4. First startup iteration “Garmz”
1,5 years building.
5 months live = cool! but houston, problem.
5. Pivot to LOOKK. This time we had a lot
of industry expert input.
Complete relaunch.
6. Pivot to LOOKK. This time we had a lot
6 Months building. of industry expert input.
After 3 months = by far better! Complete relaunch.
but still houston, problem.
7. Pivot to LOOKK. This time we had a lot
6 Months building. of industry expert input.
After 3 months = by far better! Complete relaunch.
but still houston, problem.
TOO SLOW…
TOO BIG…
= TOO MANY UNPROVEN ASSUMPTIONS…
8. Slow launches are dangerous
We learn the most when launch.
Therefore we need to launch earlier/smaller/often…
9. Launch earlier…
Focus on the actual core of your product.
Example: Buffer linked their payment
submit to a 500 error page until
bug reports came.
10. Slow launches are dangerous…
You build baggage…
Why don't people launch earlier ?
Why don't people only talk to "right" customer ?
Why do startups always launch "in one month" ?
11. #1
We want to stay in our comfort zone.
Our team wants to stay in the comfort zone.
15. Let’s get i like
investors first networking
Let’s build
i like coding
this first
We decide our strategy based on our comfort zone.
We want to excel in our comfort zone topics.
20. “people want to do the best they can”
Solutions we (tried to) apply:
Timeboxed launches (for editorial teams).
A/B Launches (for development).
Cultural Change (for everyone).
21. “people want to do the best they can”
Solutions we (tried to) apply:
Timeboxed launches (for editorial teams).
A/B Launches (for development).
Cultural Change (for everyone). (we failed at that one)
25. #3
Also you have no idea “it needs training…”
what you do…
26. #4
Discovery Validation Efficiency Scale
“Ok we did custdev.
Now back to real
business.”
This process has not
much to do with the
phase you are.
Apply lean and custdev
to feature rollouts.
(And feature removals as
we did it…)
28. First step:
METRICS!
I ♥ it, but…
Presentation about it:
http://www.slideshare.net/
andreasklinger/metrics-lesson-
learned-leancamp-barcelona
TL;DR:
* Focus on Retention
* Use your database for insights
* CustDev > Metrics
29. In early stage: CustDev eats
Metrics for breakfast.
Discovery Validation Efficiency Scale
Qualitative Quantitative
Validation Validation
(eg. customer interviews) (eg. metrics)
Source: Custdev.com
30. CustDev in a team
All of the team brainstormed assumptions on
pink post-its.
We clustered them to groups (topics).
Questions (green) for each group that should
get people to mention our assumptions.
We did this for both user types (designers,
users)
Source: Custdev.com
31. User Example
Examples
We wanted to give them the opportunity to
Examples
raise the assumption in the group without
The questions should not be leading but open the possibility for
leading to it
someone to say those assumptions.
We wanted them to pass the “mother-test”.
We wanted them to pass the “mother-test”.
Build questionnaires and set up meetings.
Designer Example
32. We had help by a saint
Salim Virani (of leanca.mp) helped us
creating those topic maps.
If you can afford him get an external
expert like him involved. Especially if you
try to do this with a larger group.
33. Speeddating aka
Industry events
We slimmed down the questions to
the most important ones and added
basic “starter questions”.
I went to tradefairs in Paris and
London and talked with several
dozens of designers in one day.
How:
* “I am doing a survey…”
* Have respect for the reason why
they are there
* Write superfast or have someone
second writing for you.
* If you can have something phsyical
to show with you. And “ask” about
that.
34. We collected the answers
And grouped them within the topic map.
Source: Custdev.com
35. We went back to our
assumptions
And moved all those down that have not
been mentioned.
We regrouped the rest in their new contex
Yellow is a new information…
Roughly 90% of our assumptions turned
out to be not true.
36. We went back to our
assumptions
And moved all those down that have not
been mentioned.
We regrouped the rest in their new contex
Yellow is a new information…
Roughly 90% of our assumptions turned
out to be not true.
But… some turned out to be by
far bigger than expected.
Now we focus on that.
37. Bullshit to externals,
but always know where
your startup really stands…
and act accordingly!
#LSMLondon – @andreasklinger
38. Startups have different stages.
Discovery Validation Efficiency Scale
Read: The Startup Owners Manual - Steve Blank
Or: Custdev.com by Patrick Vlaskovits and Brant Cooper
39. Startups have different stages.
Discovery Validation Efficiency Scale
Problem/Solution
Read: The Startup Owners Manual - Steve Blank
Or: Custdev.com by Patrick Vlaskovits and Brant Cooper
40. Startups have different stages.
Discovery Validation Efficiency Scale
Problem/Solution
Product/Market
Read: The Startup Owners Manual - Steve Blank
Or: Custdev.com by Patrick Vlaskovits and Brant Cooper
41. Startups have different stages.
Discovery Validation Efficiency Scale
Acquisition
Problem/Solution
Company building
Product/Market
Read: The Startup Owners Manual - Steve Blank
Or: Custdev.com by Patrick Vlaskovits and Brant Cooper
42. Startups have different stages.
Discovery Validation Efficiency Scale
Acquisition
Problem/Solution
Company building
Product/Market Magic Dragon Land
New business models
Read: The Startup Owners Manual - Steve Blank
Or: Custdev.com by Patrick Vlaskovits and Brant Cooper
43. Startups have different stages.
Discovery Validation Efficiency Scale
Acquisition
Problem/Solution
Company building
Product/Market Magic Dragon Land
New business models
Read: The Startup Owners Manual - Steve Blank
Or: Custdev.com by Patrick Vlaskovits and Brant Cooper
45. #1 Reason for startups dying:
The follow the patterns of later stage companies
(aka premature scaling)
Source: Startup Genome Project Report
46. Startups have different stages.
Discovery Validation Efficiency Scale
Acquisition
Problem/Solution
Company building
Product/Market Magic Dragon Land
New business models
Growth investments
B2B corporations
PR campaigns
Marketing Stunts
Read: The Startup Owners Manual - Steve Blank
Or: Custdev.com by Patrick Vlaskovits and Brant Cooper
47. recommended presentations
by @robfitz on custdev/lean:
http://leanca.mp/2012/01/intro-to-lean-startup-with-no-buzzwords-by-robfitz/
http://thestartuptoolkit.com/blog/2012/05/how-to-do-and-what-to-expect-from-
early-stage-customer-development-sales/
by @serenestudios and @stueccles on metrics:
http://www.slideshare.net/danhilltch/cohort-analysis-leancamp
http://www.slideshare.net/stueccles/lean-startup-metrics
48. And if they haven’t
died they are lean
happily ever after…
Thanks,
@andreasklinger
Now back to work! ;)
#LSMLondon – @andreasklinger