IoT for Retail: Measurence Scout – Weather-aware Foot Traffic Predictions and Analytics Dashboard

Scout is a location scouting solution by Measurence that helps physical businesses make business data-driven decisions. Using its own WiFi technology, Measurence is collecting data about the traffic of people in and around physical locations and providing analytical reports and business insights to the location managers.

In this scenario,  Measurence has developed – in collaboration with EW-Shopp technical partners – the new “Event and Weather Scout Dashboard” by adding an additional layer of information to this analytics: correlations with weather and events data.

“Event and Weather Scout Dashboard” created for businesses to optimize their ROI based on historical analysis and future predictions of customer flow vs seasonality and weather trends.

The service was built using machine learning algorithms and is able to predict the number of visitors for future marketing events 1-12 weeks in advance as well as understand the effectiveness of past events. So the marketing department can make decisions about the best time for marketing campaigns.

WATCH THE VIDEO: Intelligence solutions for the customer journey

KEY ASSETS

  1. Add layers of intelligence to Measurence Scout (correlations with weather and events data).
  2. Build vertical-specific products that enable clients to optimize their resources: sales promo activities; staff management; inventory & out-of-stock management. The key idea for these verticals is to integrate external datasets and provide in-depth analytics, analyze the effectiveness of the past marketing campaigns, and predict the number of visitors for the future campaigns and the best dates for its realization.
  3. Expand to new markets.

THE APPLICATION

The dashboard is an easy-to-use online tool that provides actionable analytics to Measurence’s customers to help them understand the relationship between their advertising spent, weather conditions, and foot traffic in their physical locations. For example, one of the top automotive marque uses Measurence’s product to analyze 50 of their car dealerships in Italy: they get daily, weekly and monthly data about foot traffic, marketing campaigns and weather conditions.

A predictive model-based on historical data and machine learning algorithm-estimates the number of visitors for future marketing events, helping managers analyze the effectiveness of their marketing campaings and improve their events planning. 

Fig. 2. Comparison of the number of predicted visitors, in case of the marketing campaign, and predicted visitors, on a regular day.

The dashboard shows:

  • People’s daily/weekly/monthly traffic in locations;
  • Past weather data and forecast for the next 5 days in relationship with people’s traffic;
  • Events information (date and type of the event) and prediction of the foot traffic for future events;
  • An estimate of the effectiveness in driving foot traffic of marketing campaign—both past and future events;
  • Comparison of foot traffic for different periods of time;
  • Comparison of weather temperature for different periods of time;
  • Comparison of foot traffic and weather for the same period of time but different dealerships.

ACHIEVED IMPACT

By developing “Event and Weather Scout Dashboard” we have achieved the following results:

  • The pipeline to enrich traffic analytics data with weather and event data and automatically populate the Measurence Scout dashboard has been deployed and now serves 50 car dealerships
  • Using events to estimate foot traffic has improved the accuracy of the model by at least  20% for the days with a planned event; as a consequence, the dashboard returns predictions that are aware of the event schedule, improving on previous results based on plain historical data analysis. In addition, the event-based predictive model supports what-if analyses, where different event schedules can be fed into the model to check the estimated impact of the different marketing schedules on foot traffic.      
  • One paid pilot request has been developed for a new segment — insurance; with this pilot, we are now investigating how the weather impacts on insurance for flexible workspaces

“Event and Weather Scout Dashboard” helps companies:

  • Better allocate their advertising budget: based on quantitative historical analysis of customer foot traffic vs seasonality and weather trends, companies can make better-informed decisions on how to invest their advertising budget to generate higher ROI.
  • Determinate the best timing for Marketing Campaigns: by applying event-based predictions to different kinds of marketing campaigns and different dates, marketing managers can decide the schedule for their events that maximizes the ROI of their marketing investments.

INDUSTRY IMPACTED

  • Automotive
  • Retail
  • Insurance
  • Out-of-home advertising
  • Commercial Real Estate

RESEARCH & INNOVATION DOMAIN: Data integration, analytical and predictive modelling, user interactive interface including visualizations

TOOLS INVOLVED: ArangoDB, statistical and predictive modelling were realized in python, JavaScript library highcharts

CONTACT PERSON: Olga Melnyk omelnyk@measurence.com – Alessandro Prioni  aprioni@measurence.com