Plan the growth

More relevant means more revenue

Transcript

1. Example e Commerce buy the outcome, not the exercise
2. The challenge “The consumer profile is not clear due to the fact that multiple different datasets are working independently from each other” Consumer Unknown CRM, Newsletter Opt-ins, Consumer Behaviour (webshop), Google Analytics
3. The situation Consumer Questions “When are we communicating this?” “Who is the customer?” “What are we going to tell the customer?”
4. The situation Consumer Fact “We can’t develop effective and relevant communication needed for growth hacking without knowing the when, who and what”
5. The situation Consumer Strategy “We connect with one or more types of data to clarify the consumer, to find out when the best time of communication will be and learn to recommend the most suitable item or content at that specific time frame”
6. Properties “Are consumers finding what they want, or are they having the type of experience that encourages them to take the next step and become a customer?” Relevance
7. Properties “In growth marketing, relevance means understanding a consumer’s behavior across all digital channels and then being able to use that intelligence to take action in real-time by offering the most relevant and personalized experience to consumers” Intelligence
8. Providing the fix Measure Segment Training Model Analysis ++ relevance == more revenue
9. The chase “ Machine Learning (data mining, data analyse) 1. Searching for patterns within data 2. On the basis of these patterns produce a model that describes the data 3. Provide better decisions as a result of this model” Training model
10. The chase “Collect data on user behavior and activity from all the visitor interaction channels, including websites, social media and mobile devices. Final focus lies on the e-commerce instances” Measure
11. The chase “After data is collected, analyze it to obtain a holistic view of customer behavior across all channels and over time” Analysis
12. The chase “Visitors who share one or more preferences or characteristics” “Segmentation Division of visitors into groups (or segments) that are then tested or used against defined objectives” Segment
13. The situation Questions Revisited “When are we communicating this?” “Who is the customer?” “What are we going to tell the customer?” “Men 20-30 years old best time is 19:45 AM weekdays” “Biggest spenders are single female 20-30 years old” “We recommend items based on the patterns and decissions”
14. Outcome “Understand what works and what doesn’t, the interrelationship and the chases between the touch points, and the ability to easily take action on this understanding” Real-time relevance is king
15. Real Life buy the outcome, not the exercise
16. One 1. Input (webshop) 2. Mechanical Turk(refresh and collect) 3. Process (train model with conversions) 4. Update (decisions and scoring) Plan your growth
17. Two “the more data we collect and feed into and from a consumer during the training of the model, the more relevant the bucket will become over an period of time” relevance conversions
18. Three “based on experience we create a 0.3% raise of conversions the first year” uid: 12345 ctr: 0.053 ctr his: -0.002 kscore: 32 items:{iid 12345, iid 12346} uid: 12346 ctr: 0.173 ctr his: +0.021 kscore: 48 items: {iid 1245, iid 12346, iid 33221, iid 223344} uid: 12347 ctr: 0.233 ctr his: +0.035 kscore: 51 items: {iid 12345, iid 12346, iid 443325, iid 33221}
19. Outcome “happy faces because we identified the right signals and filtered all noise resulting in a more relevant recommendation“