top of page

Our Research

Solution built for jewelry business

We use prudent and scientifically-proven methods to help solve growth challenges in the jewelry e-commerce space. We practice techniques that are frequently used by Big Techs and leverage modern natural language processing and statistical modeling to help guide your business to achieve optimal outcomes on social media.


Stage 1: Data Extraction

We track more than 5,000 jewelry businesses on Instagram, ranging from international fashion icons, established jewelry companies, uprising DTC brands, influencer brands, regional and local jewelry stores and many more.


Based on our research, each new entry to our collection has to pass our eligibility criteria:

  • Active on social media

  • Substantial online/offline presence

  • Offer online purchase

  • Available in English

We refresh our collection every week to ensure more companies and brands are collected and the existing ones are made update-to-date. With each account, we leverage social media APIs to extract raw data and pre-process it for advanced data processing.


Stage 2: Data Processing & Analysis

With hundreds of thousands of data points in our curated database, we deploy a variety of advanced data analytical methods to extract insights and understand the competitive social media landscape for jewelry business.


The 5,000+ jewelry businesses are categorized into nine groups based on their follower numbers to approximate a bell-shaped distribution for reliable competitive analysis.

The analytical methods we have used include:

  • Traditional statistical learning

  • Time series analysis

  • Natural language processing and text mining

We provide summary stats for each jewelry group, such as group size, average followers, group follower growth rate, to present an overview of their social media presence at different levels. We aim to create a social media success recipe by continuously monitoring the best performing jewelry businesses and decomposing their content into repeatable steps.

We continuously fine-tune the correlations between post publishing time and post interactions, offering the potential of maximizing organic engagement when releasing content.

We leverage natural language processing and text mining to analyze key words, hashtags and emojis in both brand profiles and posts, to better understand their business types and track their social activities, such as giveaways and promotions.


Stage 3: Data Visualization

After applying multiple stages of data processing, we pipe our datasets into interactive data visualizations using Microsoft Power BI. The visualizations provide insights that can enable businesses to actively monitor competitor movement and help answer challenging questions like:

  • Who are the competitors that are similar to me?

  • Which jewelry businesses are trending on social?

  • When does a huge follower spike happen?

  • What does a highly “liked” or “commented” post look like?

The convenience of our data visualization and the flexibility to customize it for answering your own questions can provide a powerful tool to deeply understand and expand your own jewelry business on the competitive social media landscape.

bottom of page