About AmpIQ

AmpIQ enables customer-centric marketing campaigns. Use segment insights to build high-value segments. Use those segments to add audiences to campaigns. Build campaigns that send those audiences to any combination of downstream marketing workflows.

AmpIQ promotes customer-centric marketing by enabling new insights on every customer, smart segmentation, and seamless campaign integration from an intuitive interface that makes it easy to apply the right segments, test and optimize, and then grow loyalty and lifetime value.

• AmpIQ generates customer- and segment-level insights, including for brand and channel behaviors, product preferences, revenue sizing, and recommended actions.

• AmpIQ enables out-of-the-box predictive models, segments, and personas designed to improve customer lifetime value with a visual interface that allows you to explore and create segments on your own.

• Campaigns can be activated cross-channel to optimize channel mix, drive improved experiences, and improve return on investment.

• Apply closed-loop measurements, multivariate testing, and control groups to transactions, revenue gains, and key performance indicators for customers.

• Send these results to any 100+ integrations and any of your downstream workflows, including to marketing, customer experience, advertising, and attribution tools.

Common workflows

The most common workflows for AmpIQ involve segment insights, predictive attributes, and campaigns. Use AmpIQ to:

• Discover customer segments with product affinities, and then identify the right products for that segment

• Review customer insights from the Segments tab to identify growth opportunities

• Perform opportunity sizing on a per-segment basis

• Build segments by customer value, i.e. “top 1%”, “top 5%”, “top 10%”, etc.

• Review data by predictive model, including predicted customer lifetime value (pCLV), churn propensity, next best product, and discount sensitivity

• Build segments that leverage customer attributes, behavioral data, and predictive attributes

• Build detailed campaigns with A/B groupings, and then send them to any customer channel

Requirements

AmpID has specific data requirements for data source quality, columns and tables, fields, and additional data sources. (This link opens a topic in a new tab.)

Campaigns

A campaign is a message or offer that is sent to a specific group of customers or recipients.

Use the Campaigns tab to activate a variety of use cases across your marketing stack. For example:

• Messaging a group of high customer lifetime value (CLV) customers on Facebook for a VIP event

• Emailing customers with an affinity toward blue button down shirts for a product liquidation

• Targeting a segment of customers who are more likely to add a visit to the hotel spa

• Messaging a group of customers likely to churn with a special offer

• Showing an ad to re-engage customers who have unsubscribed from email

• Reaching out to customers that have enrolled in your loyalty program

• Identifying customers within existing lists, and then pushing targeted subsets for downstream messaging to any channel.

• Using a control group, along with any number of recipient groups, to measure the quality of a campaign.

How campaigns work

A campaign is a message or offer that is sent to a specific group of customers or recipients.

A campaign is defined in the Campaigns tab, from which you configure the segments that will be associated with the campaign, the downstream locations to which the campaign will be sent, and the time at which the campaign is to begin.

Segments are associated with a campaign. For example:

• A segment that includes the list of customers for this campaign.

• A segment that defines a list of customers who, if also present in the segment that includes the list of customers, will be excluded. This segment is optional, but will take priority over the list of customers in the included segment.

A campaign must be associated with at least one destination, such as Klaviyo, Sailthru, Facebook Ads, Google Ads, HubSpot, etc. A recipient group may be configured to send data to more than one destination. Each recipient group configured for a campaign defines its own list of destinations.

At the scheduled date and time, Amperity will run the segments and build the customer list, after which Amperity will send the results to each configured destination.

When this process is complete, your campaign is ready to be scheduled.

Amperity supports the following types of campaign workflows:

• One-time campaigns

• Recurring campaigns

One-time campaigns

A one-time campaign represents a specific campaign message that is configured to be sent only once.

A one-time campaign can be configured to run in a similar manner as a recurring campaign, but with the purpose being to communicate messages to customers they should see only once. For example:

• A welcome message to a customer who has joined a loyalty program.

• A message to a customer who has signed up for a credit card.

How to configure a one-time campaign

1. Give your campaign a name.

2. Choose a segment that represents the list of customers who will be the recipient of your campaign. You may also choose a segment that contains a list of customers that will be excluded from the campaign.

3. Configure your recipient groups. A control group may be configured, along with at least one recipient group. Click Add recipient group to add more recipient groups. Rename them as necessary.

4. For each recipient group (including the control group, if configured for this campaign), assign at least one destination. You may assign more than one.

5. Configure recipient list delivery. One-time campaigns you have two options: at a scheduled date and time in the future or as soon as possible.

6. Provide the campaign launch date. Amperity uses the campaign launch date to improve results tracking for campaigns.

7. Click SCHEDULE to put this campaign into the queue. Amperity will process the segments, and then send the results to the configured destinations. Please allow for enough time for Amperity to complete this process before kicking off the campaign in the downstream system.

