.png)
The Choice Step vs The New Capture Step
The Capture Step just got an upgrade. With our newest update, conversation designers can now build & repurpose complex user responses easier than ever. Let’s find out how.
Conversation designs are made up of a series of turns. These turns often consist of a combination of system responses and user responses. i.e. A system says: “Hello, how can I help you?” and a user can respond by selecting through a series of intents via buttons/pre-defined utterances or by speaking, i.e. “I’d like some help with my banking account.” Both Capture & Choice blocks are examples of how a conversation design can record a user response or “actions” within a conversation. Let’s break it down further.
When designing any conversation experience, most conversation designers need to capture information to better inform the experience, direct the user to the right outcome, and repurpose the responses gathered to create something custom & memorable. By effectively capturing information from their end-users they can direct the conversation to the next step in their flow and progress the experience – this is often accomplished through either a Choice Step or Capture Step.
Gathering more information on how users interact with your assistant makes no matches and fallbacks happen less frequently, makes your assistants more helpful, and makes it easier for you to design flows that are more personal to the end user and what they’re looking for.
There are many ways to gather information on how your user intends to use your experience. Whether you’re asking a simple yes or no question, prompting an open-ended response, referencing meta-data (i.e. session=1 vs. 2) or pulling from an internal database. By understanding how you can gather your data, conversation designers can create more contextual experiences to better meet their users where they are and are more likely to create lasting positive experiences with their designs.
To save user information, Voiceflow has two blocks that most customers use: the Choice step and the Capture step.
The similarities between the two steps are just as plentiful as the differences - they are both closed prompts, where the assistant is leading the conversation in order for a specific objective to be met, with varying levels of constraints.
They both require the user to input some form of information, versus the user giving a command to the assistant.
So let’s break down the best times to use the choice step and the capture step.
When to use the Choice Step
The biggest difference between the choice step and the capture step is that the choice step is meant to branch the conversation path based on intention, but can ALSO capture information as a result of that branching through invoking intents. The Capture step which cannot branch the conversation but can only capture information.
The most common use for the choice block would be creating scenarios with a Yes or a No answer, taking the user down one of two paths.
The Choice Step is a great option if you want to maintain control of where the user can navigate within the experience, in particular if it’s a linear flow.
This is also great for eliminating ambiguity of potential for no match results.
For example, you can use a yes or no question to ask your users if they’d like to restart the conversation once they’ve reached the end of a path as a way to loop the conversation back to the front, or end on the user's terms.
You can also use the Choice step as a way to confirm the information saved from the Capture step as a form of verification. This is a great way to level set in your conversation before moving onto parts of your assistant that require the information given in the Capture step to be accurate in order to move forward.
When to use the Capture Step
The capture step is best used when you are looking to capture information without branching the conversation.
Some examples of unique entities or responses could be: username, email, name, zip code, or even the reason why they came to your experience today.
Another example of using the capture step and leveraging the captured information for confirmation later on in your design. Here you can see we have a calling assistant capturing the name of the contacts to call (i.e. {usercalled}). However, in this instance, the design captures the {usercalled} response, and then moves on to prompting the user for more information i.e. {usernumber}.
Using The Choice Step And The Capture Step
Knowing when to use the choice step and the capture step can be a difficult choice in your conversation designs. So here’s a trick:
Take a look at your choice step - do you only have one intent in that step? That’s a sign that your choice step should be a capture block.
The capture step is best used when you’re designing a conversation that will have entity-only utterances. When you’re asking for a specific piece of information versus asking a broader question that could a multitude of responses about different topics.
Here’s an example of the capture and the choice blocks working together:
You can see the kinds of questions for which the capture and the choice blocks are most appropriate in this example.
Breaking off your flow into separate paths is best done with the choice block, whereas capturing specific entities to save to your system from your users is where the capture step shines.
Know What Information To Capture & When
There are a few clues within your conversation designs that could indicate you may be using the wrong blocks for the most efficient conversation design.
Remember that if you’re asking a yes/no question or have a question with just one intent, you probably want to use a choice block. If you’re capturing information like name, email, PIN, address, or any more personal information, the capture step is the best option to pick.
Want to see for yourself how to use the Choice and the Capture steps together? Check out our latest template here!
