Insurers Need to Resolve the Doolittle Dilemma

The Doolittle Dilemma: As Eliza Doolittle sang in the “My Fair Lady” musical song ‘Show Me: “Words! Words! Words! I’m so sick of words! First from him, now from you! Is that all you blighters can do?”

Insurers need to resolve the Doolittle Dilemma. Words continually cascade through the operations of insurance firms. Words are the essence of an insurance firm’s analog and digital existence.

Increasingly more words (whether written, spoken, or captured in applications and other documents) are flowing into, out of, and through the value chains that constitute the strategic, tactical and operational business processes of insurance companies. Insurers have historically ignored this extremely rich mine of unstructured content they own primarily because they lacked the ability to discover and manage it.

Now, semantic technologies are available to open insurer’s eyes to the rich set of competitive opportunities at their fingertips by leveraging both structured and unstructured content.

What are semantic technologies?

Semantic technologies enable data to understand and manage the meaning of and relationships between data. This becomes more powerful when semantic technologies are used from a specific perspective or context.

A context can range from focusing on a specific issue that a single department manages to an issue that spans the entire company. As insurers experiment with and implement semantic technologies, they will soon realize that the content in the company can be best considered as an ecosystem of concepts that both provides a context and supports larger contexts similar to a Matryoshka Doll.

For insurers, that context could be improving the product development process or being able to more effectively comply with e-discovery requirements. 

Potential insurance applications of semantic technologies

Insurers who begin to experiment with semantic technologies have to resist the temptation to feel as overwhelmed as Alice did when she went down the rabbit hole. One of the strengths of semantic technologies is that it enables Humpty Dumpty ‘s desire to become the master of words become reality.

“When I use a word,” Humpty Dumpty said in rather a scornful tone, “it means just what I choose it to mean — neither more nor less.” “The question is,” said Alice, “whether you can make words mean so many different things.” “The question is,” said Humpty Dumpty, “which is to be master – – that’s all.” (Through the Looking Glass)

Insurers have struggled in the past with striving to find the one correct definition of a concept such as premium or client. Now they can use semantic technology standards such as OWL (Web Ontology Language) and RDF (Resource Description Framework) to move beyond the false need to establish that single definition of the truth.

Internal and External Content

Insurance applications leveraging semantic technologies can achieve greater intelligence from internal and external content:

  • Internal content: Examples include text in documents, policy applications, claim forms, claim adjuster notes, correspondence, depositions and other legal filings, e-mail, instant messages and database text data fields. For the purposes of this report, internal content lives in the home office, field offices, claims offices, and in the insurer’s agencies and brokers.
  • External content: Examples include text in the social networking mediasphere whether used by policyholders, prospective customers, producers and other stakeholders. It also includes transcriptions of telephone calls involving customer service, claims, field operations, regulators, and agencies or brokers. I am defining the mediasphere is the collection of blogs, instant messaging, virtual worlds and other web sources in which conversations flow through and reside for a very long time.

Four major domains of intelligence

Collectively, semantic technologies enables insurers to more effectively manage internal and external content to strengthen their competitive position by achieving greater intelligence in four domains:

  • Market intelligence: monitor and manage the growing amount of content that relates to opinions, feelings, remarks and announcements created by policyholders, prospects, claimants and others about the insurance company, its products, services and operations.
  • Operational intelligence: identify and capture company and external information relevant to creating new or strengthening existing business processes.
  • Financial intelligence: monitor and manage external and internal content related to regulatory and legal matters to be in a stronger position to defend lawsuits, remain in regulatory compliance and improve company’s risk management requirements.
  • Legal & Regulatory intelligence: identify and capture content required for eDiscovery and broader regulatory requests.

Adoption Hurdles

As expected, the largest hurdle insurers face adopting semantic technologies is cultural. A sampling of several major challenges are shown below in the four hurdle areas of culture, operations, skills and expertise, and technology:


  • Realize that unstructured content is as critical to analysis and decision-making as is structured content.
  • Overcome the traditional inertia preventing content of any type to be shared enterprise-wide (i.e. create a culture of sharing data as a corporate asset).
  • Be aware that unstructured and structured content can be combined to provide richer answers to decision-making and to crafting stronger business processes.
  • Become comfortable with minimal human intervention in the data discovery process.
  • Think about what can be asked of the content or should be asked of the content without continually being blinded by asking the same questions time and again
  • Shift from asking questions that a “point” response satisfies to instead asking questions that a “descriptive” response satisfies (i.e. shift from “how many” or “how much” to “tell me about…”).
  • Estimate the costs of not being able to find the right information regardless of its being structured or unstructured content.


  • Create processes to continually identify new and existing content regardless of where the content resides in the insurance firm’s value chains and data repositories
  • Determine the business applications that could benefit from the discovery, sharability, and analysis of unstructured (and structured) content.

Skills & Expertise

  • Become knowledgeable about semantic technologies standards
  • Understand objectives and capabilities of semantic technology solutions encompassing content management, query, and presentation tools.
  • Become knowledgeable about concept extraction and data normalization across content sources.


  • Become knowledgeable about content storage techniques.
  • Become knowledgeable about true join operations from the business intelligence space.
  • Look for scalable semantic technology solutions.
  • Look for semantic technology content management solutions that index new content in real-time.

As an aside, this C.O.S.T. heuristic can be used to identify adoption hurdles for other technologies and their applications.


A company’s mere existence creates some level of data, information, and communication flows. The company’s real and perceived actions and behaviors amplifies the frequency and tone of the data, information, and communication flows.

Insurance companies are entirely based on information and their actions have financial implications for a host of stakeholders including but not limited to policyholders, producers and claimants. That is why it is imperative that insurers effectively manage all of the streams of data flowing into, out of, and through the internal and external value chains regardless of the type or structure of the data.

Semantic technologies offers insurers a golden opportunity to resolve the Doolittle Dilemma and make informed decisions from textual data regardless of where the data reside.

Note: Not discussed but critical to the topic of semantic technologies are several issues including but not limited to taxonomies (creating and using them), eDiscovery, and a deeper dive into semantic technology terms.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.