Recurring campaigns

A recurring campaign is sent automatically based on a state change, such as an accepted return, a change to a loyalty program, or an alert based on credit card status, with a predefined campaign message and cadence to a list of recipients.

A recurring campaign has the following components:

1. A state change that initiates a campaign message.

2. A segment that defines a list of customers to which the campaign message applies. This segment can be configured to limit the list to certain types of users, such as only business travellers, high-value customers, and so on.

3. A launch cadence that defines the frequency–daily, weekly, monthly, quarterly–at which the campaign messages are run.

How to configure a recurring campaign

1. Give your campaign a name.

2. Choose a segment that represents the list of customers who will be the recipient of your campaign. You may also choose a segment that contains a list of customers that will be excluded from the campaign.

3. Configure your recipient groups. A control group may be configured, along with at least one recipient group. Click Add recipient group to add more recipient groups. Rename them as necessary.

4. For each recipient group (including the control group, if configured for this campaign), assign at least one destination. You may assign more than one.

5. Configure recipient list delivery. Recurring campaign require a cadence (daily or weekly), a start date and time, and then an end date. The end date has three options: never, on a specific date, or after a specified number of occurrences.

6. Click SCHEDULE to put this campaign into the queue. Amperity will process the segments, and then send the results to the configured destinations. Please allow for enough time for Amperity to complete this process before kicking off the campaign in the downstream system.

Campaign messages

Use the Campaigns tab to activate a variety of use cases across your marketing stack. For example:

• Use Facebook Ads to message a group of high customer lifetime value (CLV) customers about a VIP event

• Use Klaviyo or Sailthru to email customers with an affinity toward blue button down shirts for a product liquidation

• Using Facebook Ads and Klaviyo to target a cross-channel segment that contains customesr who are more likely to add a visit to the hotel spa

• Use Campaign Monitor or Mailchimp to email a group of customers who are likely to churn with a special offer

• Using Facebook Ads and Google Ads to show ads to customers who have unsubscribed from email, but you want to re-engage

• Use Shopify data to identify your loyal customers, and then send them additional offers

• Identifying customers within existing lists, and then pushing targeted subsets and/or apply exclusions for downstream messaging to any channel

• Using a control group, along with any number of recipient groups, to apply A/B tests that help measure the quality of a campaign

Churn prevention

A churn prevention campaign seeks to engage customers as their likelihood of making a transaction diminishes. This type of campaign uses a series of messages based on thresholds to apply deeper discounts and different messages as probability declines. This approach also allows different strategies to be applied at each stage in the process or based on the likelihood of that customer completing a transaction.

Use a query to identify users based on predicted customer lifetime value (pCLV), specifically to identify the probability of a user from making another transaction. The probabilities could be broken down into the following thresholds:

• Cooldown: 50-60% probability of future transactions.

• At risk: 35-50% probability of future transactions.

• Highly at risk, 20-35% probability of future transactions.

• Lost, 0-20% probability of future transactions.

and then define a message that is specific to each threshold with custom discount levels, offers, and calls to action.

Important

Part of a churn prevention campaign includes being careful about which messages are sent, and when they are sent. For example, use a frequency safeguard to set a threshold to ensure each user receives an email once per month, once per quarter, once per X days, and so on. Use another frequency safeguard to ensure that a user is eligible to receive a message targeted at the correct probability threshold.

Member activation

A member activation campaign seeks to welcome customers as members to loyalty programs, credit cards, and other opt-in activities. This type of campaign uses a series of messages to help pull the customer closer to additional activities and benefits based on what they had just opted into.

For example, use a query to select all users who have signed up for a loyalty program within the previous 24 hours. Send the results of this query to a downstream email campaign manager, and then send a welcome message to those users.

Predictive models

A predictive model is a feature of AmpIQ that predict customer behavior, such as predicted customer lifetime value (predicted CLV), churn propensity, product affinity, and lifecycle events.

Note

Predictive models require at least four years of historical data (five years or more is recommended), along with the following tables already configured in your customer 360 database:

1. Customer_360

2. Merged_Customers

3. Transaction_Attributes

4. Transaction_Attributes_Extended

5. Unified_Itemized_Transactions (updated to include product catalogs)

6. Unified_Transactions

Common inputs to models

Columns from the Merged_Customers, Unified_Transactions, and Unified_Itemized_Transactions (including product catalogs) are used as inputs to predictive modeling. For multi-brand tenants, the amperity_id column from the Customer_360 table is also used as an input to predictive modeling.

