Analysing 1,000s of images at scale for a global jewellery brand
Analysing viewers’ conversations to inform future marketing initiatives for the UK’s longest running commercial network.
Understanding needs
In order to help the client identify dominant and emergent themes within the jewellery market, we proposed objectively sizing semiotic codes over time and detected shifts in ‘out of category’ labels using outlier detection methods.
Connecting data
Sourced 1,000s of user generated and official brand images online each wave.
Created a data science pipeline to convert computer vision outputs (objects, colours, emotions) into semiotic codes.
Driving change
Saved STRAT7 semioticians hours of time manually analysing each images, thus saving time and cost for the client.
We were also able to add analytical rigour and objectivity to questions like ‘Which brands and codes are growing the most’ and ‘Where is white space opening up’?