Building a data-driven product is scary business. You need to get the right platform both for today’s needs and for tomorrow’s possibilities – and then, you need to go beyond the technical to build a go-to-market plan that will set you up for success. Learn the five keys to building a great analytical product from someone who has done it before — and failed! Hear Kevin Smith speak about the mistakes he’s made building data products and how you can benefit from his lessons learned.
9. ‹#›
A typical embedded analytics journey
Stage 3:
Excitement
Project Progress
Enthusiasm
Hey! We’ve got all this
data! Let’s build a data
product!
This platform is
awesome! Let’s
implement it!
OMG! Let’s launch
and watch the
money roll in!
Wait. Why aren’t they
buying? Whose idea was
this anyway…
Stage 1:
The Big Idea
Stage 4:
Sadness & Despair
Stage 2:
Anticipation
16. 16
Hold a Product Workshop
• 6-8 hours
• In person
• Leadership team present
• What do you want to build?
• Align your strategy & tactics
• Start to develop your pricing/tiering
model
• Pick personas
• Identify boundaries
19. ‹#›Concepts from Nir Eyal “Hooked, How to Build Habit-Forming Products,”
If this is your goal
Goal: an enticing, engaging product
that keeps you coming back
20. ‹#›
You get there by
linking users’ goals to
resolution via
analytics
21. ‹#›Concepts from Nir Eyal “Hooked, How to Build Habit-Forming Products,”
You need to
know:
• Who are you?
• What are you
trying to achieve?
• Why can’t you do
it today?
22. ‹#›
Start by
picking
2-3 key
user
personas Potential
Data Product
Users
CMO
CTO
CPO
Head of
Sales
Head of
Operations
Executive
End Users
Internal
Users
Tactical
End Users Installation
team
Sale Reps
Marketing
team
Customer
Service
reps
Your CPO
Your Head
of
Operations
Your Head
of Sales
Your
Finance/
Billing team
Your
Marketing
team
23. ‹#›
“I feel like I’m operating blind. I don’t know what campaigns are doing well and which are doing poorly until it’s way too late to take
action and correct any problems”
Salesforce.com
Marketo
Excel spreadsheets
Funnel tool
NetSuite
360 view of customers
360 view of campaigns
See gaps and coverage
See geographic overview of campaign
reach
Can’t see performance by region, team,
or campaign
Can’t perform “what if” analysis
Guesses at best course of action
Can’t see if campaigns touch customers
multiple times or if gaps exist
Key Wishes or NeedsTools Used TodayFrustrations or Pain Points
Needs to develop marketing campaigns that attract new customers and
encourage existing customers to try new products and features
Mission
Key Characteristics
The focus of all demand gen programs
A trusted advisor to the executive and sales teams
Pulled in many different directions
VP of Corporate Marketing StrategyRole or Title
Build persona cards for key users
24. ‹#›
Lay out the mission and workflows
Mission: Create high performing marketing campaigns
Workflow: review and adjust performance
Review status of
campaigns
ID
underperforming
campaigns
Determine regions
or segments
underperforming
Adjust campaign
target list or
collateral
Project revised
results
1
25. ‹#›
Then identify the pain points (gaps)
Mission: Create high performing marketing campaigns
Workflow: review and adjust performance
Review status of
campaigns
ID
underperforming
campaigns
Determine regions
or segments
underperforming
Adjust campaign
target list or
collateral
Project revised
results
Pain point:
I can't compare campaigns
to past performance
Pain point:
I can't see performance vs
target
Pain point:
I can't see benchmarks for
performance
Pain point:
I can't predict what may
result from my changes
Pain point:
I have to wait 24 hours to
see my changes reflected
Pain point:
I can't easily filter
campaigns by dimensions
Pain point:
I have to guess at the best
changes to make
2
26. ‹#›
And determine the analytics to close the gaps
Table with
filter
Clustered
bar chart
Predictive
analytics
Analytics
for this
persona
Bar/line
chart
Mission: Create high performing marketing campaigns
Workflow: review and adjust performance
Review status of
campaigns
ID
underperforming
campaigns
Determine regions
or segments
underperforming
Adjust campaign
target list or
collateral
Project revised
results
Pain point:
I can't compare campaigns
to past performance
Pain point:
I can't see performance vs
target
Pain point:
I can't see benchmarks for
performance
Pain point:
I can't predict what may
result from my changes
Pain point:
I have to wait 24 hours to
see my changes reflected
Pain point:
I can't easily filter
campaigns by dimensions
Pain point:
I have to guess at the best
changes to make
3
27. ‹#›
Repeat for all personas, missions, & workflows
Mission: Create high performing marketing campaigns
Workflow: create campaign
Workflow: review and adjust performance
Review status of
campaigns
ID
underperforming
campaigns
Determine regions
or segments
underperforming
Adjust campaign
target list or
collateral
Project revised
results
Identify target
markets
Build target list
Select collateral
to serve
View potential
results from
campaign
Launch campaign1
2
Workflow: identify new audiences
Review audience
list coverage
ID underserved
audience
segments
Adjust list to
increase coverage
Project revised
results
3
4
29. ‹#›
The Card Sort
Mission:
Understand
campaign
performance
Mission:
Compare
performance between
teams
Mission:
See if we could be
using other more
effective tactics
Use for
arranging
dashboards
1
Use for
ordering
workflows
2
Mission:
Understand
campaign
performance
Workflow:
Check overall
progress toward
quarterly goals
Workflow:
ID any outliers to
improve or replicate
Chart #1:
Bar chart showing
overall campaign
performance
Ch
Workflow:
Bar chart showing
performance /$
spent
Ch
33. ‹#›
“Throw it all in and let the
customer decide…”
is not how you build a great data product
34. ‹#›
Creating a great data product is about
having an opinion and making decisions.
