To Sherwin-Williams, color is personal—it’s human. Sherwin-Williams Coil Coatings has partnered with architects for 150+ years to bring their visions of the perfect building to life. With the world’s largest color set of 50,000+, Sherwin-Williams has proven that they can create any color imaginable for their customers.

But in a world of endless color, how do you find THE ONE ? "Red" might mean crimson to some, yet a bright apple red to another. Historically, finding the perfect color meant paging through physical color chips, online catalogs or working with a Sherwin-Williams Color Technician— all underwhelming tasks for such an emotional topic.

We created a solution that brings together data, language, and innovation—opening architects up to a world of color they never thought possible. We’ve spent years solidifying Sherwin-Williams as the color innovation experts, and knew we could bring architects a bold new way to experience the color selection process.

Speaking in Color is a React web app that allows the customer to use natural language to find their perfect color. Describe a color via a phrase, a memory or stream of consciousness and our NLP machine will identify interpret that color and then allow the customer to tweak it as desired.

We use a combination of 3rd-party and integrated proprietary code to deconstruct spoken or typed language into component parts of speech and use the resulting output to run an image query on Google’s custom search platform. We then use a version of Google’s Vision AI to derive the composite colors of the images returned collectively, and force-rank the most likely colors according to volume in those images. Our custom algorithm then allows users to finesse the color in a way that transliterates statements like “make it dimmer,” “add warmth,” or “more like the 1980s” into mathematical RGB adjustments.

Essential to Speaking In Color's user experience was its ability to start the color selection process from anything that inspires the architect. To accomodate this, it required an infinite data set to query and build results from. The vastness of imagery available online provided the source. Our job was to distill that query into a useable set. This was achieved through image analysis and machine learning to interpret the HEX color data into the most common hues to form a palette.

Once the palette was was created and the architect chose a starting color, they needed to modify it. Here again imagery from the internet was used to interpret what is meant by very subjective terms, like earth, misty, moody, etc. Returning a primary hue, contrast and chroma values, that our AI would combine with the previous color. This process repeats until the user finds there one true color.

Sherwin-Williams set out to create a more meaningful and personalized approach to finding the perfect color for architects. This data driven approach proved so inspiring, challenging centuries of how color was identified and created, that it’s utility and experience transcended the B2B segment for which it was designed.

It's human centric experience, is now being rethought and engineered for the Sherwin Williams consumer markets. Allowing its retail paint store customers to discover, formulate, mix and acquire the one in a million color that they’ve dreamt of.