Talaria - Forecasting - Disruptive Insights Ideation Challenge

Challenge Overview

We aim to create a data model that forecasts subscriber churn and additions for an imaginary broadband and handheld provider.  Validation is a priority and will be performed against historical data.  

Through what-if scenarios, the validated model would be used to forecast performance for the next year and each of four additional years that follow.  The model would focus on the UK+EU markets and be sensitive to GDP, household income, increases and decreases in addressable markets - or any/all the major factors that typically impact buyer choice.  The model should focus on the consumer market.

Given 5 years of annual “actuals” for broadband and 5 years of quarterly “actuals” for mobile, what analytic techniques should be used to create such a model? We are interested in both novel techniques and effective ones that are described in the literature.

Your task in this challenge is to provide a briefing paper that describes an approach to the problem.  Your paper should include references in order to support your approach or to provide references to existing approaches.  Extra consideration will be given to submissions that include POC implementations of the concept.

Contestants may submit multiple solutions.  To include multiple solutions, package each in a zip file, bundle them in a single zip file, and submit the single zip file.

Background

Telecom providers sell products such as broadband and mobile phone contracts. These contracts sell products of different types and capabilities, which are outlined below:

• Mobile
  • Handset
  • SIM-only
• Broadband
  • Copper
  • Superfast
  • Ultrafast

For each of the five products the customer would like to forecast the following:

• Volume
• Net adds / churn - The number of new subscribers, or gross adds, minus the number of customers that drop service, which is called churn
• ARPU - Average Revenue Per User (if possible using industry standards)

Our model would eventually need to account for:

• Different subscription lengths
• Shocks introduced by competitors, eg price disruption and seasonal releases of new product lines (eg Samsung in Spring, Apple in the Fall).

An ideal, complete vision for this application would allow:

• Forecasts that are within 20% of eventual actual results for the following year
• What-if scenario building by users
• Ability to load new datasets in order to test them for relevance against the validated forecast (eg improves accuracy)

We will provide a list of the potential input variables in the challenge forum.

Submission Requirements

Your submission should include a text, .doc, PPT or PDF document that includes the following sections and descriptions

• Overview: describe your approach in “laymen’s terms”
• Methods: describe what you did to come up with this approach, eg literature search, experimental testing, etc
• Materials: did your approach use a specific technology?  Any libraries?  List all tools and libraries you used
• Discussion: Explain what you attempted, considered or reviewed that worked, and especially those that didn’t work or that you rejected.  For any that didn’t work, or were rejected, briefly include your explanation for the reasons (e.g. such-and-such needs more data than we have).  If you are pointing to somebody else’s work (eg you’re citing a well known implementation or literature), describe in detail how that work relates to this work, and what would have to be modified
• Data:  What other data should one consider?  Is it in the public domain?  Is it derived?  Is it necessary in order to achieve the aims?  Also, what about the data described/provided - is it enough?
• Assumptions and Risks: what are the main risks of this approach, and what are the assumptions you/the model is/are making?  What are the pitfalls of the data set and approach?
• Results: Did you implement your approach?  How’d it perform?  If you’re not providing an implementation, use this section to explain the EXPECTED results.
• Other: Discuss any other issues or attributes that don’t fit neatly above that you’d also like to include.

Proof of Concept

• Extra consideration given to PoC code in Python and Jupyter Notebook (or anything else, so long as we can run it without a license) to illustrate particular approaches / solutions.

Judging Criteria

We provide for 5 awards.  One first-place award is guaranteed.  
These submissions will be evaluated subjectively by the client based on the following criteria:

1. Completeness and Effectiveness (50%)
   1. Did you complete the sections as required above
   2. What're the key insights we can get from your analysis?
   3. How will these discovered insights benefit the client?
2. Feasibility (50%)
   1. Does your submission include enough detail for us to understand if this approach is feasible?
   2. Is your solution more likely feasible than other submissions to the challenge?
3. Proof of concept (PoC) solutions are appreciated and will be weighed over otherwise similar submissions.

Final Submission Guidelines

• Documentation
• Code