Duke Brand Guide Overhaul

Building a new Duke University brand guide for a diverse, decentralized user base

UX research • brand systems • documentation • web accessibility

duke-brand-home-all-devices.png

Overview

Duke University houses hundreds of schools, institutes, departments, programs, and initiatives — each with its own identity and communications strategy. The Duke brand guide provides a hub for these programs to access centralized brand standards and recommendations on sub-branding their projects. In fall of 2018, I led an large reimagining of the guide to meet our users’ needs.

The Problem

Before this overhaul, the Duke brand manual existed in a fragmented trinity:

  • A print design-focused binder that could only be accessed by a select few

  • A single page style guide applied to the homepage (and meta brand bars to be applied as frames to other pages) without many guidelines or flexibility

  • Additional caveats and pointers that lived in the heads of our brand officers

Both Duke and its audiences quickly outgrew that system.

What began as a handful of university websites and departments with dedicated designers grew into scores of sites, campaigns, and apps governed by a decentralized network of communications professionals with diverse needs. Whereas earlier, it was easy for the central brand officers to police the implementation of the Duke brand across a few larger entities, new technologies summoned the genie from the lamp: web fonts, online graphic design apps, open-source CMS solutions like WordPress, and even smartphones considerably lowered the barrier for anyone (or any school unit) to create and manage their own identity and brand assets.

All of these conditions combined made it incredibly difficult for the branding team to keep the different school programs compliant with Duke branding. New logos and websites would constantly pop up that increasingly disconnected from the Duke brand. Rather than needing one manual that addressed a system, Duke now required a flexible framework that met the needs of a growing ecosystem of diverse contributors.

The old style guide — outdated, clutter, and not very mobile-friendly

The old style guide — outdated, clutter, and not very mobile-friendly

The People

Our users

We identified 3 distinct categories of users with varying skillsets and levels of internal access:

  • Internal graphic designers and developers

  • Internal communications professionals and program managers who either have to take on some design responsibilities or communicate with interns and external vendors who are fulfilling their design needs

  • External design, development, and print/production vendors (both freelancers and agencies)

Our stakeholders

Our primary stakeholders were the decision-makers who were accommodating branding ambitions within their program budgets and calling the shots for the users:

  • senior administrators

  • school deans and department chairs

  • program directors

Upstream from them were a secondary group of stakeholders who applied indirect influence rather than direct control over branding:

  • engaged alumni and donors

  • prospective students (and their caretakers)

My Role, the Team, and the Timeline

I led the overhaul of the brand guide, which involved conducting stakeholder and user research, design, front-end development, content/resource production, and writing documentation.

To gather the necessary input, I worked with our digital strategy director, our web services team (particularly their creative director and visual designer), and three photo journalists. These team members often lent me ideas and resources that would go into the new guide or helped in reviewing components of the site.

Also, I shouldn’t overlook that the other front-end developer on my team bailed me out in a few tight spots as we closed in on the project deadline.

This project was done completely in-house with zero budget. Throughout the fall of 2018, I worked with various programs to test out color palettes and new sub-branding treatments. After initial validation of these across a few projects, I had about a five weeks to produce the new online guide in parallel with a handful of other large design projects I was working on at the time.

The Design Process

Identifying 5 critical failures through initial user interviews

To get an idea of what resources would be useful for the new brand guide, I met with designers and communications professionals throughout Duke to understand their needs and pain points with the old style guide system.

A set of five overarching pain points emerged.

 

A confounding color palette

Users were frustrated with massive inconsistencies in the color palette when working in different color spaces. They felt many of the colors were also quite garish and overpowering. Many of them only used a subset of colors in the palette that they felt were reliable and useable.

Irrelevant co-branding

Users felt the options available for co-branding their programs and units were too restrictive. They wanted ways to incorporate new glyphs into their logos and they needed compositions that translated well to social media. With so many shortcomings, the decision-makers our users reported to often encouraged them to ignore the guidelines.

Restrictive fonts

Garamond 3 Ltd and Interstate, the two typefaces mandated in the old style guide, were difficult for most users to implement on the web. Many users had already moved away from these fonts and wanted recommendations for free, web-safe alternatives they could use.

 

Outdated resources = noise

Users found many of the digital resources in the style guide were very “Web 2.0” and not responsive. Others noted that the PowerPoint templates were low quality and no longer useful. Most users didn’t see the need for many of these resources and mostly unconsciously skimmed over them unless they were asked about them explicitly. Overall, users felt that the style guide was dense, noisy, and took time to navigate.

An absence of photo guidance

The number one demand that users expressed was greater access to university photography. They needed it for their websites, social media, publications, and e-newsletters — and slapping in thumbnails from Google Images was no longer going to cut it. Other users who sat in more strategic positions wanted guidelines for capturing photography that would give a Duke look or tone to their products.

Compiling The Feature List

From these interviews and discussions, I was able to distill a wish list of content and features desired by both our users and stakeholders:

  1. Less noise; more breathing room and organization

  2. New logo files for the Duke wordmark (including new web-supported formats such as PNG and SVG)

  3. New co-branding guidelines with additional options and templates

  4. Information on phased out or restricted university trademarks

  5. A new color palette with:

    • Color values that yield enough combinations that can satisfy WCAG 2.0 level AA color contrast ratios

    • Better consistency between hues in RGB, CMYK, and Pantone color spaces

    • New complementary colors that aren’t eyesores

  6. A quick way to identify accessible color combinations

  7. A new suite of recommended fonts

  8. Guidelines for web font implementation and font pairings

  9. Resources for free on-brand photography

  10. Guidelines for creating brand-compliant photography

Using this feature list, I conducted a content inventory between my team’s resources vs. resources our web services team had built out. Those assets (mostly colors and web fonts) became the foundation for the new brand resources.

