Fear the fRaNkenSTaCk! How to automate, orchestrate & personalise your entire customer lifecycle
1. Fear the fRaNkenSTaCk
How to Automate, Orchestrate & Personalise
Your Entire Customer Lifecycle
Ed Fry @hull
2. SaaStock attendees have an
average of 12 tools
That’s excluding tools out of the page. Your CRM. Your database. Server side tech.
3. MY GOAL
This will make you think.
Yeah, it’s right after lunch. But I’m intolerant to boring talks too. I work at Hull. We help SaaS startups manage their
customer data chaos. It fascinates me. I have 114 slides, tons to tell, and just twenty minutes. Let’s go!
4. We’ve had customer data for a while. This was the first ever Salesforce website.
5. We’ve had Salesforce since 1999.
We’ve had Salesforce since 1999, Mailchimp since 2001, HubSpot since 2005…
6. What’s new about
customer data?
Why is customer data suddenly a “hot topic” for SaaStock? What’s changing?
7. Three trends b r e a k your data
(because of three little lies.)
There are three trends which are taking your customer data and breaking it apart. And the force behind these
trends are how we are working differently. And how we lie to ourselves. So let’s unpack these three little lies...
8. THE FIRST LIE
“We’re customer centric”
Of course, every founder, investor, and team want to be “customer centric”. However...
9. Who has a view over the
whole customer journey?
Do you know your customers? Everything about them from their first website visit through years of subscription?
10. When your startups was founded, you probably knew every customer - like a small, local cafe or bar. Your CRM
was your brain. But in real life scaling startups, the go-to-market teams aren’t organised this way.
FOUNDERS
SALES + MARKETING
11. MARKETING
SALES
CUSTOMER
SUCCESS
When you hire in a startup, you hire specialists. Marketing. Sales. Customer Success. But each of these see only
part of the entire customer journey. The full story gets scattered across these separate, specialist teams.
12. MARKETING SALES
CUSTOMER
SUCCESS
Each of these teams builds their own support function to scale and optimise their processes. But they build silos.
MARKETING OPS SALES OPS SUCCESS OPS
Your teams are siloed.
13. MARKETING SALES
CUSTOMER
SUCCESS
Instead of working together, they lasso together the data and tools they need. There’s no bigger picture. There’s
no strategy. They’re what some teams call “cowboys”. (I realise that’s sexist. Everyone can try and lasso data.)
MARKETING OPS SALES OPS SUCCESS OPS
Each team only looks after their own.
14. “Cowboys in the Wild West!” This is my favourite expression I’ve heard for explaining data governance. You are
not customer centric. Your go-to-market operation is team-centric. How do you re-align siloed, specialist teams?
You are team centric.
(not customer centric)
15. MARKETING SALES
CUSTOMER
SUCCESS
Revenue operations combines sales and marketing operations. It aligns teams for landing new deals and
accounts. But, it still leaves all the rest of the customer journey out of the picture after the first purchase.
REVENUE OPERATIONS SUCCESS OPS
Some teams run revenue operations
16. MARKETING SALES
CUSTOMER
SUCCESS
How do you optimise for the accounts that are easiest to grow, have the lowest churn, and maximise lifetime
value if customer success is out of the picture?
REVENUE OPERATIONS SUCCESS OPS
Some teams run revenue operations
17. MARKETING SALES
CUSTOMER
SUCCESS
Customer operations provides the intelligence across the entire customer lifecycle. It optimises for the best-fit
accounts and lifetime value at every interaction, across every team and tool, over the entire customer journey.
CUSTOMER OPERATIONS
Break every silo with customer operations
18. Customer Ops provides the
software and metrics for the
entirety of the customer journey.
“
Tomasz Tunguz at Redpoint has talked about customer operations before. Instead of individual teams optimising
for individual team metrics, optimising for the whole company’s metrics like customer lifetime value.
19. Customer Ops provides the
software and metrics for the
entirety of the customer journey.
“
This is the important bit - to understand the whole customer journey. It is the job of customer operations to build
that complete picture from all the data for every team and tool. So let’s talk about tools.
20. THE SECOND LIE
“We have a working
marketing stack”
Teams have the feeling that every tool in their arsenal is deliberately chosen, functioning well, and providing value.
21. You have abundant choice.
It’s like a pharmacy. Marketers can choose from a bewildering array different tools to cure all manner of ills.
22. In the past six years, we’ve seen massive growth in the number of these tools - from 150 to over 5000 since 2011.
150 to 5000 tools
in six years
23. We live in the age of the
Martech 5000
This year, Scott Brinker unveiled the Martech 5000 supergraphic. How does this make you feel?
