Forecasting New Product Performance Like A Meteorologist

Forecasting Product PerformanceLike A Meteorologist

ProductCamp Boston 2012

#pcampBoston

A.Chakravarty 6/9/2012

Problems are just as frustrating…

Jim Kosek – Funny Weather VideoThree Key Forecasts related to Product Mgt:• Domino Effect – Power outages, Traffic backup, Black Ice• Redundancy Effect - Snow Drifts, Windy• Anticipated Failure – Heavy Snowfall

http://www.youtube.com/watch?v=Awwg0SO4Nck

PPT Notes provide addt’l information.

Workshop Overview• Forecasting – why is it important

• Forecasting Impacts Decisions

• Limitations

• Metrics & Data

• Methods & Techniques

• Leveraging forecasts workshop

– Key Product Metrics

– Internal partnering for product success

– Method and Technique

Review – 20 min.

Case – 25 min.

DISCLAIMER: This session will not discuss how you should forecast, but you should walk away having a better understanding of what to think about and the complexity of good forecasting for new products.

New Product Forecasting• Error Rates are HIGH!

• So Why Forecast?

1. New Product Development and Forecasting Problems. R. Simon, Journal of Business Forecasting 2009-2010.

Error rate in Guessing a Coin Toss (50%)

Forecasting – Why?New Product Development (NPD)

On average ~20% of company sales are New Products1

Newer products typically command higher profit margins1

In the US, 50% of revenues and 40% of profits are from New Products1

~26% of revenue from Engineering companies are from products < 3 years old1

~70% of today’s manufactured goods will be obsolete in 6 years1

In Fashion and High Tech that’s closer to 2 years1

35%-45% of New Products fail immediately2

According to Herb Baum, Former CEO of Campbell Soup Company, in consumer business, “80% of all new products fail, only 4% reach the 20 MM level and 0.5% break the 100MM mark.”3

1. New Product Forecasting 2006. M. Singh. MIT2. Doing it Right: Winning with New Products. R. G. Cooper. Product Management Institute 2006.3. Journal of Business Forecasting C. L. Jain, Editor. Winter 2009-2010.

Reduce Market Costs of Failed Products

Drive Sales & Profit

Forecasting Impacts Decisions

Product Launch

Sales Support

See Notes

Cover/Clothing

Time/Resources

Emergency Planning

Event Planning

Weather Forecastslet us decide:

Product Forecasts let us decide:

Forecasting NPD Limitations

Short term Accuracy

Significant data collection

Many data points

Change Impact

New patterns emerge

Long term Impacts

Weather ForecastsAre limited:

Product ForecastsAre limited:

High failure rate

Almost no accuracy

Data collection is sparse, erratic, and not necessarily sales drivers

Incomplete/Unknown data points

Product Change Impact

Environment Change Impact, game theory, competitive pressures

Long term Impacts are strategic only

Forecasting is difficult!

Metrics for Product ManagersQ

uanti

tati

ve

Metr

ics • Empirical

Data• e.g. Sales• Unique

Visitors• Subjective

Data• Market

Research• Surveys

Qualit

ati

ve

Metr

ics • Subjective Data

• Delphi• Sales Force

Composite• Focus Groups

• Pattern Data• Pattern

Recognition• Look-Alikes• Scenario Analysis

Metrics for Business

Product Metrics

Decrease Expenses

Increase Revenues

Business Metrics

Product metrics must impact a business metricProduct metrics need to be translated into a P&L statement

Example: Unique Visitors X Est. Avg Revenue per Visitor = Revenue

Persistence Today = Tommorrow

Trends – Nowcasting Extrapolation of current variables

Climatology Historical Extrapolation

Analog Patterns Looks like June 6th, 1874 so…, Scenarios, Looks-Like

Numerical Weather Predictions Computationally heavy, multi-variable predictive algorithms – many

variations

Basic Meteorological ForecastingKey Weather Forecasting Techniques:

Methods & TechniquesThree broad sets capture dozens of methods

1. Managing the Mysterious: How to Forecast New Products. Logistics Summit & Expo, Mexico 2010. Kenneth Kahn

2. New Product Forecasting 2006. M. Singh. MIT

57%44%

39%

14%

30%

19%15%

MktResearch

ExecutiveJury

Sales ForceComposite

ScenarioAnalysis

Looks-LikeAnalysis

Trend LineAnalysis

MovingAverage

Subjective QuantitativeMktResearch

Top methods used are shown by percentage

Some things to keep in mind to forecast

• Timeframe

• Assumptions

• Units of Measure – atomic

• Project Timeline and Milestones – including failure to meet impacts

• Ongoing, Cost Structure Estimations (for P&L Pro-Forma development)

• Risks & Mitigation

• Estimation technique for revenue or cost savings

• Revenue Derivation algorithm

• Validation & Assumptions Check

Workshop Preparation…

• TASK 1: Break out into groups of 2-5 people each and introduce yourselves.

THE FORECAST CASEWORKSHOPWeatherCommerce.com is a clicks-only online business that sells Thermometers, Barometers, Hygrometers, and Weather stations. The site has recently launched a new ecommerce sales engine to sell a new line of product – the Baltimore Weather Gauge. We need to forecast product performance post launch for 12 months.

• TASK 2: Each team take 5 minutes and create a list of 3-5 Key Product Metrics to measure new launch success for the Baltimore Weather Gauge

EXAMPLES:

• Monthly Sales (Units Sold) for 12 Months

• Monthly Sales ($) for 12 months

THE FORECAST WORKSHOP

• TASK 3: Each team takes 5 minutes to list 3-5 Key Resources and Key Deliverables that they would reach out to develop their forecast.

EXAMPLES:

• Marketing - Customer Market Survey, Realistic Price Points, Existing Customer Interest

• Ecommerce Sales Operations – Site Traffic and Conversion Rates for similar products.

THE FORECAST WORKSHOP

• TASK 4: Each team takes 5 minutes to list one method they would rely on to forecast – and a quick justification of why they would choose it.

• Note: There are no right or wrong answers. Many companies use up to 3 methods at the same time for validation and cross-check.

EXAMPLE:

• Online Market Surveys –With many existing customers through an ecommerce platform, we can quickly gauge product sentiment.

YOUR ORGANIZATION

• TASK 5: What techniques would you use for your company’s products?

THANK YOU!

Ananda S. Chakravarty

@achakravartyhttp://www.linkedin.com/in/anandachakravarty