The following columns are common inputs to predictive models:

Source table	Input columms
Customer_360	The amperity_id columns is used as an input to predictive modeling for multi-brand tenants.
Merged_Customers	The following columns are common inputs: amperity_id birthdate city email gender given_name phone postal state surname
Unified_Itemized_Transactions	The following columns are common inputs: amperity_id is_return item_discount_amount item_quantity item_revenue order_datetime order_id product_category product_description product_id product_subcategory Important The product_category, product_description, and product_subcategory columns must be joined to the Unified_Itemized_Transactions table.
Unified_Transactions	The following columns are common inputs: amperity_id order_datetime order_id order_cancelled_quantity order_cancelled_revenue order_discount_amount order_quantity order_returned_quantity order_returned_revenue order_revenue purchase_brand purchase_channel store_id

AmpIQ models

AmpIQ provides models for the following behaviors:

• Churn propensity is a predictive model that determines the likelihood that a customer will be active at any point in time, based on their purchase history. The churn propensity model outputs a score between 0 and 1 that represents a customer’s probability of return.

• Predicted customer lifetime value represents the total value of all orders a customer is predicted to make if they return to make another purchase during the next 365 days.

• Product affinity is a predictive model that identifies which customers are likely to purchase by using a combination of historical purchase data and lookalike audiences. The predicted affinity model outputs a ranked list of customers with three recommended audience sizes.

Segments

A segment defines a specific attribute profile that can be used to initiate marketing campaigns using the list of customers that match that profile.

Use the visual editor in the Segments tab to easily build a segment from a series of drop-downs and picklists. As you build the attribute profile refresh segment insights to see how many customers match and to see how much value they bring to your business. When the segment is ready, activate it, and then use it to initiate marketing campaigns.

The Segments tab contains a list of active segments along with recommended segments highlighted across the top of the page.

How visual segments work

The Visual Segment Editor is the user interface for building segments in AmpIQ. This editor uses a series of drop-downs, picklists, conditions, and values to define an attribute profile. Refresh segment insights to see how many customers match the profile, and then active it as a segment.

The Segment Editor is available from the Segments tab. Click Create from the Segments tab to open the editor, and then start building your segment.

Follow this sequence when building a segment:

1. Give your segment a name.

2. Select a table that contains the attributes you want to use, and then use the drop-downs, picklists, conditions, and values to define the attribute profile.

   Note

   The Customer_360 table is the default. For many segments it’s the only table you will need.

3. When more than one attribute is present within an attribute group use AND or OR to filter values on more than one condition.

4. Click Add section to add additional attribute groups.

   Each attribute group has the same requirement for using AND or OR to filter values. There is another AND versus OR choice that must be made in-between each each attribute group.

5. For each attribute group that is added to the segment, select the database table that contains the attributes, and then select attributes.

   Tip

   Any table that is listed in the lower right may be used to build the attribute profile. Open the Data Explorer to learn more about the individual attributes within these tables.

6. To view updated segment insights based on the currently-defined attributes, click REFRESH. This will validate the segment, after which segment insights are updated to match the condition-value pairs associated to all of the attributes in the segment.

7. When the segment returns the parameters that meets the goals for your campaign, click ACTIVATE.

Segment names

A good segment name is clear and concise, is not longer than necessary, uses patterns to help lists of segments stay organized, and considers how it might be represented in downstream workflows, such as lists of segments in the Campaigns tab in Amperity, but also external systems such as Braze, Campaign Monitor, Facebook Ads, Google Ads, Klaviyo, and Mailchimp.

Database tables

You can use any database table that is available to the Segments tab to build attribute groups. All of the tables in the currently-selected database are available from the drop-down menu in each attribute group and are shown in the bottom right corner of the Visual Segment Editor.

Important

Tables must be configured to be available to the Segments tab. This must be done by a member of your team who manages databases and tables from the Customer 360 tab.

1. You may change the selected database by choosing another one from the drop-down menu.

2. The list of tables is refreshed to show the tables in that database.

Attribute groups

An attribute group is a series of attributes in a database table that are used together to define a segment. You may use as many attribute groups as the segment requires. Each attribute group must specify a single database table. Multiple attributes within the same attribute group must be assigned the AND or OR operator.

1. When you open the Visual Segment Editor only one attribute group is visible and it is assigned to the Customer_360 table.

2. You may choose another table from the drop-down menu. You must define at least one attribute (2) by selecting an attribute from the drop-down, selecting a condition, and then (if required) providing a value.

3. As you add attributes to an attribute group, additional rows appear. Continue defining attributes until you do not need anymore. If you want to remove an attribute, click the minus icon.

4. If you need more attribute groups, click ADD SECTION. This will add another attribute group. This new group will behave the same as the first one. Add as many attribute groups as necessary for this segment.

5. In the new section, select a database table.

6. Define at least one attribute using the same steps as then previous attribute group.

Note

Each attribute group may use AND or OR within an attribute group, and then also between each attribute group. See AND vs. OR for more information about how to use AND and OR within and between attribute groups.

AND vs. OR

The Visual Segment Editor builds segments using a series of WHERE clauses. Each clause defines a single attribute group. There are two locations in the Visual Segment Editor where you must specify the use of AND and/or OR operators within a WHERE statement.

