Unilever reserves the right of final interpretation

Coding your Future

——2020 Unilever Hackathon

As the development of artificial intelligence and digitalization, Unilever (China)has launched many internal digital projects. 2020 Unilever Hackathon will be held to find the most innovative business solutions and the best IT technology tycoons currently. In this Hackathon, we focus on the competition not only in AI technology but also in innovative projects. Participants can learn from each other via this competition, showing their extraordinary wisdom and talent! Moreover,generous cash rewards and benefits are waiting for you!

Sign up Deadline: April 13th (Before Beijing Time 23:59)

Who Can Participate?

2020 Unilever Hackathon is open to everyone. It doesn't matter whether you are a student, a developer or a start-up team, Unilever welcomes your participation as long as you have a strong interest in the competition topics. Come on and join this competition!

* Participants can choose to participate individually or in teams (team size within three members).

Competition Benefits

  • 1

    Competing with IT Masters

  • 2

    Cash rewards
    and Unilever gift packages

  • 3

    Unilever OFFER and the
    chance to cooperate
    with Unilever

  • 4

    Trip to London

Competition Benefits

1. Competing with IT Masters

Hackathon is famous for its competition in IT programming. Unilever provides a platform for all the technology enthusiasts to communicate and interact with each other. On this platform, you will compete with numerous outstanding technical and digital talents. Here you may be the strongest Hackers that Unilever is looking for!

2. Cash rewards and Unilever gift packages

We have prepared generous cash rewards (excluding topics marked with “*”) and Unilever gift packs for winners. Team with best performance of each topic will win the champion. The rest two groups will be the Runner-up.

★Champion: BD Master / U + Creator Up to ¥20,000 in cash (or equivalent gifts)
★Runner-up: Winning Award Unilever Gift Pack

3. Unilever OFFER and the chance to cooperate with Unilever

For student participants, if you met the requirements of 2020 Unilever Leadership Internship Programme, you would get a Green Pass or even an Offer. For other social elites, you would get an exclusive Unilever internal referral. For enterpriser participants, you would have the opportunity to cooperate with Unilever business.

4. Trip to London

Winners of the 2020 Unilever Hackathon will have the opportunity to represent Unilever China to participate 2021 Unilever global business competition final round in UK. You can compete with the global elites!
*Reminder: The expenses including round-trip air tickets and accommodation will be covered by Unilever.

Competition Timeline

The most comprehensive competition strategy and detailed schedule are as following. Please read carefully!

Online Registration: March 23rd to April 13th (Before Beijing Time 23:59)

Unilever's second Hackathon will be officially launched on March 23rd. Participants must fill in the entry form online and sign a confidentiality agreement. After the registration, you will receive a confirmation email, including the invitation of briefing and introductory session, the detailed requirements, data,business brand proposals, competition introductory letter, etc. Please sign up as soon as possible!

Questions and Answers Session: March 30th-April 3rd

Participants who have successfully registered will receive confirmation emails and will be invited to participate in the briefing and introductory session. Please check your mailbox frequently.

Submission of Preliminary Works: March 27th to April 22nd (Before Beijing Time 23:59)

Preliminary Works can be submitted during this period, and the format can be found in the requirements of each topic in detail.

Preliminary Selection: April 23rd-April 29th

All preliminary entries will be assessed based on the competition criteria.(The topics marked with“ * ” have only one round.)

Preliminary Selection Result: April 30th

Top 5 of each topic will enter the semi-finals. Details will be sent to you.

Preparation for the Semi-finals: April 30th to May 17th (Beijing Time 23:59)

All entrants need to optimize their preliminary entries or finish new topics of semi-finals. Semi-final Q&A session will be held between May 4th-May 8th. The detailed requirements will be emailed to you.

Questions and Answers Session: May 4rd- May 8rd

Participants who have successfully entered into semi-final round will receive a confirmation email, including an invitation of Q&A session. Please check your mailbox frequently.

