stop throwing money out the window

Post on 08-Sep-2014

741 Views

Category:

Technology

2 Downloads

Preview:

Click to see full reader

DESCRIPTION

Presentation at WebExpo Prague 2013. Description below ---- We have learned how to build software: Extreme Programming gave us the developer tools and Scrum the project management tools. But we are still investing a lot of money in our ideas and most of them fail. 9 out of 10 startups are unsuccessful. Why is that? One reason is that we still make assumptions about our users' needs. Repeat after me: "I am not my user!" This talk will discuss about minimum viable products, validated learning and continuous deployment: how to write the minimum amount of code that can teach us something about the user and only then developing the full feature (instead of waiting to have the perfect feature that maybe nobody wants).

TRANSCRIPT

STOP THROWING MONEY OUT THE WINDOW@FlaviusStef, Mozaic Works

v.2, includes design thinking

Me

2001-2006Web Developer

2007-2010Startup co-founder

2011-2014Coach and TrainerAgile, Lean, Org. Change

Follow me on Twitter

More on Slideshare

1. WASTE2. LEAN STARTUP3. TECHNIQUES

4. EXAMPLES5. PRINCIPLES

1. WASTE

WASTE LEAN STARTUP TECHNIQUES EXAMPLES PRINCIPLES

Syneto WCF – Active Directory

Cost to implement Total revenue0

10000

20000

30000

40000

50000

40000

6700

The Tutorial

Not the actual app

2 months250 man-daysUsage: - 30%

My claim

MOST BUSINESSES WASTE MONEY BECAUSE THEY INVEST IN PRODUCTS OR FEATURES

NOBODY WANTS!

Throwing money out the window?

Build something nobody wants

Build more than the users need

Opportunity cost

More?

The iPotty – Amazon Remote Headband – Inventorspot

Shoe Umbrella – Luux

More?

The Hair Hat – FlairHair

Wheel Mate Steering Wheel Desk – Techeblog

We assume too muchWe assume too much

Photo: itspaulkelly on Flickr

People will understand how it worksI need it, so the others do tooI understand what the user needsThe user won't be distractedThe customers will want to payIt's useful to integrate with system <X>

2.LEAN STARTUP

WASTE LEAN STARTUP TECHNIQUES EXAMPLES PRINCIPLES

Runexperiments

Photo: zhouxuan12345678 on Flickr

Do research

Create hypothesis

Design experiment

Run experiment

Analyze data

Ask question

Hypothesis false

yes

The lean startup

The lean startup

But:- What to build?- How to measure?- What to learn?

Design Thinking

Who is the user?

What is their problem?

How could we solve it?

What's the MVP?

Is it working?

BUILD

MEASURELEARN

What is the problem?Who is our customer?

What is the next step?

Is it useful?Best alternative?

What is the solution?

3.TECHNIQUES

WASTE LEAN STARTUP TECHNIQUES EXAMPLES PRINCIPLES

What is the problem?Who is our customer?

LEARN

InterviewsShadowing

Innovation GamesEmpathy mapLean CanvasStorytelling

Olark, Qualaroo

earlier

later

LEARN

EMPATHY MAP

OLARK, QUALAROO

LEARN

LEAN CANVAS

LEARN

Document HypothesesCreate business case

Prioritized BacklogLean Canvas

MMFConcierge

UI SketchesIncremental build

BUILDWhat is the next step?What is the solution?

earlier

later

BUILD

We believe that (capability) Will result in (outcome) We will know we are right when we see (measurable impact)

DOCUMENT HYPOTHESES

We believe that having a mobile homepageWill result in different usage based on the consumer's countryWe will know we are right when we seeat least one country having 10% more mobile traffic

BUILD

DOCUMENT HYPOTHESES

BUILD

xx x

x

x

x x

x x x

Minimum Marketable Feature

BUILD

CONCIERGE

BUILD

MEASURELEARN

Paper PrototypingFake/Landing Page

AnalyticsA/B testing

Client DemosTrack $$$ earned

Is it useful?Best alternative?

earlier

later

BUILD

MEASURELEARN

FAKE/LANDING PAGE

BUILD

MEASURELEARN

FAKE/LANDING PAGE

BUILD

MEASURELEARN

PAPER PROTOTYPING

ANALYTICS

BUILD

MEASURELEARN

BUILD

MEASURELEARN

ANALYTICS

InterviewsShadowing

Innovation GamesEmpathy mapLean CanvasStorytelling

Olark, Qualaroo

MVPConcierge

UI SketchesIncremental build

Paper PrototypingFake/Landing Page

AnalyticsA/B testing

Client DemosTrack $$$ earned

BUILD

MEASURELEARN

Document HypothesesCreate business case

Prioritized BacklogLean Canvas

4. EXAMPLES

WASTE LEAN STARTUP TECHNIQUES EXAMPLES PRINCIPLES

The Syneto Email Digest

B

ML

Needs email digest (interview)

License sales

Profitable

Needs minor tweaks

Req. GUI with 8 actions/email

Needs only 1 action

Needs no extra actions2x Client Demo

5x Client Demo

Review instrumentation logs

Wait for feedbackB

M

L

B

M

L

B

M

L

L

Basic CLI

Improved CLI, released beta

Built GUI with 4 functionalities

Added one extra action

Nothing more built

B

M

L

B

M

L

Initial list

● Release Email● Delete Email● Whitelist Sender● Blacklist Sender● Release All● Delete All● Whitelist All● Blacklist All

for e

ach

ema i

lfo

r all

emai

ls

What we implemented

● Release Email● Delete Email● Whitelist Sender● Blacklist Sender● Release All● Delete All● Whitelist All● Blacklist All

for e

ach

ema i

lfo

r all

emai

ls

What the customers asked for

● Release Email● Delete Email● Whitelist Sender● Blacklist Sender● Release All● Delete All● Whitelist All● Blacklist All

for e

ach

ema i

lfo

r all

emai

ls

Let's do the math

3 features x 2.5 days / feature = 7.5 days

7.5 days x 115 EUR / day = 862.5 EUR

Photo: xcode on Flickr

B

ML

Team retrospective“Something” for SMs

NPS

What's hot, what's not

There is a need. Also received client req.

Facilitation is important

Interest validated

Is there actually interest?~ 30 attendees

Client accepted

Interviews with existing clients

Check # of registrationsB

M

L

B

M

L

B

M

L

L

Leadership Workshop (2h)

1d workshop structureLean Canvas

Built 1dCreated 2d “official” structure

Scheduled 2d for fall

Second half

B

M

L

B

M

L

TODAY

5.PRINCIPLES

WASTE LEAN STARTUP TECHNIQUES EXAMPLES PRINCIPLES

Generate multiple alternatives

DI VERGE CON VERGE

Don't listen to the HiPPO

Identify your assumptions

Weather forecastIn 10 days

Clear

Follow the 80-20 rule

New features Existing features

No measurement, no learning

Release early, release often

CommitAutomatedacceptance

UX testing

Useracceptance

Scalabilitytesting

Production

git pushautomated

manual

● Build (war, gems, minified js/css)

● Run unit tests● Static analysis● Deploy artifacts

CI Server

mon

itors

launches

WRAP-UP 1. Start by empathizingex: Empathy Map2. Identify assumptionsex: Lean Canvas3. MMFs are minimalex: Concierge4. Measure to learnex: A/B testing5. Continue refiningex: Incremental development6. Follow the money

B

ML

… more?

THANK YOU!

Twitter: @FlaviusStefBlog: flaviusstef.wordpress.comMore slides: slideshare.net/fstef

top related