no such thing as a romance feed; when we lost Ann Marie, we lost that loving feeling.

We’re at another inflection point where margins, value, and customer loyalty are at odds with Internet access, matrix shopping options, and price driven purchase behavior.

As an industry we face a few multi-billion dollar questions:

  • Can we affordably change our systems, processes, and interactions?
  • Can our data be enriched with personality – ie, can we craft a romance feed from a big data knowledge base?
  • Can a deeper relationship at multiple points in the purchase funnel create loyalty by adding value to the customer?

How do we technologists (including myself) want to be regarded: as sales advocates delivering revenue or programming pragmatists wedded to legacy systems and outflanked by newbie start-ups?

Semantics and semantic technology are technical and marketing solutions that enable us to create “knowledgebase and data response personalities” akin to individual emotional, attitudinal and behavioral response patterns.

First, we start by creating product/benefit data packages, which are the substance and content from a search of rich databases or deep web data. Created from a search query, packages span two layers of information:

  • Benefit data: the surface web or presentation layer of matrix results, displayed in a browser template, for the customer to consider and interact with
  • Product data: the deep web of highly structured product, inventory and pricing data housed in one of more internal siloes.

Nothing particularly new here, but as Dr Alfred Lanning says in iRobot:

“I’m sorry. My responses are limited. You must ask the right questions.”

In other words the key is to accurately query for data that is then reconciled with and mapped to the true meaning of the question; ie. data fields aligned with the semantic meaning.

Semantic technology gives us another layer for metadata and ontological meaning that currently does not exist; over time we’ll create a rich knowledge base.

The business and marketing outcome is ability to wrap this package in and with many other knowledge and data stores (CRM/loyalty programs), data feeds (Facebook, news, weather, TSA), and other linked data repositories (TripIt, TripAdvisor).

This product/benefit package can be further linked to and marked-up with smart branding data so we may present richer or more relevant benefits outside the norm.

Eventually we may enable machine reasoning to deliver greater value to the consumer in a modified matrix with fewer and more relevant choices, enhanced with knowledge points; at the minimum we can increase the number of templates to deliver more tailored results based upon performance and yield.

Further examples are discussed in a prior Tnooz article from us.

For example purposes, two primary points of reasoning through inference and deduction involve dates and places.

There is a high probability that a family traveling to Orlando in April is on school Spring Break and wants a rental car for flexibility, as compared to the college student, also on Spring Break, who wants a bus to Coco Beach for a premium hotel room on the Boardwalk.

A same day round trip to Bentonville, Arkansas, means a visit to Walmart, and a mid-January trip to Detroit means the Auto Show or CES if the destination is Las Vegas. There are thousands of smaller but no less obvious times and places where people travel and seek obvious value-adds, bundled packages or are willing to pay a premium price when placed in the proper context.

Using product/benefit packages to pull data and present unique merchandising offers with relevant information to the consumer is the way to put knowledge and personality back into the process.

The richness of semantic tools and processes will permit everyone to spin the data to their own unique presentation, to enhance brand personality and deliver an engaging attitude. This will also expand distribution opportunities, and enhance affiliate programs.

Though the yield management department already knows this “in theory”, for the first time, marketing and sales can work with the IT and Internet staff to put this into practice. The web page is no longer about the price, it’s all about the options and value-add of things that make sense for the trip.

  • The Bentonville trip is enhanced by seeing the Avis Chauffeur option to pick-up, wait, and return the car thus being on-time or catching an earlier flight.
  • Booking air, room, and wheels for the The Detroit Auto Show is enhanced with points to OpenTable for dinner reservations at an exclusive table, which makes the client happy to renew a contract.
  • Downloading a branded app with the schedule and locations at CES in Las Vegas is free when booking via this supplier or agent.

Adding a layer of semantic technology allows data locked into multiple data stores to be extracted from legacy systems and displayed in more powerful and profitable ways at lower cost and accuracy.

Over time, this data will become knowledge enriched with analytics to permit value calculations that support the bottom line.

NB: This is a guest article by Larry Smith, a partner at US-based Thematix.

NB2: Image via Shutterstock.

 
 
Special Nodes

About the Writer :: Special Nodes

Special Nodes is the byline under which Tnooz publishes articles by guest authors from around the industry.

