04-20-2017 02:58 PM
Here's what's new in the April 2017 release of SAS Customer Intelligence 360:
SAS 360 Discover is available for production use by customers. The web marketing and mobile marketing features of SAS 360 Engage are designated production quality and are available for sale with a limited availability status. This status is prequalified by the SAS Customer Intelligence 360 team to ensure a good fit with customer requirements. All other features are evolving and are not intended to be used as production-quality features.
Data views are a new object type introduced in this release. A data view enables you to combine multiple events and rules to define a business scenario that you can use in targeting and conversion metrics for activities and tasks. For example, suppose you want to track all purchases of running shoes that cost more than USD 150. You can define a new data view for this condition by combining two basic events in a logical expression:
website purchase event where cart value is > USD 150 and product category is running shoes or mobile app purchase event where cart value is > USD 150 and product category is running shoes
You can then use this data view as a primary or secondary metric across multiple activities or tasks. Whenever a customer purchases more than USD 150 in running shoes, it is considered a conversion for the data view.
You can also use data views in targeting for tasks. For example, you could use the “Over USD 150 running shoes purchase” data view in the targeting criteria to target customers who have purchased more than USD 150 in running shoes.
These changes are part of the new data view object type:
Existing published activities and tasks (with events used as primary or secondary metrics) continue to work without any changes.
This release includes changes to the left pane for navigating to areas within SAS Customer Intelligence 360. You can now navigate to different areas via a persistent left navigation bar with icons for direct navigation. Click an icon to navigate to a solution area in SAS Customer Intelligence 360.
The navigation bar can also be expanded to show the labels that correspond to each icon.
You can now select items directly from the default list views and complete actions such as deleting, opening, or renaming selected items. You can remove any selections by clicking Clear Selections in the upper right corner of list pages.
As an administrator, you can now configure a setting to allow only specifically designated spots on a web page to be selected when users create new spots in SAS Customer Intelligence 360. On the Page and Session settings page, you now have an eligible spots setting with two options:
If you select the second option, you specify selector text that can be used to identify eligible spots. If you are a web developer, you can create a subset of eligible elements by adding the selector text to the HTML for the pages on your site.
When users create new spots in SAS Customer Intelligence 360, they can select only elements that have been defined as eligible. You can control which website elements can be selected as spots for use in tasks.
Multivariate testing is a major new capability that is introduced in this release. The goal of multivariate testing is to determine which combination of variants performs the best out of all possible combinations. For example, suppose you create a spot set with four spots. Each of those spots is associated with four creatives. By creating a multivariate testing task, you can test all possible combinations of the four creatives in the four spots. Each combination is a variant, and in this case there are 256 possible combinations, or variants.
Multivariate testing is similar to A/B testing, but there is an important difference. An A/B test determines the effect of a change to one variable at a time. However, a marketer is often interested in changing more than one variable at a time, because there might be an interaction between variables that results in a combined effect that performs better or worse than single changes. In a multivariate test, multiple variables can be tested together to uncover the most effective combination for reaching a single conversion goal: signups, clicks, form completions, or shares. A multivariate test can eliminate the need to run several sequential A/B tests on the same page.
The most difficult challenge in multivariate testing is the amount of visitor traffic that is required to achieve meaningful results. Because of the full factorial nature of these tests, the number of variations in a test can add up quickly. In a test with many variations, the traffic that is allocated to each variation is lower.
In the preceding example, you were testing all possible combinations of the four creatives in the four spots, and there were 256 possible combinations, or variants. Suppose that in order to provide reliable results, each of the 256 versions of the page must be shown to at least 10,000 visitors, or a total of more than 2.5 million visitors. Such a test would take so long that the results would be meaningless.
SAS has developed a solution to the problem of optimization in the field of experimental design. Originally designed for applications in manufacturing, where each variant is expensive or difficult to perform, a multivariate test can be optimized by minimizing the cost while meeting a minimum level of reliability.
The constraint in SAS Customer Intelligence 360 is the limited traffic to share over a large number of different combinations. The OPTEX procedure from SAS/QC samples the design space as efficiently as possible to reduce the number of required variants.
In addition, SAS Customer Intelligence 360 fits a model using the measured data to predict the outcome for the untested variants. The marketer gets reliable results about the conversion rate of all the possible combinations of content from a subset of measurements. In one example, the required visitor count dropped from more than two million to 200,000 after optimization. The test ran in only 10% of the original full factorial design.
Use of optimization and model fitting techniques to reduce the time required to perform multivariate tests is unique to SAS. SAS Customer Intelligence 360 also enables the marketer to impose constraints, such as requiring specific combinations of content, or excluding some combinations. Optimization takes these constraints into account when generating results.
You can now modify more properties in data descriptors without creating a new descriptor for uploading data. With this release, you can update the values for these properties: identityAttribute, channelContactInformation, segmentProfilingField, predefinedValues, and uniqueValuesAvailable. This change eliminates the need to create a new descriptor in order to update these values for situations such as a mistake in the initial descriptor or changes in how the data is used.
Previously, you could upload data descriptors with data items that already existed in SAS Customer Intelligence 360. This ability became problematic when many data items had the same label, which made it difficult to know whether you were using the right data item in other areas (for example, when building segments).
Now when you upload a descriptor, the REST API checks to see whether a data item already exists. If it does exist, the REST API returns an error message indicating that the data item already exists. To successfully upload the descriptor, the data item’s label needs to be changed to a unique value.
Mobile marketers can now customize their large template in-app messages so that they comply with the branding guidelines associated with their mobile application. More specifically, when you configure an in-app message task, you can define custom, creative elements of the message, including these:
In this release, mobile marketers can define a stand-alone push notification task where the creative (push message) is sent in bulk to its defined target audience based on a defined, set schedule. You set the start and end dates.
When viewing the details for a planning item, you can now see the level name of the planning item displayed in the title. In previous releases, when you looked at the planning details for the Big Event, after a few minutes you might have forgotten whether you opened the Big Event Plan or the Big Event Campaign. Now, in this release, when you open this item, you know this Big Event item is your Plan.
It is important to avoid allocating more money to cost centers than you have available in your budget. When associating a budget to cost centers, you now have a configuration option to restrict overbudgeting. When overbudgeting is restricted, the sum of the cost center allocation cannot exceed the budget for the planning item. If you are authorized, you can edit this configuration setting from GeneralPlanning SettingsConfigurations.
Within the Calendar view on the Plans list page, you can display a Gantt view, including the option to define a custom date range for the display period. For example, instead of a monthly, quarterly, or yearly view, you might want to view planning items for a six-month range such as Jan 1, 2017 – Jun 30, 2017.
When you submit a planning item for approval, you define a set of approvers and duration (in days) for each approver. Now, you can track the start date and due date for each approver as well as the overall approval start and due dates. This enhancement provides you with greater visibility into the approval timelines prior to approval initiation.
Preproduction services are intended for evaluation purposes only and not for production use. All services related to the creation and sending of email are currently designated preproduction. Therefore, the following enhancements to email are considered preproduction.
SAS Customer Intelligence 360 now captures hard bounce events that might occur when you attempt to send an email to an incorrect email address. The incorrect email address is recorded in the system. SAS Customer Intelligence 360 then prevents further attempts by the tenant to send emails to this address.
Using SAS Customer Intelligence 360, you can now capture complaint events that might occur when a recipient marks an email as spam. If the spam information is available from the destination ISP or domain, then the system captures it and creates a record so that you do not continue to send email. However, not all ISPs share the details of spam complaints registered by email recipients.