Semi-finals Selection: May 18th-May 22nd

Announcement of Finalists: May 23rd-May 24th

Top 3 of each topic, 21 groups of participants in total, will participate in the final competition.

Final Competition (online + offline): May 28th-May 29th

The Hackathon final round will take the form of both online and offline. Participants in Shanghai will come to Unilever's Shanghai headquarter while participants from other regions log in via online video conference system to compete.

Each topic will compete for a set of champions and 2 sets of runners-up.

Competition Topics Introduction

There are 2 themes that divided into 10 topics. Participants can only choose one topic according to your own interests. The detailed content is as following. For the submission format, please refer to the requirements of each topic. The details of semi-final round and the results of the first round will be announced on Apr.30. The topics marked with * have only one round of selection, with prizes but no cash reward.

Dear Hackers, 2020 Unilever Hackathon is coming! We are looking forward to your participation!

Big Data & Digitalization

Cross Category Prediction for Consumer Preference

Detail Briefing:
Consumers’preference towards products within a specific category has been explored within Unilever, e.g., target setting for skin tone or tactile target for body care. However, there is little understanding about consumers’ preference across categories. The concept is not new in marketing, and known as cross-category selling, e.g., beer and diapers displayed in the same section to boost beer purchase by males. It can be generalized to R&D in the context of product optimization and well fit with our personalization Competition Agenda, e.g., using consumers' preference to facial cream to predict the preference to body cream. With our multi-category abilities, following questions can be asked:
  • Does the product preference interdependence exist?
  • If so, how can we quantify the cross-category dependence?
  • Is the dependency a one-way influence or 2-way interaction?
  • What are the hidden factors that can explain or drive the interdependence? Are the factors intrinsic, e.g., psychological (personality, motivation...) or physical (skin condition, micro biome...), extrinsic, e.g., long/short term media exposure, peer influence, seasonal change, etc., or both?
  • How to leverage the insights for product personalization?
Please propose your ideas of how to answer the questions above and focus on consumer preference towards product usage and experience without brand impact. To start with, please choose at least 2 categories for your proposal to understand the cross-category prediction. One of them has to be one of our BPC categories (skin, hair, oral, and deo).
Your idea could be a study plan for an empirical study (1), or data mining or simulation with existing data (2). For both options, please specify the questions to be tackled. In your final presentation for option 1, we would like to understand the rationales behind the study. It will include, but not be limited to, the materials (products/prototypes) to be developed, participants (target consumers) to be tested, study protocol, and budget estimated. For option 2, please include details on the dataset(s) you will work on, tools to fetch the data, methods for analysis, and your final conclusion.
All the work must be novel and original. Reviews or summary of existing reports are not accepted.

1. CV of participants
2. PPT Deck (understanding and dissection of the question)
3. Study proposal on how to tackle the questions

Judging Criteria:
Round I:
1. Comprehensiveness of Solution (25%)
2. Feasibility of the proposal (40%)
3. Potential for commercial use and replicability of proposal (40%)

Online Application

Increasing offline retail store coverage and estimating their sales potential

Detail Briefing:
Offline retail store is an important sales channel for Unilever. Each day, Unilever is faced with the following questions: how many retail stores are there in total? Out of which how many are not covered by Unilever (do not sell Unilever products)? Where are these stores located? Do they still exist? How are their sales potential? Given this background, Hackathon participants are challenged to update a specific city’s retail store statistics through data mining and estimate the sales potential of uncovered retail stores using AI models, thus locating the new stores Unilever should sell their products.

Round I:
1. Proposal (PowerPoint at most 5pages)
2. Proactively collected data(Excel)
3. CV & Previous Related Projects

Judging Criteria:
Round I:
1. Background of team (15%)
2. Previous Projects (15%)
3. Comprehensiveness and feasibility of proposal for commercial use and replicability (35%)
4. Readiness of external data (35%)

Online Application

Analysis of scattered static data

Detail Briefing:
Round 1: Aggregate data from all levels of operation (production lines, workshops, factories, and supply chain departments) of the plant, and display them in the form of dashboards. In the end, the static data scattered everywhere can be unified, and filtered and analyzed by related factors (such as product SKU);

Please submit other types of materials such as PPT, code, and demo app.
* Participants should be interested in data analysis and be familiar with at least one of the tools of Power BI, Tableau, and Qlik.

