View insights and observations

  • Updated

Optimizely Data Platform (ODP) automatically assesses insights and observations for an account using available underlying data (like customers, products, orders, and page views). ODP evaluates the data nightly and makes it available for standard segments, reports, and custom Liquid building. For real-time evaluation based on events, use filters and behaviors instead.

Availability

In standard segments

Insights and observations available for a customer's attributes like their country, topic of interest, and likelihood of ordering can help you build a standard segment in Customers > Segments.  

In custom report columns (rocket columns)

  1. Go to Reports > Custom Reports.
  2. Select your desired report or click New Report
  3. Expand the Column(s) menu to select the columns you want in a report. Click Apply to add the column data to the report.

In custom Liquid

You can reference the insights and observations in Liquid syntax when building out campaign touchpoints. Use the following prefixes with the Liquid suffix according to the category:

  • Customer observationscustomer.observations.suffix (for example, customer.observations.order_count)
  • Customer insightscustomer.insights.suffix (for example, customer.insights.order_likelihood)
  • Product observations
    • Customer – customer.observations.product.observations.suffix where product can be first_ or last_ (for example, customer.observations.first_product.observations.repeated_sale_ratio)
    • Product grid – product.observations.suffix (for example, product.observations.repeated_sale_ratio)
  • Product insights
    • Customer – customer.insights.product.insights.suffix where product can be first_ or last_  (for example, customer.insights.first_product.insights.up_sell_category)
    • Product grid – product.insights.suffix (for example, product.insights.up_sell_category)
  • Order Observationsorder.observations.suffix (for example, order.observations.order_number)

Proper use of percentiles

Observations and insights based on percentiles use a range of 0 to 100, where 100 is the highest percentile. For example, a value of "At Least 90" when building a segment with Order Count Percentile includes customers with more orders than 90% of the customers in the account, meaning the top 10% of customers by Order Count.

Customer information

Customer observations

Name Liquid suffix Description
Order count order_count Number of orders purchased.
Order count percentile order_count_percentile Percentile by order purchased.
First order date and time first_order_ts Date and time of first observed order.
Last order date and time last_order_ts Date and time of last observed order.
First product ID first_product_id Product ID of the highest-priced product in the first order.
Last product ID last_product_id Product ID of the highest-priced product in the last order.
Total revenue total_revenue Sum of the value of all observed orders (net).
Total revenue percentile total_revenue_percentile Percentile by observed orders.
Average order revenue average_order_revenue Average revenue across observed orders (net).
Average order revenue percentile average_order_revenue_percentile Percentile by average revenue.
First visit date and time first_visit_ts Date and time of first observed page view.
Last visit date and time last_visit_ts Date and time of last observed page view.
Session count session_count Number of observed sessions.
Session count percentile session_count_percentil Percentile by observed sessions.
Email domain email_domain The domain name of the email address.
Mailbox provider mailbox_provider The organization that provides the email box.
Discount order ratio discount_order_ratio The ratio of orders that involved a discount (line item level, order level, or discount code).
Discount order percentile discount_order_percentile Percentile by discount order ratio.
Discount usage discount_usage Restatement of discount order ratio into "Always," "Sometimes," or "Never," based on the ratio of orders with a discount.
Acquisition Source acquisition_source Same as first_source, when populated, otherwise filled in as "unknown".
First source first_source UTM source of the first event.
First medium first_medium UTM medium of the first event.
First campaign first_campaign UTM campaign of the first event.
First event date and time first_event_t Date and time of the first event.
One-time send count (last 365 days) one_time_send_count Number of one-time sends in the last 365 days.
One-time click rate (last 365 days) one_time_click_rate Click rate (based on sends) of one-time sends in the last 365 days (in decimal, that is, 0.05 is 5%).
Last seen subscribed subscribed The marketing consent status for the most recently seen email address (True or False).
Last seen email status email_status The deliverability status for the most recently seen email address.

Customer insights

Name Liquid suffix Description
Engagement rank engagement_rank Percentile engagement for each customer based on the amount of time spent with your brand, weighted to more recent days. A measure of "headspace" occupied, is correlated to future engagement.
Order likelihood order_likelihood Likelihood of ordering in the next six weeks.
Days until next order ttno_days Estimated number of days until the next order, based on those likely to order in the next six weeks.
Winback zone winback_zone Customer's designation as "Engaged," "Winback," or "Churned".
Winback type winback_type Customers with the value "New" entered "Winback" during the latest model run (evaluated nightly).
Probability still a customer probability_alive Customer's liveness probability from the churn prevention model.

Product information

Product observations

Name Liquid suffix Description

Bestseller rank (seven days)

bestseller_rank_7_day The sales ranking by units sold over the last seven days calculated at the parent product level, if applicable. "1" is the highest rank.
Bestseller rank (30 days) bestseller_rank_30_day The sales ranking by units sold over the last 30 days calculated at the parent product level, if applicable. "1" is the highest rank.
Detail views per day views_per_day The average product detail page views per day, weighted to more recent days. Calculated at the parent product level, if applicable.
Detail views per day percentile views_per_day_percentile Percentile by average product detail page views per day.
Discounted sale ratio discounted_sale_ratio The ratio of product purchases sold on discount to all product purchases, weighted to more recent days.
Discounted sale percentile discounted_sale_percentile Percentile by discount sale ratio.
First sale ratio first_sale_ratio The ratio of product purchases that were first-time buys to all product purchases, weighted to more recent days.
First sale percentile first_sale_percentile Percentile by first buy ratio.
PDP views per day pdp_per_day Average product detail page views per day, weighted to more recent days.
PDP views per day percentile pdp_per_day_percentile Percentile by average product detail page views per day.
Revenue per day revenue_per_day The average revenue per day weighted to more recent days.
Revenue per day percentile revenue_per_day_percentil Percentile by average revenue per day.
Revenue rank (seven days) revenue_rank_7_day The sales ranking by revenue sold over the last seven days calculated at the parent product level, if applicable. "1" is the highest rank.
Revenue rank (30 days) revenue_rank_30_day The sales ranking by revenue sold over the last 30 days calculated at the parent product level, if applicable. "1" is the highest rank.
Sales per day sales_per_day The average sales per day weighted to more recent days.
Sales per day percentile sales_per_day_percentile Percentile by average sales per day.
Repeated sale ratio repeated_sale_ratio The ratio of product purchases that were a repeat purchase of the product, weighted to more recent days.
Repeated sale percentile repeated_sale_percentile Percentile by repeated sale ratio.
Views before purchase views_before_purchase The average product detail page views before the first purchase of this product, weighted to more recent days. Calculated at the parent product level, if applicable.
Views before purchase Percentile views_before_purchase_percentile Percentile by average product detail page views before a first purchase.
Viewed Rank (seven days) views_rank_7_day The product detail pageview ranking over the last seven days calculated at the parent product level, if applicable. "1" is the highest rank.
Viewed rank (30 days) views_rank_30_day The product detail page view ranking over the last 30 days calculated at the parent product level, if applicable. "1" is the highest rank. 

Product insights

Name Liquid suffix Description

Up-Sell category

up_sell_category Category of products that are frequently purchased with this one.
Cross-sell category cross_sell_category Category of products that are frequently purchased after this one.
Replenishment days replenishment_period_days The average number of days before a reorder of the product.

Order information

Order observations

Name Liquid suffix Description
Order number order_number The order number that the behavior is associated with.