1. Within every attribute group

2. Across all attribute groups

Use AND to set multiple conditions, where each condition is evaluated separately and only records that satisfy each condition can be included in the results. Attribute groups that use the AND operator will typically become smaller as the number of attributes used in the attribute group gets larger.

Use OR to combine conditions, where at least one of the conditions must be satisfied to be included in the results. Attribute groups that use the OR operator will typically become larger (or at least stay the same) as the number of attributes used in the attribute group gets larger.

Important

All conditions within an attribute group must use the same operator. In addition, the same operator must be used across all attribute groups. In both situations you have a single choice: AND or OR.

Example segments

The following segments will produce very different outcomes.

Tip

These examples were run against test data used by the Amperity training and documentation teams. Try running a similar segment against your own data (without activating it) so that you can compare and contrast these result patterns against real data in your own tenant.

This segment returns a very small number of customers (fewer than 100), most of whom are active and most of whom are reachable. The 1-year revenue is ~$2500.

This segment returns a large number of customers (more than 5000), some of whom are active and most of whom are reachable. The 1-year revenue is ~$425k.

If you add another condition–surname is like "Smi"–to both examples, the attribute group built using the AND operator gets smaller, but the attribute group using the OR operator gets larger.

Consider the differences in how each operator returns data when you are choosing which attributes to add to attribute groups. You can define multiple attribute groups that use the same database table, such as one that pulls first and last names and another that groups by location using the Customer_360 table in both attribute groups.

You can get the same results as the example that uses the OR operator by creating three attribute groups with a single attribute in each group:

Segment insights

Note

Segment insights are available when the customer 360 database contains the Unified_Transactions, Transaction_Attributes, and Transaction_Attributes_Extended tables.

Segment insights within the Visual Segment Editor include the following categories:

• The number of unique customers.

• The number of customers who have been active within the last year.

• The amount of revenue generated from these customers within the last year.

• The number of customers who are reachable.

When you start building a segment, the values for segment insights represent 100% of your customers and revenue across all categories. As you apply attributes to your segment refresh segment insights and review the the updated values.

1. Segment insights are shown when you start using the Visual Segment Editor.

2. As you apply attributes segment insights are reset. Click the Refresh button to update segment insights based on the attributes you have selected. The values for each category are updated.

SQL editor

The SQL Segment Editor is an optional interface that allows you to build an attribute profile using Presto SQL. Start with a SELECT statement that returns the Amperity ID, and then apply a series of WHERE statements to define one (or more) attribute groups that match specific conditions and values.

There SQL Segment Editor has the following requirements:

1. The only field that can be returned by the SELECT statement is amperity_id.

2. All conditions and values must be contained within a WHERE clause.

3. A WHERE clause must use one of the following operators: AND or OR.

Example segments

For example, a segment that uses the OR operator to return customers whose first name begins with “Mi”, last name begins with “Smi”, and who reside in California:

SELECT
  "amperity_id"
FROM
   "Customer_360"
WHERE
  (
    (LOWER("given_name") like '%mi%')
     OR LOWER("state") = 'ca'
     OR (LOWER("surname") like '%smi%')
  )

The following example is identical to the previous example, but shows each condition in its own WHERE clause, using UNION ALL between each clause to group the results together:

SELECT
  "amperity_id"
FROM
   "Customer_360"
WHERE
  (
    (
      "amperity_id" IN (
        SELECT
          "t0"."amperity_id"
        FROM
          "Customer_360" "t0"
        WHERE
          ((LOWER("t0"."given_name") like '%mi%'))
        UNION ALL
        SELECT
          "t1"."amperity_id"
        FROM
          "Customer_360" "t1"
        WHERE
          (LOWER("t1"."state") = 'ca')
        UNION ALL
        SELECT
          "t2"."amperity_id"
        FROM
          "Customer_360" "t2"
        WHERE
          ((LOWER("t2"."surname") like '%smi%'))
      )
    )
  )

Segment overview

A segment overview highlights the most important information about a segment, including:

• How much opportunity does this segment have?

• Which channels should I engage on?

• What is the predicted revenue for this segment?

Use the information on the segment overview to help determine the best way to initiate a marketing campaign.

Each segment overview contains the following details:

1. The number of unique customers.

2. The number of customers who have been active within the last year.

3. The amount of revenue generated from these customers within the last year.

4. The number of customers who are reachable.

5. Revenue statistics, including historical revenue and predicted revenue trends.

6. The channels through which this segment has customer opportunity.

7. Segment insight charts for customer behaviors and customer attributes. These charts have a configurable date range and up to six may be selected. Click the Actions menu, and then select Customize charts to choose up to six charts for customer behaviors and for customer attributes.

8. The list of customers who belong to this segment.