Judging Criteria:
1. Relevant data can be displayed correctly (50%)
2. You can connect dashboard display by SKU or date (20%)
3. The data can be automatically refreshed periodically (20%)
4. Data import should be more convenient, and data sorting and cleaning need to be automated (10%)

Online Application

Provide a stock allocation solution (including the numeric results and calculation logic) for Unilever supply chain

Detail Briefing:
Provide a stock allocation solution (including the numeric results and calculation logic) for Unilever supply chain to maximize service level with lower stock level, under the given data and assumption

The solution should include but not constrained to:
1. Stock allocation proposals for each Unilever distribution center
2. Calculation logic/algorithm illustration and coding if any
3.Simulation codes (on R or Python) to verify the impact on service and stock level based on the stock allocation proposal
*Please refer to data and business case details in attached files

Judging Criteria:
1. Logic and the thinking of the solution (Qualitative) 30%
2. A simulation program will be provided to evaluate the Order fulfilment and Month End Stock Value (Quantitative) 70%

Online Application

*Use programming and AI intelligence to automatically write and identify articles that consumers love to read

(This topic only has one round)

Detail Briefing:
Now that the influence of Internet celebrities has exploded, Xiaohongshu APP has become a favorite place for consumers to find products information, looking for articles about various products.
Many brands cooperate with famous KOL to introduce their products, conducting product trials and article promotion. Since the function of self-learning of AI has been developing, how can we use programming and AI intelligence to automatically write and identify articles that consumers love to read?

Mini Programs / APPs (Need to provide the total number of reviews after your own experiments)
Or solution PPT (including the main programming content)

Judging Criteria:
Practice: the total number of forwards, likes and comments earned by the articles written by AI;
the degree of how the score of an existing article published in the AI applet / APP is in line with the actual situation

Online Application

OEE (Overall Equipment Efficiency)calculation of production line

Detail Briefing:
According to time loss table of the production line and logic table of OEE calculation, write OEE (Overall Equipment Efficiency)calculation program using JAVA, C language or other languages.

The solution should include:
1. OEE calculation program (including writing source code)
2. Writing logic (explained in PPT or word document)

Judging Criteria:
1. According to the provided data and the time loss, calculate the equipment shift efficiency by using OEE calculation logic, and make sure the number is accurate to one decimal place
2. Clearly describe the calculation logic
3.Provide written source program for verification

Online Application

Social listening of pollution, stress for consumer & product insights (Category: skincare)

Detail Briefing:
Today consumers face many urban aggressors, such as air pollution, psychological stress, workload a problem. These stressors may alter consumer online behaviors such Weibo posts, online searches, product choice, preference or comments on e-commerce channels. These online data source contains could be capture and analyzed to find novel insights for developing novel personal care products to address these concerns. Your challenge is to explore how urban aggressors affects consumer’s needs and product choices through systematic collection and analysis of online data. And come up with anti-stress new skincare product business plan. You can refer to the challenge download deck for details.
For example, you can explore correlation with historical PM2.5/AQI time series data with Baidu, Weixin Index data for searching skin problems, products… Another example is stress levels may peak at certain time points of year (e.g. EoY assessment etc+, which can be captured in Weixin/Baidu index….

Please be aware that your proposal must be a novel data mining report, not a review or summary of existing studies. You need to write code to capture and analysis data by yourselves.

1. Deck of the Team background, skill set
2. Data mining report (understanding of the problem, data mining approach, results, insight)
3. business proposal (Proposal on new anti-stress skincare product and marketing plan)

Judging Criteria:
Round I:
1. Team and skill (10%)
2. Data collection and mining approach (30%)
3. Results and insights (60%)

Good data visualization will be a plus!

Online Application