Grips Research
Discover the impactful applications of Grips data in cutting-edge research worldwide, revolutionizing industries and driving innovation to new heights.
Projects
ROAS Management (Ongoing)
Strategically managing Return on Ad Spend (ROAS) can enhance profitability. We analyze 500+ advertisers and show that a simple ROAS-based budget reallocation can increase sales by 5%-25% without additional ad expenditure.
Product Embedding Alignment (Ongoing)
Multi-source product embeddings often disrupt training, reducing performance in tasks like product recommendation. We propose a method for aligning product embeddings from diverse sources, as demonstrated using product embedding for a UK and Irish grocery shop.
Beyond Rankings: Exploring the Impact of SERP Features on Organic Click-through Rates (May 2023)
Supplemental SERP features, including knowledge graphs and video reels, significantly impact CTR prediction. Our analysis of 100,000+ keywords suggests a 20% increase in prediction accuracy using these features.
Interpretable Deep Learning for Forecasting Online Advertising Costs: Insights from the Competitive Bidding Landscape (Feb 2023)
Cost Per Click (CPC) prediction is essential for marketing strategy planning. Our study of 12,000 advertisers shows that Dynamic Time Warping can identify indirect competitors, improving long-term prediction. For example, data from flour sales can improve predictions for kitchen appliance marketing.
Russia’s e-commerce trade in the aftermath of the 2022 invasion: Evidence from high-frequency data (May 2022)
Despite many brands leaving Russia following the 2022 Ukraine invasion, retailers remain. Our analysis of parallel imports and 100+ brands' cross-border e-commerce sales helps estimate the effects of public pressure on companies to cease Russian trade.
Online Advertising Revenue Forecasting: An Interpretable Deep Learning Approach (Dec 2021)
Most advertising performance prediction research focuses on the advertiser's perspective. Our study, covering 200+ news websites' ad revenue, highlights the benefits of training a temporal fusion transformer on all available data.
Teaching an old fox new tricks: How sustained e-commerce sales lift from pandemic lockdowns depends on age (July 2021)
COVID-19 lockdowns spiked online sales, but customers under 50 returned to baseline shopping habits two months later. The 50+ generation, having adapted to online shopping, maintain high sales levels post-lockdown.