24. It's no wonder "Actionable and Valuable
Data" has become such a high concern
for marketers. How can you know what
to act on when you are drowning in this
mess?
“
I shared the Martech 5000 Supergraphic on LinkedIn. I found others were similarly alarmed...
25. That picture makes my
head hurt.
“
This response was my personal favourite...
26. Marketers hoard tools.
(And they don’t use them together)
Teams can hoard tools faster than ever, without thinking how their system of tools is supposed to work together.
Like overlapping live chat, syncing email unsubscribes across 5 email tools, scattered leads management etc...
27. Data is fuel
Invest more in the engine.
“
Scott Brinker at Chief Martec says it’s like assembling an engine. But who knows how to do that? It reminds me to
the haphazard engines strapped to long-tail boats in Thailand. It “works” but looks like it could break anytime!
28. Duct-tape does NOT save you.
By the time we realise it’s a problem, we’re reaching for duct-tape #cowboys. But, your crazy spreadsheets, hacks,
and scripts to clean and stick data together isn’t 100% reliable. You waste a ton of time wrestling with crap data.
29. Duct-tape not working?
Throw dollars instead.
As you scale, and wasted time starts to cost you, you’ll start throwing real money at it instead. An assortment of
“all-in-one” platforms, real engineering hours on integrations, and a customer data warehouse that’s never in sync.
30. You have a fRaNkenSTaCk
The result is a frankenstein-style abomination. A haphazard combination of tools, siloed data, and frustrated
teams. Teams who don’t have the data they need in the tools they use. Teams can’t do their jobs effectively.
31. Just imagine if your finance
team acted like this!
It would be unthinkable to have your payroll broken, your investment be the wrong number or your billing system to
duplicate and fail. If finance aren’t allowed to get away with this, then why should you?
32. You need something to organise your tools
Instead of a complete mess of “everyone and everything trying to control everyone and everything else” you need
a way to organise your tools and data flows. There’s something missing in the middle.
Something missing here?
33. Think of a “hub-and-spoke” integration network
Instead of building integrations between everything (n permutations), build one set of integrations between a
central hub for customer data. Your CRM, a customer data warehouse, or a real-time customer data platform.
34. There is no “single” source of truth.
There should be no “single” source of truth. Every tool & team should get the “truth”, the whole truth & nothing but
the truth, fed from a central hub with data they can trust. You get the truth. You get the truth. You all get the truth!
35. What makes a system of
intelligence valuable is that it
typically crosses multiple data
sets, multiple systems of record.
“
This takes thinking about a system, not just many individual tools.
Jerry Chen at Greylock has talked about “systems of intelligence” before.
36. What makes a system of
intelligence valuable is that it
typically crosses multiple data
sets, multiple systems of record.
“
And this is the key point. That a system of intelligence can cross multiple sets of your customer data.
37. THE THIRD LIE
“Our data is usable”
We’ve discussed the cross-functional “customer operations” team to break siloes & draw alignment, & centralised
“system of intelligence” to sync tools together. But what about the data itself? Your data is ready to use, right?
38. Your data is messy.
You’ve duplicate contacts. Form submissions are incomplete. Your analytics tracking is broken. Someone keeps
“dumping” data in your CRM. Someone else set everything up perfectly, then left. Your data is FUBAR.
39. No one tool can digest all your data.
Even if all your data was perfectly clean, it wouldn’t fit in every tool. Analytics. CRMs. Databases. Email tools. Ad
networks. All have different ways of storing and using data. You have to compensate for each tools inadequacies.
40. Personalisation is a data problem.
(Not a marketing problem)
But we have to digest more data than ever. Personalisation is how you cut through noise.
But you need to have data on each person and account to be personal. I’ve a 30 minute deck you can view here.
41. Data keeps pilingup!
You can’t just “sort it later” because data keeps piling up. More tools. More accounts. More complexity. More
duct-tape. The mess is getting more unusable.
42. How is your “engine” doing?
Back to Scott Brinker’s analogy earlier, what happens when you feed this mess into your engine?
43. You have data puke.
You don’t have “fuel”. You have data puke. Your system of tools cannot digest all this messy data. And without
immediately useful data, none of your tools or teams can piece together the full picture of every customer.
44. From a customized
data puke to getting
insights that drive
action which will have
a business impact.
