Post on 16-Mar-2018
Quantifying Knowledge
with :CRUX
James Webb
Group Product Manager, FT.com
About :CRUX
A technology company dedicated to Quantifying Knowledge.
:CRUX measures how much users know based on what they read,
and recommends the best content to increase their knowledge.
:CRUX’s mission is to quantify knowledge everywhere,
for everyone.
“Stay abreast of global
business, political and
economic developments
to formulate a ‘big
picture’ view.”
“Understand
the context of a
specific issue
that is relevant
to my client.”
“Be confident that I
haven’t missed out on
interesting things my
colleagues might
know about.”
Our users #JTBD
“Be updated
about news
relevant to
my industry”
“Research a
topic or company
to become an
expert in it.”
Reading the FT
supports
work needs
People often
know what they
need to know.
But when do
they know
enough?
The plan
We’ll provide users with a live
knowledge score and
recommendations to show
them how much they might
know based on what they’ve
read, with a more compelling
onward journey.
Scenario 1:First Contact
Article Journey
KQ:The
Quantified
Knowledge
Box
First Contact,Article journey
The KQ box appears
on all relevant article
pages.
Live knowledge score
displays how much the
user knows about a
Topic, based on articles
read.
Recommendations
appear for further
knowledge-rich articles
about the topic. The
expected contribution to
the user’s KQ score is
displayed next to the
articles.
On reading this story
about quantum
computing, their KQ
score gained 10%...
First Contact,Article journey
...and a further 8% by
moving onto this story
about China’s robotics
industry...
First Contact,Article journey
...and so on.
With three article reads,
this user has moved their
KQ score to 52%.
First Contact,Article journey
Scenario 2:User returns to a previously read topic
KQ:The
Quantified
Knowledge
Box
Scenario 2:Users return to a previously read topic
...the system quantifies
and displays the
negative knowledge
impact of what they
have missed out on
since their last visit...
...and recommends a
maximum knowledge
reading list covering
new developments
since their last visit
When users return to a
topic on which they have
previously built up
knowledge...
Scenario 3User reviews topic progress over time
KQ Dashboard
Users can choose
knowledge topics to
follow in myFT, and can
track their progress
history in a central
dashboard.
KQ Dashboard
...and see the expected
benefit of reading the
recommended articles. They
also see the predicted
negative impact if they
ignore this topic further.
Users can track how
they gained and lost
knowledge in a topic
over time...
Users will receive periodic
Knowledge summary emails
analysing their progress with
a cross-topic view.
Here they can compare their
knowledge progress in
different topics directly within
the email.
And see individual
achievement highlights, as
well as comparisons to other
readers.
Scenario 4: Periodic Knowledge update Email
Exploring multiple engagement prospects
KQ: Dashboard
Ways to maintain
engagement over
time
KQ: Articles
Innovative hook
from an article
page journey
KQ: Email & Alerts
Helping users understand
their time and effort
investment while offsite
How will we measure success?
RFV: The metric that matters
Recency
How recently the subscriber visited
Frequency
How many times they visited
Volume
How many articles they read
This metric galvanises many areas of the business because it correlates
with FT subscription revenue.
Becoming engaged
leads to 10% reduction
in cancellation rates
There are two ways to increase Engagement
Recency Frequency Volume
How can we
encourage
people to make
another visit?
How can we
encourage
people to read
another article?
Revenue prospects
If we can get disengaged users to visit
one extra time every 90 days, and read
one article extra per topic they already
engage on - we believe this to be worth
up to £1.5m per year.
At the end of the project we hope to have:
● A working Knowledge Score algorithm from :CRUX
● A KQ presence on articles, a topic progress dashboard and some
experiments with alerting
● Insights into which subscriber types find the ‘knowledge’ treatment
most appealing
● An indication of in-demand knowledge topics among subscribers
Outcomes
Thank you
@jameswebb