35. ‹#›
Know your boundaries
Stuff We Won’t Do
StuffWeWon’tDo
The Product
Stuff We’ll Do
as Part of the
Core Product
Stuff We’ll Do
for an Extra Fee
StuffWeWon’tDo
Stuff We Won’t Do
36. ‹#›
Category Details Typical Options
Amount of data
The number of years of information available
to be presented to the user
1 year; 3 years; 5 years
Benchmarking
Allow comparisons to aggregated and
anoymized data from others
No benchmarking; internal benchmarking; external
benchmarking
Predictive analytics
See possible results from changing
parameters
No predictive; predictive allowed
Ad hoc analytics
Allow users to create new charts and other
analytics using existing metrics
No ad hoc; ad hoc allowed
Custom metrics
Create non-standard metrics still supported by
the standard model
Standard only; modify/create new metrics
Drill levels Allow drill down to record level data Drill down 1-2 levels; drill down to record level data
Data layers
Create new insights by stacking data sources
to composite a new view.
None; pre-defined; customized
What do you have to offer?
37. ‹#›
Basic
• Included for all customers
• Uses standard template
• Read-only dashboards
• No benchmarking
• 1 year historical data
Plus Pro
• Uses standard template
• Customization for an
additional fee
• Read-only dashboards
• Internal benchmarking
included
• 2 years historical data
• Uses standard template
customized per customer
needs
• Users can modify dashboards
(ad hoc analytics)
• Includes external
benchmarking
• 3 years historical data
Use tiers as a growth path for buyers
38. ‹#›
Basic
Wow—I’ve got everything I’ll ever need in this one tier!
I’ll never need to buy more! And, I’ll expect new features to be
added for free!
The wrong way to use tiers
39. ‹#›
Basic
Ok, I’m hooked
but am I
missing out on
other insights?
Plus Pro
The right way to use tiers
Customization,
comparisons, and
more data but I
want more…
I’m set — ad hoc
analysis, more
data, and detailed
benchmarking
42. ‹#›
The four numbers you need to understand
1
2
3
4
The cost to build your product WITHOUT analytics
The cost of the analytics (platform, services, etc.)
The current price you charge for your product
The price your best competitor with analytics
charges
43. ‹#›
Here’s how you price a data product
Potential Pricing Zone
Caution
Zone
Danger
Zone
Caution
Zone
Danger
Zone
Price your data product along this range
Cost of Goods
Sold —
WITHOUT
analytics
Cost of Goods
Sold —
WITH analytics
Current
Product Price
Price of Best
Competitor
Basic Plus Pro
46. ‹#›
This is the foundation of your product.
Core Product
BI FunctionalityWhat companies
spend their time
on…
Operations
Provisioning
Training
Support
Marketing
Billing
ChangeManagement
What they
forget…
(until it’s too late)
48. 48
Walk the
processes
to find gaps
• Design the process
• Print & lay out the
steps
• Walk through
scenarios
• Note hand-offs, gaps,
waits, etc.
49. 49
Set trip
wires
• Use key launch
metrics
• Set “do not cross”
thresholds
• Know what to do if
crossed
• Do it beforehand
• Monitor
50. 50
Establish a
command
center
• Have representative
from all key
stakeholder groups
• Establish who is in
charge
• Regular reviews
• Monitor trip wires