Skip to content
Open call

Enhancing P&G’s trend spotting capability

Open date: Monday 4 March 2019   Close date: Friday 12 April 2019

P&G is looking for innovative solutions to replace its current, labour intensive, methods for data collection and analysis to help it predict future trends that might have an impact on the categories in which the company competes in.

P&G is an international company developing, producing and selling a wide range of fast-moving consumer goods worldwide. It is important for P&G to be aware of new technologies entering the markets that relate to the categories it competes in, and of social trends which may also have a big impact in product requirements.

For this challenge P&G is seeking innovative solutions to replace its existing (semi-manual) process for collecting data, performing trend analysis and predicting which trends might impact its business. Data streams will include patent applications, journals, news articles and social media (including blogs and forums).

The R&D team within P&G needs to be able to spot and react to upcoming important trends up to five years away. The information that is gathered is fed into product category teams, allowing P&G to improve existing products or develop new products in response to predicted trends.

Who should apply?

Startups and scaleups with a solution that has a clear route to market. Organisations must have the ability to build a minimum viable product (MVP) by September 2019. They must also have the capabilities to scale the solution across P&G if required.

Applicants should be available to attend the pitch day on 17 May at PROTO, Gateshead. Applications must be recieved in full by the application close date on 12 April.

Why you should apply

This is an opportunity to work with a small team of domain experts in a global organisation, to refine solutions in this area and test ideas in a live environment with access to large amounts of data. Following a successful pilot P&G would be interested in further discussions with the successful company to help explore other markets for this solution.

P&G would like to find a partner to deliver a suitable solution for this challenge. This is a fully funded commercial opportunity and does not require matched funding, with the intent that P&G plan to support the build of a prototype solution.

Additional info

P&G is looking for a solution that saves time over existing manual processes. The proposed solution must therefore save time when it comes to topic identification and categorisation. This process currently takes three months of part-time work.

Companies are welcome to apply for part of or the whole challenge

    Considerations to take into account

    As P&G is a global company, information could be in many languages including English, French, German, Chinese and Japanese.

    GDPR implications must be considered in relation to processing and holding data from social media profiles; for example the solution automatically updates if someone has deleted a post from their personal feed or triggered ‘the right to be removed’ under GDPR.

    Current solution

    If you are interested in applying to this challenge additional information can be requested which will include example datasets and a breakdown of the current job flow and outputs.

    Datasets of social media posts may range from a few thousand to hundreds of millions depending on the topic. One of the current datasets used has 645 million posts.

    The R&D Team at P&G currently uses programmes called Crimson Hexagon and Sprinklr to collect and analyse trend data. P&G is now looking for a system to take data from these programmes and suggest new categories to monitor, based on emerging trends. The system currently focuses on text-based data, however as there is a shift to consumers using social media platforms such as Instagram, it would be beneficial for solutions to be able to analyse both text and images.

    Evaluation criteria

    • GDPR compliance (pass/fail)
    • Sustainability (including upgrades & maintenance) (30%)
    • Quality of solution (low mis-classification rate and time saving) (40%)
    • Ease of implementation (30%)