spigitengage predictions webinar

55
v Welcome to the Innovation Café Today’s Presenters: Milind Pansare, Kate Bennet and Dr. Anna Gordon, Mindjet

Upload: milind-pansare

Post on 15-Jul-2015

156 views

Category:

Software


0 download

TRANSCRIPT

vWelcome to the Innovation CaféToday’s Presenters: Milind Pansare, Kate Bennet and Dr. Anna Gordon, Mindjet

Milind Pansare, Vice President Marketing

Milind is Vice President of Marketing at Mindjet. He has held product engineering and marketing leadership roles for over 25

years at Silicon Valley startups and large tech companies such as Sun Microsystems (now part of Oracle), HP and Saba. Milind

is a product expert on social collaboration and innovation management technologies. He has served as an advisor to several

startups and startup incubators and holds a degree in Computer Science.

Kate Bennet, Director of Product

Kate Bennet is responsible for the design and delivery of Mindjet’s enterprise platform SpigitEngage and for researching and

scoping the next generation of crowdsourced innovation. Kate has worked in the innovation space in both the public and private

sectors, and has experience implementing, advising on and supplying innovation management software. Prior to Mindjet Kate

was Innovation Manager for the UK government and a Director and Trustee at Warwick University.

Dr. Anna Gordon, Data Scientist

Anna Gordon analyzes data across all departments – working on a variety of projects from measuring and optimizing marketing

and sales efforts to developing algorithms and data-driven features for SpigitEngage. Anna’s team creates approaches to study

and adapt innovation networks to generate the best outcomes through modeling and measuring engagement, understanding

how cooperation evolves in the network, and how trust behaviors affect information flows.

About our speakers

MINDJET CONFIDENTIAL

ABOUT MINDJET

Mindjet enables organizations to manage a pipeline of ideas

to drive new business strategies, product development,

operational efficiencies, customer insight and employee

engagement.

3

MINDJET CONFIDENTIAL

REDUCE COSTS INCREASE PRODUCTIVITY ACCELERATE GROWTH

DRIVING BUSINESS IMPACT

Proven & Quantifiable Impact on ROI

4

MINDJET CONFIDENTIAL

CROWD INNOVATION IS FOR EVERY PART OF THE BUSINESS

5

INNOVATION CLOUD

StrategyBusiness Model, Revenue

Streams

MarketingConsumer Insights, Brand

Engagement

Product DevelopmentNew Products, Time to Market

HREmployee Engagement,

Innovation Culture

Market InnovationCustomer Insights

Offer InnovationProducts & Services

Process InnovationCost Reduction

Employee Engagement Culture of Innovation

MINDJET CONFIDENTIAL

INNOVATION IS A CONTINUOUS PROCESS

Innovation must be managed just like any other important

process. An accelerated pipeline of the best, actionable ideas

that are vetted and ready for development and

implementation.

Idea Submission Idea Vetting Idea Feasibility Idea Implementation

6

MINDJET CONFIDENTIAL

UNIQUE ADVANTAGE OF MINDJET OUR KEY DIFFERENTIATORS

SINGLE PLATFORM AND BEST PRACTICES

Crowd Science + Behavioral Dynamics + Predictive Analytics

Manage(Select & Align)

Outcomes(Execute & Repeat)

Engage(Source & Refine)

7

MINDJET CONFIDENTIAL

UNIQUE ADVANTAGE OF MINDJET

• Purpose-driven: Drive real outcomes – new products & services,

customer experience, operational efficiency & agility

• Social: Game mechanics for viral adoption & ongoing participation

• Mobile: Any device, anytime, anywhere, anyone

• Fun: Friendly, intuitive, user experience

• Culture: Make innovation a daily habit for the entire enterprise

EngageCrowd Dynamics

8

CREATE A CULTURE OF INNOVATION – SUSTAINED ENGAGEMENT

Anywhere, anytime,

any device

Engage at scale, ensure

diversity of opinions and

aggregation of

knowledge

Solve specific

organizational

challenges on a

timeline

MINDJET CONFIDENTIAL

UNIQUE ADVANTAGE OF MINDJET

• Crowd science: Industry-leading algorithms to find the best ideas

• Velocity: Speed decision-making & vet ideas faster with

Automated Graduation & unique Pairwise Voting

• Optimization: Pattern recognition & machine learning for

continuously improvement based on past results

• Unbiased: Eliminates mob behavior to get true outcomes:

prevents gaming

• Nurture: Continuously surface & refine ideas for ongoing

innovation

ManageCrowd Science

10

FIND BEST IDEAS– CROWDSCIENCE

Define idea stages and automated

graduation criteria unique for your

organization. Use the crowd to do

heavy lifting in initial stages. Apply

built-in workflow engine for expert

evaluations and vetting in later

stages.

Ensure all ideas are

carefully considered and

the best ideas are

surfaced.

MINDJET CONFIDENTIAL

UNIQUE ADVANTAGE OF MINDJET

• Repeatable: Institutionalized innovation process; proactively manage

and measure a pipeline

• Predictable: Validate the cost, feasibility, time, connection to strategy,

value of ideas with unique crowd Predictions and expert evaluations

• Measureable: Data visualizations and Analysis for measurement of

benchmarks and effectiveness

OutcomesCrowd Analytics

12

DRIVE BEST OUTCOMES– REPEATABILITY & PREDICTABILITY

Create expert workflows

matched to your

business process

Measure innovation program

health with the help of business

analytics, dashboards and

reports

Quantify your innovation pipeline

with crowd predictions of cost,

time and revenue.

