Tips and best practices for journeys

Although journeys are designed to be flexible and fully customizable, there are some fundamental strategies and practices that can help you plan, design, and manage any journey. Consider the following tips and best practices for designing and managing a successful journey.

Scope and settings

Because a journey can perform a variety of different and interrelated tasks, it's a good idea to create a well-defined scenario for a journey. Also, you should choose journey settings that support your scenario and goals. By using journey settings, you can establish constraints that determine the timing, volume, and frequency with which a journey can engage participants.

When you define a scenario, consider limiting its scope to a small aspect of a larger customer experience. Although Amazon Pinpoint supports large-scale journeys that have extensive workflows, you have more opportunities to monitor, refine, and manage a customer's experience if you design a journey to be part of a sequence of related journeys.

For example, you can design a journey that focuses on welcoming new customers and providing them with recommended first steps during their first seven days as a customer. Based on each customer's actions during the first journey, you can then add them to a subsequent journey that's tailored to their initial level of engagement. One subsequent journey might provide next steps for customers who were highly engaged in the first journey. Another subsequent journey might promote different products or services to customers who were less engaged in the first journey. By creating a sequence of smaller-scope journeys, you can continually refine and manage the customer experience throughout the customer lifecycle.

After you define a scenario, choose journey settings that support your goals for the scenario. These settings define the timing, volume, and frequency with which any part of a journey can engage participants. To choose these settings, create or open the journey. Then choose Settings from the Actions menu, and expand the Advanced settings section.

Some key goals and related settings are:

Store and use participants' local time zones

To optimize participant engagement in a journey that has a scheduled start and end time, configure the journey to use each participant's local time zone. This helps to ensure that journey activities occur when a participant is most likely to participate in those activities.

Note, however, that the usefulness of this setting depends on whether you store local time zone values in the endpoint definitions for participants. If you use this setting and the endpoint definition for a participant doesn't specify a time zone, Amazon Pinpoint doesn't include the participant in the journey. To avoid this issue, use the Demographic.Timezone attribute to store time zone information for participants. This is a standard attribute that Amazon Pinpoint provides.

Address quiet-time conflicts

If you configure an activity to send messages at a time that conflicts with the quiet-time settings for the journey, Amazon Pinpoint doesn't send the messages. Once quiet time ends, new messages are sent. If you chose to resume sending messages after quiet time ends, any messages held during quiet time will also be sent. If not, then those messages held messages are dropped.

Time zone estimation

Time zone estimation helps to estimate an endpoints time zone based on Endpoint.Location.Country and depending of the estimation methods selected either Endpoint.Address Endpoint.Location.PostalCode or both. The endpoints time zone is used to avoid sending messages at inappropriate times of the day when quiet time is configured and also when a journey sends messages based on the local time zone. Time zone estimation is only performed on endpoints that do not have a value for the Demographic.Timezone attribute.

Note

Time zone estimation is not supported in AWS GovCloud (US-West)

If a journey contains an endpoint with multiple time zones:

• The journey will start sending messages according to the latest time zone for an endpoint, when Recipient's local time zone is enabled

• The journey will stop sending messages when either all messages have been sent or according to the earliest time zone for an endpoint, when Recipient's local time zone is enabled

• The journey does not send messages to the endpoint when within quiet time of any timezone, when Quiet time is enabled

The journey will only send messages when all endpoints in the journey are allowed to receive messages based on all of the configured journey sending rules.

Journey with endpoints in multiple time zones and quiet time

For example, if you setup your journey's quiet time between 20:00 (8:00 PM) to 08:00 (8:00 AM) and there are endpoints in UTC-8 America/Los_Angeles and UTC-5 America/New_York then the journey starts sending messages at 08:00 America/Los_Angeles (11:00 America/New_York) and stop sending messages at 17:00 America/Los_Angeles (20:00 America/New_York).

Limit the number of messages that participants can receive

To help ensure that participants don't receive too many messages from the journey or project, limit the number of messages that can be sent to a participant during a 24-hour period. This can be especially helpful if a journey uses a segment that's also used by campaigns or other journeys. You might also create and use a segment that's designed explicitly for use by only a specific journey.

Optimize the number of messages that can be sent

If a journey has a large number of participants and it sends a large number of messages, factor the amount of time that Amazon Pinpoint needs to process and send all of those messages.

For example, consider a situation where you have a journey activity that sends messages to 1,000,000 participants, and the maximum sending rate for your Amazon Pinpoint account is 200 messages per second. Some participants won't receive the message until approximately 80 minutes after the activity starts. This is especially relevant if a journey includes wait activities that follow email activities. If Amazon Pinpoint hasn't finished sending all the messages by the time the wait activity ends, participants might be moved to the activity that follows the wait activity, before they've received the message.

To mitigate this risk, consider increasing the maximum number of messages that a journey can send per second, and possibly increase it to the maximum sending rate for your account. Also consider increasing the sending quotas for your account.