 

Comments

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

  1. A Semantic Personality | THEMATIX

    [...] following article was written by Thematix partner Larry Smith and first appeared in Tnooz on December 8, [...]

     
  2. David Kolner

    Thanks Larry for your POV and presenting the ‘Romance Feed’ of Ann Marie, which I feel is so far out of fashion right now in the eyes of the industry mouthpieces and conferences as to be the equivalent of bell bottoms. As someone actively trying to reintroduce these bell bottoms via the world of big data, I appreciate anyone rationally presenting this point of view and highlight this inevitable and emerging trend.

     
  3. Larry Smith

    @Kuan Sng
    Absolutely agree with your comments and observations.

    “Communications finesse” may never be comparable at a human vs machine level, but there is a lot of runway for the machine to process out some of the garbage and irrelevant results. As companies explore alternative filters (Bing/Farecast, Hipmonk/Agony) consumer will expect to have more relevant results with a recommendation, not just a huge matrix of options.

     
  4. Kuan Sng

    Interesting & thought-provoking – the intangible “Ann Marie” effect is a benefit many HNWIs are invariably used to and reluctant to forsake, hence, the success of personal concierges. The next generation of consumers will however seek out and have higher expectations of the online channel – whether or not they use agents. These high expectations, however, should not inject undue risk aversion to travel companies in their experiments. Mistakes will invariably be made in relying on machine-based algorithms, irrespective of semantic sophistication. But which quality human agents never took risks in suggesting something outlandish or improbable? Only they had the relationship/touch/finesse to communicate them. Such goodwill elasticity is something that can also be incorporated into successive online “iteractions” between consumer and company.

     
  5. Jonathan Alford

    This is a good article, Larry. It will be fascinating to continue to see themes of psychographic and societal intelligence applied to experiential traveling continue to evolve in this space.

    A few years ago now, at Traveler.com, we had planned out something similar, where we had the process and algorithms designed to help us understand not only a traveler’s explicit interests (eg, going to the Auto Show), but also their implicit needs (eg, how risk-oriented they are – “push it vs play it safe” – or their preference for the “real jungle vs the urban jungle”).

    Using some of the capabilities you mentioned and through semantic inference from interest in certain images, for example, though certainly not as evolved as it is now, we could potentially get a great sense of a traveler’s explicit and implicit profile and create the best experiences for them.

    It will be exciting to see evolve – the example I like to use is that when we looked at Disney World, a great microcosm of the travel world, we could understand that a family going there could be directed to the deluxe “pampered vs modest” Animal Kingdom Lodge, a “real jungle vs urban jungle” resort, rather than being directed toward the Ft. Wilderness campground (modest) or Contemporary Resort (Urban jungle).

    Then, if we know Dad has a higher risk-orientation (“push-it vs play it safe”), we could direct him toward the Richard Petty Driving Experience to fulfill that “need for speed”, which in high likelihood, neither he nor a travel agency would know about to match him with it.

    So opportunities from what you describe can not only help create and deliver better experiences to travelers, but also create upsell/cross-sell opportunities and reduce supplier customer acquisition and operating costs.

    And of course, with that understanding of travelers, it was natural to extend it to the concept of providing recommendations – and affirmation – for travelers based on not only their trusted network of friends/family, but also from other travelers who think like them and share similar personality perspectives. Facebook Connect didn’t invent it :) And of course there have been other cool startups that recognized similar things as well.

    Unfortunately our investors defaulted in early 2009 before we could get it off the ground and the site has been dormant, but that’s another story with some good battle scars for me, and the the premise you’re describing is still quite sound.

     
    • Larry Smith

      @Jonathan Alford
      Thanks for the compliment!

      As you touch upon, consumers are forming certain expectations and are granting permissions for companies to leverage. The Petty Driving School example might first manifest itself as pictures posted to Facebook, then migrate to a car rental booking where the Online Travel Agent or Car Rental Company should know to make the first recommendation as a premium priced sports car. This should make the rental faster and more enjoyable for all involved.

       
 
 

Newsletter Subscription

Please subscribe now to Tnooz’s FREE daily newsletter.

This lively package of news and information from Tnooz’s web site provides a convenient digest of what’s happening in technology that drives the global travel, tourism and hospitality market.

  • Cancel