“
I’m not the first to talk about “data puke” at all. Avinash Kaushik notably talks about data puke vs. insights in
comparing web reporting vs web analysis. One is a chart (sometimes a pretty chart). The other is useful insight.
45. From a customized
data puke to getting
insights that drive
action which will have
a business impact.
“
He looks at it from an analytics perspective. How can we find “insights”. But to scale, automate, and personalise,
we need to create rules, processes, and actions with data - we want to make our data actionable. How?
46. First, combine your customer data in one place.
Combine your data into one place. Like with integrations, most SaaS companies we’ve talked to choose to
combine their data into one of three places - a CRM, a data warehouse, or a customer data platform.
47. Next, connect data about each person
Once you have all your data in one place, you need to match up the different bits of the customer journey from
each tool using email addresses and IDs. You want the “full story” for each person and account - a master profile.
48. Raw data is not immediately useful
But your data isn’t always ready-to-use in all of your tools. It needs processing. Turning into something else first...
49. Compute new data before syncing.
Data cleansing. Data enrichment. Segmentation. Scoring. Transformation. Qualifying. All computations and
actions that need to happen before data is useful in all your end tools.
50. Now sync ready-to-use data to your tools
Now you can feed the beast. Immediately useful, actionable, clean, digestible data to all your tools.
51. How do you use this together?
With “customer operations” you have aligned teams. With a “system of intelligence” you have a network of tools.
With clean, centralised, ready-to-use customer data… what can we do with it?
52. Don’t use this as a
mental model.
The problem with customer data is there’s a lot of moving parts. It’s not just about aligning sales and marketing.
It’s not just about integrating Intercom and Salesforce. It’s about the whole system of tools, teams, and data.
53. Welcome
to my brain.
How do you make your tools, teams and data work all together? Some call it orchestration. I want to share a
simple model for thinking about customer data orchestration that you can take away and use
60. This forms the a
closed loop.
It is actionable!
And we can react!
Action
Messaging
Tracking
Profile
61. Not all data is
ready-to-use.
Messaging
TrackingAction
Profile➔ Gasp! “Data puke”
➔ Can’t sync data as is
➔ Wrong format
➔ Missing data
➔ Needs processing
63. Sync it back
to the profile.
Action
Profile
Compute
➔ Cleansed
➔ Enriched
➔ Segmented
➔ Scored
➔ Transformed
➔ Qualified
Tracking
Messaging
64. Use clean data
everywhere.
Profile➔ Messaging, email, live chat
➔ CRM
➔ Ads
➔ Website
➔ In-product
➔ Analytics & intelligence
Action
Messaging
Tracking
Compute
65. Use the
customer stack
to explain
orchestration as
closed loops.
Profile
Compute
Tracking
Messaging
Action
66. Let’s look at three orchestrations...
I want to walk you through three examples of orchestrations with the closed loop.
67. RECIPE #1
Personalised onboarding
This is familiar to most of you. Onboarding is key to activation, and personalisation helps us deliver the most
relevant message.
74. React to user
actions and
update our
personalised
campaigns.
Profile
Action Tracking
Compute
Messaging
75. RECIPE #2
Product qualified leads
Many startups, including many at SaaStock, have a free trial based sales model. Whilst this is great for getting
signups, it’s hard for sales to turn a firehose of (often junk) leads into actionable sales leads. Enter PQLs...
76. PQL outreach grew
account value by 30%
Using product usage data to direct sales to the right accounts and cue them with what to say, DigitalOcean this
drove an average 30% increase in account value across a cohort of their 500,000 customer accounts
77. Product qualified leads eliminated 30%
of leads from Appcues’ sales funnel
You don’t need to be DigitalOcean’s size and scale. Working with Appcues, we cut 30% of their trial leads from
sales funnel, whilst also giving sales the insights and reasoning how the account was getting value for outreach.
79. Use product usage
data to inform sales
who is getting value.
Compute
Profile
Action
Messaging
Tracking
➔ Analytics tools
➔ Product database
80. But “streams” of
product usage
data are not easy
to sync straight to
CRMs. Compute
Profile
Action
Messaging
Tracking
81. ➔ Every action
➔ For every person
➔ In every account Compute
Profile
Action
And it’s
overwhelming
to sales reps.
Messaging
Tracking
82. Transform “streams”
of user actions into a
CRM-friendly,
rep-friendly insights.
Action
➔ projects_created = 24
➔ last_active = 2 days ago
➔ trial_limits_modal_viewed = TRUE
Profile
Tracking
Messaging
Compute
83. Enrich profiles with
company data too.