Limit the number of times that participants can enter a journey

Depending on the nature and design of a journey, limit the number of times that a single participant can enter the same journey. If you don't set this limit, a participant could enter a journey, complete several activities in the journey, arrive at an end activity, and then start the journey again. You might prefer to have each participant start and complete a journey only once.

Note that Amazon Pinpoint doesn't allow a participant to enter a journey if they're already an active participant in the journey. For example, Amazon Pinpoint doesn't add a participant as a new participant if the participant starts a journey and you subsequently update the participant's endpoint definition in a way that affects their inclusion in a segment (based on segment criteria) or the journey (based on activity conditions).

Maximize opportunities for participants to start a journey

The journey entry activity, which is the first activity in a journey, determines how often new participants are added to the journey. Because new or existing customers could become participants at any time, it's a good idea to configure the entry activity to add new members to the segment frequently. You can also configure the segment to add new participants automatically based on specific user attributes or events. For an example of how to configure a segment in these ways, see Building Your First Journey in Amazon Pinpoint on the AWS Messaging and Targeting Blog.

Segments

Segments are key. They determine who can participate in an overall journey and specific journey activities. When you create segments for a journey, consider the following best practices:

Create a dedicated test segment

If you have a regular group of people who test your journeys and messages, create a segment that contains only their endpoints. You can then use that segment as a consistent testing framework, especially if you use the journey testing feature that Amazon Pinpoint provides. For tips about how to build this segment, see Review and test a journey.

Use multiple segments

Although you can choose only one segment for the journey entry activity, that segment can include multiple smaller segments. Later in the journey, you can then use a multivariate split activity to divide participants into separate groups based on their segment membership. This approach can help you provide a more tailored experience for each participant. It can also help reduce processing times for email activities, because those activities will send messages to a smaller, more targeted audience.

It's also a good idea to segment participants based on actions that they explicitly do or don't do. You can do this by using split activities. For example, you can use a yes/no split activity to send participants down a Yes path if they click a link in a message, and a No path if they don't. The absence of an action can be an opportunity to reengage a participant through a follow-up activity.

Don't delete segments and endpoints

We encourage you to maintain segments that are part of an active journey. If you delete a segment that's being used by an active journey, the journey could fail and stop running. If the journey does continue to run, any participants who were part of the segment might be removed from the journey prematurely. In addition, those participants will be reported as “dropped” in the analytics data for the last activity that they were part of. This compromises the usefulness of your analytics data—you won't be able to distinguish between participants who left a journey independently and participants whom you removed.

Leverage custom attributes

To identify and add journey participants to segments more easily, consider adding a custom, journey-specific attribute to endpoints when your application creates or updates endpoints. You can then use this attribute to identify a user or endpoint as someone who should participate in a journey.

Activities

Activities are the building blocks of any journey. Therefore, when you choose the type and settings for each activity, and the relationships between activities, consider the following guidelines:

Optimize the entry activity

The entry activity, which is the first activity in a journey, determines how often new participants are added to the journey. You can either add participants based on an activity – for example, adding users who download specific music – or add participants from existing segments. Because new or existing customers could become participants at any time, it's a good idea to configure the entry activity to update (add participants to) the associated segment frequently. By doing this, you maximize opportunities for participants to start a journey.

Prepare for changes to segment and participant data

An activity's evaluation of segment conditions is based on the latest data for each participant (endpoint) in the segment, and this data might change over time. For example, a participant's favorite food could be pizza when they start an activity. That participant could subsequently change their preference to hot dogs. If this happens, subsequent activities will evaluate the participant based on the participant's preference for hot dogs, not pizza. One way to prepare for these kinds of changes is to use split activities that predict the changes and send participants down an appropriate path.

If an endpoint is ACTIVE, Amazon Pinpoint will send messages through campaigns and journeys. If an ACTIVE endpoint enters a journey and becomes INACTIVE before completing the journey, Amazon Pinpoint will continue to send messages to the endpoint.

Take advantage of the Else path

A multivariate split activity can contain as many as four paths (each with its own criteria), in addition to an Else path. The Else path is for participants who don't meet any of the criteria for the other paths. Therefore, it provides an excellent opportunity to handle unexpected or unusual cases that you might not have considered when you designed the journey.

Consider delays in receiving event data

Some event data, such as email opens, is based on information that we receive from participants' email providers. Some providers send us this information immediately, while others send it less frequently. Those delays can impact participants' experiences. When Amazon Pinpoint evaluates events as a condition of an activity, it moves a participant to a No path if it doesn't have any event data for a participant. To mitigate this risk, add buffer time to the evaluation schedule for activities that immediately follow email activities.

Avoid consecutive email activities

We recommend that you insert a wait or other type of activity between two or more email activities. This can help account for the amount of time that Amazon Pinpoint needs to process and send messages, and any delays in participants receiving messages.

Use re-entry intervals

Set a re-entry interval for when endpoints re-enter journeys. By setting a re-entry interval you'll space out the time between when users receive your messages, creating a better user engagement and also being less likely that your messages are processed as spam.