SpigitEngage Predictions

A new premium module for SpigitEngage that enables customers to:

Leverage the crowd to quantify the value of ideas in the innovation pipeline.

Predictions: Business Benefits

BUSINESS BENEFITS

Use the crowd to forecast revenue, time to market

and implementation costs.

Make better decisions with crowd predicted

quantitative data on each idea, in addition to experts.

Predict the ROI of the Innovation Program.

INNOVATION PROGRAM CHALLENGES

“Good” ideas, but lack of quantitative

metrics.

Too many ideas, not enough experts to

analyze business value.

Hard to measure and predict success of

the program.

The Crowd Knows: Sociology of predictions

Sociological validation of the wisdom of the crowd

1987, Prof. Jack Treynor, University of Southern California:

850 jelly beans, 56 students. Group estimate was only 2.5% off.

2007, Michael Mauboussin, Columbia Business School:

1,116 jelly beans, 73 students. Average guess was only 3% off.

The collective knowledge of the crowd creates

better estimates.

Financial analysts have long studied the wisdom of the crowd and use this for quantitative models to drive

Investment Strategies:

There’s a long history of leveraging crowds to predict results

1968- Locating a lost submarine 1998- Who Wants To Be a Millionaire

Experts guess correctly 65% of the time.

Audience guesses correctly 91% of the time.

20 miles search area. Diverse experts submitted

guesses. Collective estimate was only 220yds off.

Crowd is historically more accurate than

experts with an error of less than 1.64%.

Election results

Presidential elections

Essential qualities for accurate crowd prediction

Diversity of opinion – to get different information.

Independence from one another – so one strong leader doesn’t dominate.

Decentralization – so that errors are balanced by the others.

A method for aggregating opinions – so that all opinions are included in the

decisions.

1

2

3

4

Predictions:

Product Functionality

Predictions: Crowd Analytics

Using Predictions for automated graduation

Best practice:

Use Predictions as a

graduation criterion.

“Top x” in Best Overall move

to the next stage.

Setting up Predictions

Configured for each challenge:

1. The roles that are able to cast prediction

votes.

2. The stage that an idea must be in to

have Predictions cast on it.

3. The Prediction questions that will be

shown.

4. The currency that will be used for the

Revenue and Cost questions.

Recap

Predictions is a time-boxed phase used within a challenge.

Diverse crowds make forecasts on revenue, cost and/ or time.

Results highlight low-hanging fruit vs the more disruptive investments.

Internationalized module – available with a wide choice of currencies and in 11

different languages.

1

2

3

4

Under the hood: Crowd Science

Under the hood

Sociological experiments validate the necessity of Crowd Dynamics in the platform to

ensure unbiased predictions.

In order to make good decisions, Crowd Analytics in the platform enable you to

visualize the data meaningfully.

In addition to these, ensuring the validity of the data requires advanced patented

algorithms – Crowd Science.

Under the hoodThe Science of Convergence

Much like Maps on your phone that leverage multiple geo-location signals.

Predictions creates convergence through many data points and use of crowd science algorithms.

Under the hood When a large, diverse crowd is asked a numerical

question, their answers take the shape of a

probability distribution, with the majority of answers

centered around the middle.

In the Predictions algorithm, we draw random

samples from a distribution like the “bell curve”, and

update it as we receive information, or votes, from

the crowd.

If the majority of the crowd votes up, we move the

distribution to the right, if the crowd votes inward

toward a single point, we shrink the distribution, until

we hone in on the crowd’s estimate.

A Bell Curve: Common User Voting Behavior

Progression of Distribution as Votes are Received

Progression of Distribution as Votes are Received

Progression of Distribution as Votes are Received

Progression of Distribution as Votes are Received

Progression of Distribution as Votes are Received

Progression of Distribution as Votes are Received

Progression of Distribution as Votes are Received

Progression of Distribution as Votes are Received

Progression of Distribution as Votes are Received

Progression of Distribution as Votes are Received

Progression of Distribution as Votes are Received

Progression of Distribution as Votes are Received

Under the hood A measurement of certainty based on

the extent to which the crowd agrees

is factored into the final computation.

High agreement happens when the

majority of the crowd vote in the same

direction, so the distribution is able to

shrink around the crowd’s estimate.

Low agreement occurs when people

vote in different, disagreeing

directions, so the distribution can

never converge.

Black = low agreement

Red = high agreement

Possible Distributions of Users’ Votes

Under the hood

To compare ideas with varying results and calculate the “best overall” ranking, we normalize

by using an equation approximately equal to

Each variable is weighted by it’s agreement score, so a high agreement score will leave the

final numbers as they are, whereas a low agreement score will pull down revenue, and

increase cost and time.

revenue- cost

time.

Summary: Predictions

Real-world business application requires:

Crowd Dynamics to ensure unbiased outcomes;

Crowd Science to deliver patented algorithms for accurate data;

Crowd Analytics to meaningfully visualize big data and make better decisions.

Predictions is a new premium module for SpigitEngage that enables customers to quantify the innovation

pipeline.

vQ&A

now.mindjet.com/predictions