Messaging
Action
Compute
➔ employee_count = 500
➔ industry = software
Profile
Tracking
84. Create a real-time
segment of product
qualified leads
Messaging
Action
Compute
Profile➔ Latest actions
➔ Updated segments
Tracking
85. Sync product
qualified leads to
your sales CRM.
Action
Profile
Tracking
Compute
Messaging
➔ New PQLs added
➔ Fading PQLs removed
86. Have reps engage
each user based
on the actions
they’ve just taken
Action
Profile
Tracking
Compute
Messaging
88. RECIPE #3
The Reveal Loop
The Reveal Loop is a powerful automation for building sales pipeline in B2B.
89. RECIPE #3
aka. “10x account-based marketing”
It works by identifying and engaging best-fit accounts at scale. Some called it “10X account-based marketing”
90. Add $10,000 per day net new revenue
We’ve worked with a number of customers implementing this. One of our best results was with our partner
Segment who made this their third highest source of sales opportunities within three weeks with just one channel.
91. Most people (95%?)
never give their email
address after visiting
your website.
Tracking
Compute
Profile
Action
Messaging
92. So how can we
re-engage them? Tracking
Compute
Profile
Action
Messaging
96. Use Reverse IP lookup
to identify anonymous
website traffic.
Action
Profile
Compute
Tracking
Messaging
➔ Send IP address
➔ Return company domain
name
98. First, match against
your CRM records
or user database.
Profile
Messaging
Tracking
Compute
➔ Customers
➔ Existing sales opportunities
➔ Partners
➔ Competitors
Action
100. Then, prospect for
contacts amongst
qualified accounts
Action
Profile➔ Job role
➔ Seniority
➔ Contact details available
Messaging
Compute
Tracking
101. Create a real-time
segment of “revealed
prospects”
Action
Profile
Messaging
Compute
Tracking
➔ Every stakeholder associated
with each “revealed” account
➔ Account are eligible for outreach
➔ Account is qualified
➔ Target job role to talk to
102. Sync the real-time
segment of “revealed
prospects” to all
your tools
Messaging
Profile
Action Tracking
Compute
➔ Sales CRM
➔ Sales email cadences
➔ Website audiences
➔ Ad audiences
➔ Live chat
103. Now you can
re-engage best fit
accounts across
multiple channels
Messaging
Profile
Action Tracking
Compute
➔ Retargeting ads
= 1 person x 1 channel
➔ Reveal Loop
= many people x many channels
104. And you can react to
the entire account’s
engagement.
Profile
Action Tracking
Compute
➔ Ads clicked
➔ Website visits
➔ Email opens and clicks
➔ Live chat conversations
➔ Meetings booked and held
Messaging
106. Orchestration is like using lego blocks.
Think how different pieces can plug together.
107. Think how you can (re)use all of these pieces
➔ Reverse IP lookup
➔ Live chat
➔ Data enrichment
➔ Sales CRM
➔ Product qualified leads
➔ Sales engagement notifications
➔ Prospecting for related contacts
➔ Automated drip emails
➔ Ad audiences
➔ Transform events streams into properties
➔ 10X retargeting
➔ The Customer Stack
➔ Analytics tools
➔ Backend database
➔ Marketing automation
➔ De-duplicate accounts
➔ Ad platforms
➔ Calendar views
➔ Meetings held
➔ 10X account-based marketing
108. You can do this!
(You just need to be a bit creative)
It takes a mindset to make it work, and to solve for your tools, teams, and data - not just individual challenges.
109. What works for startups like
yours? Let’s meet 1:1 at SaaStock!
hull.io/get/saastock
Whilst I’m here in Dublin, I’d love to help you guys work through different orchestrations that might work for your
startup. I’ve a Calendly link above to book a time. But in case we don’t get time, here’s the high level principles...
110. Where is your data?
CRMs. Backend databases. Analytics tools. Marketing automation. Data enrichment. Ad audiences...
111. How do you need to transform your data?
Cleansing. Enriching. Segment. Scoring. Transforming. Qualify. Make data useful for your end tools
112. Where can you engage
leads and customers?
Email. Ads. Live chat. Sales meetings. Website. Snail mail...
113. Get the slides,
Our customer data academy,
(and book a meeting here at SaaStock)
hull.io/get/saastock
You can get the slides, our customer data academy, and grab a slot to talk about your three orchestrations
through the link above - hull.io/get/saastock.
114. Get the slides,
Our customer data
academy, and three
orchestrations
hull.io/get/saastock
Questions?
I’m @edfryed
ed@hull.io