Email messages

In addition to general tips and best practices for sending email, consider doing the following before you create a journey:

Create a dedicated “From” address

Consider using a dedicated email address or domain for all the messages that you send from a journey. This provides a consistent experience across all the messages that participants receive from a journey. It also gives each participant an opportunity to adjust their email application settings to ensure that all of a journey's messages arrive in their inbox. In addition, if you subscribe to the Deliverability Dashboard, using a dedicated address or domain can make it easier for you to access advanced analytics data for specific journeys. To learn how to set up a dedicated address or domain for sending messages, see Verifying email identities.

Verify that you set up the email channel correctly

Before you publish a journey, make sure that your Amazon Pinpoint account has production access for email. If it doesn't, your account is in the sandbox environment, which means that participants might not receive messages from the journey. (In the sandbox environment, you can send only a limited number of messages and you can send messages to only certain email addresses.) Also, make sure the sending quota and sending rate for your account can support the number of messages that you plan to send from the journey. To check the sending quota and rate for your account, you can use the Email Settings page on the Amazon Pinpoint console.

Design a collection of related message templates

During the early stages of the planning process, it's a good idea to design and create a message template for each email activity that you expect to include in the journey. If you do this, you can ensure that all the messages have a consistent design. This also ensures that each message is specific to and optimized for the corresponding phase of the journey. For example, in a journey that welcomes new customers, you might have three email templates. There is one template with introductory information, another with intermediate information for users who clicked a link in the first message, and another with revised introductory information for users who didn't click a link in the first message.

Reviewing and testing

Amazon Pinpoint includes a review feature that checks for and warns you about configuration errors in a journey. It also simplifies the process of finding and fixing any errors. To find the activity or setting that has an error, click the error description.

To fix an error, follow the recommendation provided in the Review your journey pane. We encourage you to use this feature to review and fix errors before you publish a journey. As a best practice, we also encourage you to complete this review process multiple times before you publish a journey.

Amazon Pinpoint also includes a testing feature that streamlines the testing process. After you complete the review process for a journey, you can use this feature to send a group of test participants through the journey.

To ensure that only test participants can enter the journey, you can create and use a dedicated test segment with this feature. To expedite testing, you can configure this feature to reduce or eliminate wait times for and between activities. We strongly recommend that you use this feature to test all aspects of a journey, including each message that a journey sends, before you publish a journey.

To learn more about reviewing and testing a journey, see Review and test a journey.

Analytics

After you publish a journey, Amazon Pinpoint automatically starts collecting and aggregating analytics data for several types of standard metrics that apply to the overall journey and individual journey activities. We strongly recommend that you review these metrics regularly and frequently.

Among other things, these metrics provide key insight into issues to address, such as failures and errors that might have occurred when Amazon Pinpoint attempted to evaluate or perform an activity. Overall, these metrics can help you determine what is or isn't working well in a journey, which can help you design more effective journeys in the future. For detailed information about the available metrics and how to view them, see View journey metrics.

Amazon Pinpoint automatically stores your analytics data for 90 days. Depending on a journey's projected duration or your organization's long-term storage and reporting needs, you might want to store the underlying event data for more than 90 days. To do this, we recommend that you configure Amazon Pinpoint to export data to Amazon Kinesis Data Streams or Amazon Data Firehose. If you export data to Amazon Kinesis, you can also use other services and applications to perform deeper analysis or reporting. For more information, see Streaming events with Amazon Pinpoint.

Lifecycle management

As you move a journey through various phases of development and execution, keep the following in mind for each phase of the journey's lifecycle. Also note that you can stop (cancel) a journey at any time if any issues arise.

Phase	Description
Draft	The journey is being developed and hasn't been published yet. In this phase, you can change any aspect of the journey, including segments, activities, and settings for the journey. You can also leverage Amazon Pinpoint features for reviewing and testing the journey. You can repeat the review and test processes as many times as you want.
Active	The journey has been developed, reviewed, tested, and published. Depending on the journey's schedule, it might currently be running or scheduled to start running at a later time. In this phase, you can't add, change, or remove activities from the journey.
Closed	The journey has been developed, reviewed, tested, and published. It has started running and is closed to new participants. Depending on the journey's schedule and settings, it might have also passed its scheduled end time. Or the journey might have passed its scheduled start time, and it has an entry activity that's set to never add new segment members. In this phase, you can't add new participants to the journey, and no existing participants can enter the journey for the first time. However, any existing participants who are currently waiting to start an activity can resume the journey.
Stopped	The journey was developed, reviewed, tested, and published, and then subsequently stopped. You can't restart a journey after you stop it. You'll need to recreate the journey again. If you stop a journey, Amazon Pinpoint continues to perform activities that are currently in progress until those activities are complete. Amazon Pinpoint also continues to collect and aggregate analytics data for those activities until the activities are complete. It also does this for any activities that were complete when you stopped the journey. In this phase, you can't add, change, or remove any activities from the journey. In addition, Amazon Pinpoint stops evaluating the journey and doesn't perform any activities that haven't started.