The new extended color palette on desktop and mobile

The new extended color palette on desktop and mobile

Sprinting Towards An MVP

I started by drafting the overall structure and information architecture of the site. One of the chief insights I had heard from users was that the old style guide was too dense, cluttered, and seemed like a transactional afterthought. To counteract this, I divided the content into pages with a main navigation that I validated with some peers through card sorting exercises.

With a compressed project timeline, I decided to build the site within an existing customized WordPress template we had in-house. This decision allowed me to focus on developing the content and stage things quickly to put in front of users and stakeholders to observe their reactions and interactions.

A mid-way draft of the information architecture for the site.

A mid-way draft of the information architecture for the site.

Deep Dive

Iteratively Designing The Typography Section

The final design of the Typography page

The final design of the Typography page

The typography section was among the most iterated upon pages in the new guide. Initially, I started with a list of our recommended fonts (with specimens and download links).

When I put this font specimen sheet in front of users, many of them requested more guidance on using each font. I included recommended use cases for each font to accommodate this, and I added a new section on different sample font pairings.

A set of somewhat more advanced users wanted more direction on how to implement these fonts on the web. These users weren’t developers, but they were advanced enough at using their website CMS tools that they knew how to drop in strings of HTML or CSS code if given a snippet. To meet the needs of these “smart enough to be dangerous” users, I added subsections within each font listing and font pairing suggestion that gave a few high level steps on how to link the fonts into their websites.

With so much content on this page — we had nine font specimens and four example font pairings — I found that users were a bit overwhelmed when the typography page would first load and the scrollbar on the right would grow smaller and smaller as the page fully rendered. Many users would take a long while to scroll and navigate around the page until they found what they were seeking. And even though most were looking for font pairings rather than reading through the individual font specimens, they would get tired before scrolling down to that section — even after I created a shortcut to it at the top of the page.

To address this, I packaged much of the content into keyboard-accessible accordions and tabbed sections so that the font information could be displayed progressively. This way, users visiting the page could scroll down quickly, see the first font pairing, get that confirmation that they were in the right place, and then feel encouraged to explore further. Despite most of the non-heading content being initially hidden under this progressive disclosure model, most users felt less overwhelmed and spent more time examining and engaging with different options on the page rather than succumbing to content fatigue.

The initial layout (left) vs. the updated, progressive disclosure layout (right)

The initial layout (left) vs. the updated, progressive disclosure layout (right)

One feature goal that I did have on the drawing board was adding brief previews of the fonts in a card-style layout to make them easily scannable without having to click into each accordion. However, due to the timeline and the MVP accordion structure being good enough for the vast majority of users, we deprioritized this feature so that we could shore up other areas of the brand guide in time for launch.

The final resource on this page was a last-minute addition. Late in the revision process, we realized that some users might still need occasional access to the old fonts to make infrequent updates or edits to older archival projects. We added a font request form for these users at the very bottom of the page so that they could gain appropriate licensing to the discontinued fonts.

Creating A Flexible Sub-Branding System

The final design of the Branding page

The final design of the Branding page

Perhaps the greatest gulf in the old style guide was maintaining a brand system that was flexible so that individual units could be creative and establish a sense of identity, but also was structured enough that their sub-brands reinforced the overall Duke brand rather than diluting it.

During my interviews with users and stakeholders, two predominant desires for sub-branding emerged over and over again:

  1. Logo compositions that translate well to social media — especially within the confines of a profile photo thumbnail

  2. The ability to include glyphs/icons that were representative of their program into their logos.

To address this, I developed a new sub-branding system that still emphasized the primacy of the Duke wordmark, but also provided greater flexibility for individual units to get creative with their sub-branding.

The old style guide sub-branding treatments.

The old style guide sub-branding treatments.

Examples of the new brand guide’s sub-branding. Still structured, but with more room for variation.

Examples of the new brand guide’s sub-branding. Still structured, but with more room for variation.

Launch & Results

I presented the new brand guide website at our monthly communicators meeting in December 2018 (about a third of our 200+ university communications professionals were present), and they were quite pleased to see the new resource and eager to use it in new rebrands, website redesigns , and other projects they had under way. By January 2019 we had fully launched the new site and made it publicly viewable.

The Duke Chapel website redesign incorporated the updated color palette and typography

The Duke Chapel website redesign incorporated the updated color palette and typography

Duke units/programs that revised their sub-branding to comply with the updated guidelines

Duke units/programs that revised their sub-branding to comply with the updated guidelines

The greatest measure of success for a project like this is not about transactional data like visits or downloads. Rather, it’s revealed qualitatively by seeing greater compliance integrated into larger branding & communications projects undertaken by these communicators. Although I left Duke in January of 2019 (and therefore was unable to witness much of the subsequent impact), there were already a number of communicators with established brands or projects who bought in to the new brand guidelines and took steps to pivot or apply the new guidelines in their final product. These units were sprinkled across Duke in different schools and institutes that we had difficulty ensuring brand compliance with in the past, which demonstrated that the new guidelines were clear and flexible enough that they found value in attaching their departmental brands to it.

 

Up Next: Design Deliverables

Snapshots of other design assets I have produced for teams, stakeholders, and clients

UX research & design • customer experience & service design • analytics & automation