power of predictive intelligence
TRANSCRIPT
1The Power of Predictive Intelligence
2
PurposeMarketing Operations and CRM will be implementing Predictive Intelligence (PI) for Contigo. Our purpose is to ensure your team understands the data’s engine’s value, while identifying our needs for completing this implementation.
3
What is Predictive Intelligence?
Predictive Intelligence (PI) is a data engine that collects consumer data on all channels and leverages that information to build unique consumer profiles based on a person’s behavior across those channels.
The data engine then has the ability to consolidate data obtained from all users and generate unique product recommendations to a specific consumer, based on what “look-alikes” have shown interest in.
The recommendations can then be tracked to identify the influence they had on a consumer’s engagement, giving Contigo the power to derive strategies or personas for the future.
4
We Know Your ObjectivesImplementing Predictive Intelligence, will support Contigo and CRM’s collaborative objectives of:
• Improving consumer engagement through PI’s influence of consumer experience
• Increasing current product consideration and awareness through PI’s influence of web page visits
• Expanding consumer data capture through PI’s improved consumer insights
*Note: Objectives above are based on most impactful for CRM and do not reflect all Contigo brand objectives
5
The Vision for Implementation
• Deploy quick ROI campaigns
• Leverage existing or in progress templates
• Cross-sell products• Enrich consumer
database• Reengage & Retarget• Optimize crawl
learnings
• Cross-channel Experience
• Email and Web• Activate off derived
preferences• Predictive Web
Analytics
CRAWL WALK RUN
Dec. 2016 – Feb. 2017 March – April 2017 May 2017 +
6
• Rechecks the site later
• Decides to subscribe
Day 1 Interest
2 WeeksEngage
1 WeekSubscribe
3 MonthsUnsubscribe/Purchase
No Click
Unsubscribes from future messages• Likes Contigo
• Visits the site• Checks out
water bottles on the site
• Receives Welcome• Engages with kids
items + Tumblers on web
An Improved Customer Experience
PI ENGINE
Travis
Travis’ Experience Today
Travis’ Experience Tomorrow
• Receives Welcome• Includes water
bottles + other web items viewed
• Engages with mugs + kids items + Tumblers on web
• Receives Product email on Mugs
• Doesn’t include items he’s likely interested in
Interest
• Rechecks the site later
• Sees similar water bottles from last visit + others
• Decides to subscribe
Other Recommen
d Items
Interest Other Recommend
Items
• Receives Product email on Mugs
• Includes similar water bottles + kids items + Tumblers viewed on web
Consumer Experienc
e Continues
1 MonthsEngage
Where we are now…Where we can go…
7
InterestInterest
Day 1 Interest
2 WeeksEngage
1 WeekEngage
3 MonthsContinued Experience
• Likes Contigo• Visits the site• Checks out
water bottles on the site
An Improved Customer Experience
PI ENGINE
Travis
• Receives Welcome• Includes water
bottles + other web items viewed
• Engages with mugs + kids items + Tumblers on web
Interest
• Rechecks the site later
• Sees similar water bottles from last visit
• Decides to subscribe
Interest
• Receives Product email on Pitchers
• Includes similar water bottles + kids items + Tumblers viewed on web
Consumer
Experience
Continues
1 MonthsEngage
Other Recommend
Items
• Visits the site• Checks out
water bottles + Mugs on the site as well.
• Looks similar to Travis.
• Receives next promo • Focused on mugs;
recommends Tumblers + water bottles.Tiffany
Other Recommend
Items
• Engages with kid’s items on site
Other Recommend Items
Note: Predictive Intelligence tracked Travis’ and Tiffany’s engagement and behavior through their cookies, and built recommendations along their journeys, based on that collection of data.
