1. The document discusses next-best-action personalization using an API-first approach to predict customers' next actions based on their context in real-time.
2. It proposes an API-first hyper-personalization stack that is fully programmable, combines real-time and batch data, and is multi-tenant and self-service.
3. The stack includes components like event-based models, score combiners, customer journey modeling, and segmentation to power use cases in retail and hospitality.
3. Retail next-best-action scenarios
3
consider &
evaluate
shop &
book
returninstore purchase
“Remind me
what I usually
buy”
“Open
Wishlist”
“What else
should I do?
“How do I get
there?”
Skip the checkout
line
“What are my
friends buying”
“Relevant
Reviews ?”
“Special Store
Deals?”
“Pickup in
Store”
“Alert me if
there’s a
special today”
“What else
should I
instead”
Recommend
other products
to purchase
“What new
products are
for me ?”*
“Any new
deals”*
4. Hospitality next-best-action scenarios
4
consider &
evaluate
shop &
book
go return
pre-
departure arrive stay
“Alert me when
my flight
changes.”
“Let me
upgrade my
room before I
arrive”
“Where should I
go eat?”
“How do I get
there?”
“Allow me to digitally
communicate with the
hotel, and alert me that
my room is fixed”
“Allow me to
digitally check-
in”
“Help me rent
a car”
“Notify me
when my room
is ready”
“I want to pre-
order room
service”
“Alert me if
there’s a hotel
special today”
“Enable me to
digitally check
out”“Access to
more video
entertainment
choices in my
room
“I want to use
my points to
book”*
“I need to book
a hotel”*