This is from a very long time ago.
Shared as part of this post: https://medium.com/@talraviv/tearing-down-the-teardown-incl-personal-example-2d8c5ce2dce6
3. • Modeling this off of teams I’ve worked on and
how we adopted new tools (admittedly all
tech, which is not great)
• Ideally I’d spend a good amount of time
learning about this qualitatively (phone chats,
time with support, user research…)
• You’ve watched The Office
🚨 Assumptions
4. • Excited about life with slack, first to register
• Will put in extra effort to help adoption
Champion
• Open to trying new things, next to sign up
• Positive, but won’t make exceptional effort
Why Nots
• Risk-averse, let others dive in first
• Neutral at best, need to feel motivation
Laters
6. • At least 10 total messages exchanged
by 3+ unique users in 24 hours 🚨
• KPI would be % of teams with at
least four invitees to reach that in a
cohort
🚨 Assumption! Ideally I’d take today’s retained teams, go back in
time, ask what typically happens in the first 24h (or, in what frame of
time do we see 10, 20, 30 messages written?)
What is a “meaningful convo”?
7. Champion invites at least four people (given)
Champion writes in a public channel
Two “Why Nots” write in a public channel
Exchange ten messages within 24h
What are the funnel steps?
9. • No one to say anything to yet/empty
room/blank canvas
• Expect adoption to “just happen”
because everyone sees what they do
Why wouldn’t champion write in a
public channel?
10. • Not in a public channel by default
Why wouldn’t the champion write
in a public channel? (con’t)
End of current
onboarding leaves
me in DM with
slackbot
11. • Nothing [interesting] to respond to
• Not in a public channel by default
Why wouldn’t two Why Nots write
in a public channel?
12. • Conversation not catching
momentum, not feeling the real-time
Why wouldn’t two Why Nots write
in a public channel?
Subject could be
more human
Could use a clear
call to action
Ideally highlights real
people talking
13. • Conversation not catching
momentum, not feeling the real-time
• Conversation not interesting or
engaging
Why wouldn’t three users reach 10
messages?
14. • Unsure why slack is necessary
• Need external credibility/social proof
• Need to see teammates using it
• Don’t know how to use it
• Champion fears the “Blank Canvas”
🚨 These may well apply to “Laters,” but
assuming they’re not relevant to “Why Nots”
🚨 Non Hypotheses
16. May improve if:
• Let Champion in on what we know about
successful adoption
• Help Champion use their energies to more
directly evangelize to Why Nots
Hypothesis: Champion expects
team to magically adopt slack
17. May improve if:
• Direct the end of onboarding to #general
Hypothesis: New users start in DM
thread instead of public channel
18. May improve if:
• Provide Champion examples of easy yet
meaningful conversation starters (that
make sense even if no one arrived yet)
Hypothesis: Initial conversation not
interesting enough to catch
Hypothesis: Champion has no one
to write to
19. May improve if:
• Use email to more assertively highlight the
early action on slack
• Put stronger emphasis on enabling desktop
notifications
• Put more emphasis on downloading the
desktop/mobile clients
Hypothesis: Conversation not
catching momentum
21. 🏆 = if this works, it’ll be impactful
⚡ = low cost to try (or at least test)
🌲 = likely to work
🚨 Would definitely want to pulse the
different people on the team for any
historical knowledge on all three fronts
What makes an experiment worth
trying?
22. • Let Champion in on what we know about
successful adoption
• Help Champion use their energies to more directly
evangelize to Why Nots
• Direct the end of onboarding to #general
• Provide Champion examples of easy yet
meaningful conversation starters (that make sense
even if no one arrived yet)
• Use email to more assertively highlight the early
action on slack
• Put stronger emphasis on enabling desktop
notifications
• Put more emphasis on downloading the desktop/
mobile clients
⚡
🏆
⚡ 🌲
🏆 ⚡ 🌲
🏆 ⚡ 🌲
🏆 ⚡ 🌲
🏆 ⚡
Rating each experiment
🚨 Probably the most assumptions on one slide in this presentation
23. Winners
1. Provide Champion examples of easy yet
meaningful conversation starters (that make
sense even if no one arrived yet)
2. Use email to more assertively highlight the
early action on slack
3. Put stronger emphasis on enabling desktop
notifications
🏆 ⚡ 🌲
🏆 ⚡ 🌲
🏆 ⚡ 🌲
25. 1. Give conversation starters
We want the first conversation a team has on slack to be
meaningful and involved.
We want to encourage people to talk about topics that: 1) are
relevant to all types of teams, 2) don’t quickly expire, 3) can
discuss before everyone gets on slack, 4) anyone on the team
can contribute
Tada! Slack itself is a great initial topic that fits all those.
A side benefit would be encouraging knowledge sharing of slack.
Background
27. That would only show to users when fewer than three
messages have been written in the channel
1. Give conversation starters
Once people have started talking, revert back to the
standard channel meta text and CTAs:
28. What to measure
• I would measure success against our KPI: what %
of teams who see this get to 10 msg/3 ppl/24 hours
• I’d also want to measure the intermediate steps in
the funnel to see if that suggests why this worked or
didn’t, and next steps.