Consumer
Experience
Continues
Implementing the “Look-Alike” Consumer
8
Contigo Catalog
Imported into ExactTarget
Includes:• Product Attributes• Price• Category
PI Data Engine
Configured on brand site to collect consumer data
Collects and tracks:• Unique user behavior• Product engagement • Views, purchases, carts
Travis’ Profile Data
Built automatically based on user behavior
Email engagement• Opens, clicks, forwards, time in
email, etc.Web engagement• Views + content engaged, time
on page, searched items, etc.Purchase behavior and history
Logic Build
Consolidates Travis’ data + Other Similar Consumers
Others Like You • “Users who
have bought this product also bought these products
Your Recently Viewed• “Other products
you may like”
Recommendation Lists
Determined based on brand’s desired scenario
Ex:• Bought Bought• Recently Viewed• Top Sellers
How the Work is DonePulled from the catalog, the collection, the build, the recommendation…
“Look-Alikes” Profile Data
Consolidates data captured from other similar consumers
Email Web
Travis’ Recommendation
GenerationBuilt to increase consideration and
drive trial and/or purchase
9
DEC JAN FEB MARCH APRIL MAY
2017
ABANDON CART
ABANDON BROWSE
TIME TO UPGRADE
ON SALE
BACK IN STOCK
PRODUCT/PROMOBIRTHDAY SERIES
ANNIVERSARY SERIES
RATE & REVIEW
WELCOME
REENGAGEMENT
TEST, LEARN AND OPTIMIZE
NEWSLETTER
PRED
ICTI
VE IN
TELL
IGEN
CE C
AMPA
IGN
WALK – WEB IMPLEMENTATION #1 RUN – EMAIL IMPLEMENTATION #2CRAWL – EMAIL IMPLEMENTATION #1
A Best Practice ApproachThree phases of a staggered implementation - Crawl, Walk and Run
PRODUCT
CATEGORY
CART
CONFIRM
HOME
Note: Staggered approach allows CRM to maximize value and opportunities for an early return on investment through three revenue based triggered email deployments.
2016
10
Detailed Email Deployment ProcessPredictive Intelligence Email Timeline
Work streams Week 1 Week 2 Week 3 Week 4 Week 5
Contigo
Deployment
Campaign Template Development
HTML Development (Min. 2 Resources)
Implementation & Quality Assurance
Finalization & Go Live
Phase Checkpoint
To Be Confirmed
5 Days
5 Days
4 Days
Implementation Notes:According to these implementation steps, deployment should take a minimum of 21 days after the initial creative kickoff discussion amongst teams. (Discussion is not reflected on this plan)
*Risk: Deadlines are aggressive, for each day an action item is delayed, an extra day will need to be added.
***Requests: Contigo team has requested 14 days for campaign template development in order to complete internal reviews. That is not reflected on this plan.
Checkpoint Notes: Checkpoints are broken into thirds of the implementation, with the third being the “Go Live” of the email deployment.
7 Days
Note: This deployment leverages existing templates. Following these templates and substituting modules based on the email context, enables us to limit costs, time and resources necessary for deployment within this timeframe.
**Assumptions: Resources are 100% dedicated to this deployment.
11
Measuring Performance and KPIsOur plan for tracking and measuring success, based on best practices.
*Goals will be measured once all web tracking is in place. **Goal is measured against the eight fields Contigo is currently capturing data through the preference center, and will be based on derived PI insights.
Note: Influence refers to a user clicking on a PI recommendation (or reference item in triggers).
To Be Confirmed
Increase current product consideration and awareness
Expand consumer data capture
• Increase influenced site visits by 3% by end of Q32017*
• Improve recommendation clicks by 2% by end of Q32017*
• Increase of total orders influenced by 10% by end of Q32017*
• Increase influenced revenue to 15% by end of Q32017*
• Increase consumers profiles captured by 5% Q32017.*• Increase customer data capture by 20% by end of
Q32017**• Capturing these new fields:
• Unique consumer affinities• Unique consumer product views
Industry Standard Benchmarks
Improve consumer engagement
Brand Objectives
Note: Objectives above are based on most impactful for CRM and do not reflect all Contigo brand objectives
12
What We Need From YouTo Be Confirmed
• Excel file of most current Brand catalog• Continued partnership with current allocated resources
• Weekly collaboration meetings throughout on-boarding• Modify existing templates for proposed new campaigns• Web implementation ONLY: Access to eBiz or Meanbee resources
• Alignment on recommendations lists and Logic Build• Essential Brand business rules and requirements• Education on Brand limitations for product recommendations
13
Steps to Getting StartedBegin Catalog Migration Feed – In Progress for Contigo – Approximately 3 weeks
• Imports brand products into PI extension and stores product data on the site• Brand submits to Salesforce Partners
Embed PI Engine on Site - In Progress for Contigo – Approximately 4 weeks • Configured on the site to begin collecting and tracking the users behavior, includes purchases made
Begin Data Collection – 30 Days after Engine is Embedded –Ongoing Process**• Merges data collected from the PI Engine with catalog information to start generating the data
recommendations
Recommendation Configuration and Logic Build* - Timetable TBD• Recommendations can then be configured and added to brand emails and brand site**• Executed by brand resources
*Note: This process will be completed in collaboration with the brand team in order to meet campaign objectives.
**Note: Results may take up to 30 days to formulate. As the Engine acquires data over time, recommendations improve and become more unique.
To Be Confirmed
14
Appendix