Are Digital Assistants the New Face of Search?
Digital assistants are increasingly becoming an essential part of our lives, both at home and at work. And as they become more prevalent, organizations are starting to ask the question: How can intelligent assistants help with search?
Understanding Intent Through Voice Search
To understand how digital assistants might help with search, first it's worth looking at how people use voice search and artificial intelligence (AI). When we type in a query, we tend to use shorter phrases. This is because we can type faster than we can speak. But when we use voice search, our behavior unconsciously shifts to accommodate the new technology. The average length of a question spoken aloud is three words or more, while the average length of a question asked by typing is between one and three.
Voice searches often produce longer query strings than other types of searches because consumers are more likely to ask a question (represented by phrases such as "who," "how," "what," "where," "why" and "when") and expect a response that is complete and comprehensive.
When using digital assistants for search, the challenges are similar: the users use natural language and longer queries. Instead of entering a query like “vacation,” they use complex questions like “how many vacation days I have left for this year?”
We need to prepare our system for this and, obviously, support both written and voice queries. But we also need to keep in mind that users will be expecting one single response, not a huge list of potential results — and these responses might be more complex answers to their questions.
Related Article: Enterprise Search Procurement: Proof of Concept or Pilot?
What Search Vendors Are Doing Right Now
Because of these fundamental distinctions, the transition from traditional to intelligent searches is revolutionary. Here are a few considerations:
Language
Disagreement between the spoken and written forms of a language is relatively frequent. Natural language search works a little differently from how we're accustomed to using traditional search, where we may type in our requests. Natural language search is becoming more and more user-friendly, enabling us to communicate with digital assistants in a way that sounds much like a regular conversation.
Questions
Natural language search allows us to ask complex questions rather than just typing in a word and hoping that the search engine figures out our intent. This technology is based on a branch called Natural Language Processing (NLP), which is a way for computers to understand human language. It involves analyzing text to extract meaning from it. NLP is already used in many different applications, such as automatic summarization, sentiment analysis and topic detection.
Natural language search is still in its early stages, but it has the potential to revolutionize how we use search engines.
Precision
Despite natural language search's immaturity, the technology is becoming better at replacing a list of search results with straightforward, authoritative responses. It is in this context that digital assistants really excel, when they can provide such data in an understandable style. When the user asks the intelligent assistant, he/she wants to know the one and best answer, rather than get a long list of results in the traditional ten-blue-lines style.
Personalization
Intelligent digital assistants such as Siri, Alexa and Cortana, as well as enterprise systems, all include user identities. Whether you're doing a search using your laptop in the office or voice search on your phone in your living room, the results will be tied to your profile. This connection paves the way for personalized search results.
Generally speaking, the more often users use digital assistants, the more specific their searches will be, as the engine also learns from and about their behaviors.
Learning Opportunities
Related Article: Why Communications Plays a Starring Role in Enterprise Search Success
Training the Engine (aka How the Engine Learns)
New technologies tend to get more complex over time, and digital assistants are no exception. They learn from their users and provide better results over time. By constantly gathering data and feedback from users, digital assistants can slowly but surely become more efficient and effective — though there's still a long journey ahead.
Curated vs. Auto-Responses
Besides training the engine and relying on its intelligence to provide reliable, trustworthy responses, it’s also important to have an option for human curation. A selected team of subject matter experts in the organization should be able to review, edit or update the responses as needed. This ensures that the information provided is accurate and up-to-date.
It’s also important to have a plan for when things go wrong. If there is an issue with the digital assistant’s response or if it’s not providing the right answers, it’s important to have a way for users to submit their feedback and quickly fix the problem in the backend.
AI, the ‘Added Ingredient’ for Enhanced User Experience and Engagement
Tech giants like Microsoft, IBM and Google are pouring resources into exploring cutting-edge uses of machine learning, AI and other related technologies, and personal assistants are powered by enterprise AI. The engine may have access to a user's email, calendar, social network, geo-location and mobile app data before analyzing the user's behavior and providing answer suggestions. Oftentimes, the AI engine has already analyzed the data and made some recommendations before the user has even had a chance to ask a question.
Related Article: Has Microsoft 365 Been Clinically Tested?
A Promising Start, With a Few Cautions
Digital assistants are becoming increasingly powerful, and they will only continue to become more so. With that said, it is important to remember that they are still tools and should be used as such. They are not perfect and should not be relied upon exclusively. Also, to get the most benefits, users must learn and use this new technology — in a very similar way as they need to learn how to enter their classic queries into a traditional search box.
The journey ahead of us is long, but promising. There are still many unanswered questions, and we need to work on both the user and technology sides to get it right. We need to think about how users can best interact with this new type of technology and design digital assistants that are not just helpful but also trustworthy.
Learn how you can join our contributor community.
About the Author
Agnes Molnar is the CEO and managing consultant of Search Explained. Agnes is an internationally-recognized expert in the fields of modern search applications, information architecture, and Microsoft technologies.
Connect with Agnes Molnar:
Featured Research
Buyer's Guide
Read Now
Research Report
Read Now
eBook
Read Now
Research Report
Read Now
White Paper
Read Now
Research Report
Read Now
Guide
Read Now
Guide
Read Now
Related Stories
Reworked CONNECT 2024 - Pre-Registration is Now Open!