• I’d also keep an eye on other established KPIs to
make sure it didn’t hurt, say, invites or integrations.
1. Give conversation starters
29. What success looks like
1. Give conversation starters
Before (🚨) After
4+ invites sent / week 1,000 1,000
1 person writes publicly 50% 50%
2 more people write publicly 50% 50%
Exchange 10 msgs in 24h 25% 50%
6.25% (63 teams) 12.5% (125 teams)
If successful, we could activate 62 more teams per week
I’d expect the same number of unique people to exchange messages,
but that the # messages exchanged within 24h will increase:
30. 2. Use email to highlight early action
Background
When a team first adopts slack, not everyone is online at the
same time. This likely makes it hard to reach the ‘aha’ moment
of real-time conversation with your team.
This test would amplify use of email to create a sense of action
happening on slack, and encourage people to visit slack more in
the first 24h and participate.
31. 2. Use email to highlight early action
Quick sufficient test
Focus email copy on who’s talking + include a
natural call to action
Simplest version: Don’t show chats, just indicate
that people are talking (to amplify significance)
(Extra variant to test afterwards: Include all chats
and interactions in the email across channels)
32. Tatiana and Jim are talking on Slack
in the #general channel
See what they’re saying
Your team is collaborating on Slack - a great way for busy teams to
make life simpler, more pleasant, and more productive.
Too many emails? Eek.
Good thing there’s plenty of ways to stay posted:
Turn off emails and turn on desktop notifications
Download the desktop app or mobile app
Just turn the emails off, thanks
33. 2. Use email to highlight early action
Who gets this email, and how often?
First 24h: Send when there has been at least one person talking,
and at least two hours have gone by since last email
After 24h: dial back to every 12 hours
Sent to all invited team members, including those not registered
• If already registered, button directs to that channel
• If not yet registered, button first takes to register
• Don’t send if they’ve downloaded a client or enabled desktop
notifications
34. 2. Use email to highlight early action
What to measure
• I would measure success against our
KPI: what % of teams who get this email
pattern get to 10 msg/3 ppl/24 hours
• I’d also want to measure open and click
through rates to see if that suggests why
this worked or didn’t, and next steps.
35. What success looks like
2. Use email to highlight early action
Before (🚨) After
4+ invites sent / week 1,000 1,000
1 person writes publicly 50% 50%
2 more people write publicly 50% 70%
Exchange 10 msgs in 24h 25% 50%
6.25% (63 teams) 17.5% (175 teams)
If successful, we could activate an extra 112 teams/week
I’d expect this to increase both the rate of additional teammates
messaging on slack, as well as the total messages they exchange
36. 3. Emphasize desktop notifications
Quick sufficient test
Right now it’s not an emphasized action. What might happen if we
communicated it’s a priority?
Create an inline step that is 100% focused on encouraging user to
enable desktop notifications
Simple first experiment: put this step after the username is created,
but before entering slack. Good copy and personality is key to not
spammy feel.
37. 3. Emphasize desktop notifications
Slack is 3,924X better* with notifications
* This statement has not been evaluated by the Food and Drug Administration. May contain nuts.
OK that might be an overstatement. But you can make sure you
never miss a personal message, shoutout, or important Ritual
Coffee announcement by letting us to notify you instantly.
Feeling unsure? We hear you. You can always control how many
you receive, or turn them off entirely at any time.
Give notifications a shot
➔
38. 3. Emphasize desktop notifications
We really recommend enabling notifications, but no worries, we
won’t take it personally if not. Remember:
• Notifications make Slack super awesome
• You can always turn these off later
• There’s a “do not disturb” mode anytime you want quiet
➔slack.com would like to
send you desktop
notifications
We’d be honored if you enabled notifications
39. 3. Emphasize desktop notifications
What to measure
• I would measure success against our KPI: what % of teams
who get this inline step that get to 10 msg/3 ppl/24 hours
• I’d also want to monitor how many people enable
notifications compared to before to understand the story,
regardless of what happened with the KPI
• I’d want to monitor the rate of conversion from
registration to entering the slack app to better understand
what we might try as next steps
40. 3. Emphasize desktop notifications
What success looks like
I’d expect this to not significantly affect the first few people, but to
definitely increase the number of messages exchanged
Before (🚨) After
4+ invites sent / week 1,000 1,000
1 person writes publicly 50% 50%
2 more people write publicly 50% 50%
Exchange 10 msgs in 24h 25% 70%
6.25% (63 teams) 17.5% (175 teams)
If successful, we could activate an extra 112 teams/week
41. These are assertive tests, and there’s definitely many
more ideas like these
There’s a fine line between good growth and
spamminess. And it’s not a dichotomy - it’s solved by
being human and respectful.
I’d want to find the balance together with the rest of the
product team for how assertive tests can be while
everyone feeling comfortable with what ships to users.
That’s it. Phew.
Off my chest