AI-based robots while continuing to create content without back-end strategy. This creates more and more problems later when the brand adopts the AI conversational interfaces for customer relationship.
Implementation of Haphazard – succumbing to the temptation to intervene, by adopting and implementing tools without worrying about their evolution or their integration into the global content ecosystem of the brand.
To avoid these extremes, follow the advice of Noz Urbina, founder and content strategist of Urbina Consulting, who states that a solid underlying content strategy can – and should – inform your chatbot strategy. If you have invested in content to the rich structure then you are probably more ready for the chatbot than you think.
In his lecture on the Intelligent Content 2018 Conference, Chatbots: How to Integrate Them into Your Existing Content Strategy Noz shows how to reuse many existing content strategy elements to use them.
First and foremost, Noz suggests to everyone to work on the same idea of what chatbots are and what they are not.
RELATED CONTENT: How to Architect Your Content Strategy
What's a chatbot? Terms and Definitions
A chatbot is simply "software that automates the conversation with people, especially on the Internet," according to Kristina Podnar's definition . Robots fall into the broader category of conversation interfaces or language-based user interfaces and can be text, voice, or a mixture of both.
Here is an example of typical chatbot interaction.
the answer is verbal.
In a mixed presentation, the bot displays the current temperature and displays a temperature forecast graph. The user is asked to press to get more information.
To determine the intent, the bot recognizes a command from a predefined grammar or uses Natural Language Processing (NLP) to parse the input.
A fixed grammar bot consists of known commands, such as "get weather". Noz explains that this style of bot is useful for chatbot devices used in cars or in other cases interaction.
The weather response example contains fixed and variable parts. "Temperature right now" is a fixed answer. The city and the temperature are variables.
NLP-based bots are the front end of more sophisticated cognitive systems, as Val Swisher explains in a presentation on the processing of natural language at the Intelligent Content conference.
NLP analyzes component sentences. Think of a sentence as a bread. Every word or phrase is a slice in the bread.
anaphora resolution .
content strategists must suddenly become coders. Much of the chatbot preparation work overlaps the work done by content strategists to improve the customer experience, ensure content ranking for search and prepare for personalization.
Printing of your travel cards and job analysis
The functional components of chatbot technology should not scare content marketers who are well versed in the real tasks and travels of their audience.
Remember the beginnings of your content marketing efforts and content experience. Your team has probably deliberated carefully to determine the travel cards including analyzing the issues that your audience might have each time. These questions have probably determined a part of your content plan, whether or not you are considering a chatbot.
"As part of our journey, we create pairs of questions and answers," says Noz. "We match questions with answers, whether from the content we have or the content requirements we need to create."
If you have not performed a task scan, now is the time. The quickest way, as Mr. Noz says, is to check the search logs to see what people are looking for and to ask customer service and support what are the questions people are still asking. and even.
To learn how to do a correct task analysis, Noz offers these resources:
Once you have developed all the user tasks, you can decide which ones are suitable for chatbots and which ones are not. A common trap is to believe that your new robot has been the focus of all efforts because your content has already covered everything. But remember: a bot that runs a job successfully is more valuable than a robot that tries hard to do all things.
Careful job analysis can help you decide where channel transfers make sense. A channel transfer is the point at which a user asks for something that your bot does not understand or can not answer. This is a opportunity for transparency and openness that users will appreciate. Instead of asking more clarification questions, your program marks a human representative to intervene or provides a link where users can search for additional answers.
Based on the capabilities of your bot and your own goals, determine if your bot's support will be narrow and deep (as a digital expert) or superficial. According to Noz, it is impossible to be both deep and broad, because the resources of each are unlimited.
"You must choose. The reason we do a T and not a square is that we can not cover everything, "he says. "It will not work."
Wait, What is a piece of content ?
A "content" content is simply a unit of content – usually a small part of a larger work. Here is an example from a presentation given by Noz a few years ago. The revised version divides a formerly dense paragraph into shorter, more pronounced pieces that make it easier to understand the content.
A content block can provide a simple answer to an ideal question for a chatbot response:
presented an excerpt user-friendly.
is another great chatbot resource that documents the structure of your content (content types, metadata, and tags).
CMI Chief Strategy Advisor Robert Rose stated : "An excellent metadata strategy is in itself as important as the content created."
You may have to make your metadata and taxonomy even more specific to support chatbots. Indeed, a conversation interaction is like a human conversation powered by a search engine. But search engines do not work the same way as the human brain.
"When you say something, I look in my brain and I answer," says Noz. "The more I know about your situation, the easier it will be to get the right answer the first time."
The same goes for search engines. A taxonomy establishes connections between the metadata tags. Connecting concepts and content through taxonomy helps your bot return the correct answers.
One of Noz's clients experienced a 70% increase in the accuracy of chatbot responses by refining the taxonomy, without changing the underlying content.
Setting your metadata and taxonomy for a chatbot requires a lot of work. But Noz offers this encouragement: "What you must have for chatbots is extremely useful everywhere else."
How to adopt a customer-focused strategy for your content
Content Strategy Job Supports Chatbot Policy (and Vice versa)
The more your brand adopts chatbot technology strategically, the less it will be necessary to work (read: investment) later, when the public expects to automatically interact with brands – to their conditions. In other words, the more content you create without considering conversational technology, the more difficult it will be for you to convert that content when you scramble to catch up.
Fortunately, if you've structured and categorized your content at this point, your team is probably more ready for the chatbot than you thought.
Here is an excerpt from Noz's speech:
For more practical help in designing the perfect chatbot strategy for your brand, consider joining other fellow technologists and content marketers at the upcoming ContentTECH Summit. Record here for event updates.
Cover image by Joseph Kalinowski / Content Marketing Institute