Gladly
/Head of Design
The right information at the right time
To help one customer, an agent needs three pieces of information: who they’re talking to, what has already happened, and what to do next. Gladly’s workspace held all three — but not at the depth agents needed. Order details sat a level too deep, and agents dug for them mid-conversation while the customer waited.
Context
Gladly is a customer-service platform built on a premise the industry was still catching up to: conversations belong to people, not tickets. Voice, email, chat, SMS, and social all flow into one lifelong thread per customer. I joined as Head of Design in May 2021 — hiring and managing the design team while taking ownership of the surface agents live in all day: the workspace.
The workspace was where the premise met reality, and it was falling short in a specific way: agents couldn’t see the order and its history at the level of detail the conversation demanded. They could see that an order existed — not what was in it, what it cost, or what had already happened to it. Getting depth meant leaving the conversation. And inside the org, there was no shared way to settle layout questions: every “where should this information go?” debate started from scratch, opinion against opinion.
Approach
Named the zones before moving the pixels. I introduced a simple model: helping a customer takes three key pieces of information — the customer, the conversation, and the answers and actions that resolve it — and all three have to be visible at once for an agent to be efficient. “Zones” became the org’s common language for the workspace: layout debates turned into questions about which zone information belonged to, not contests of opinion.
Redesigned the workspace to the model. Customer context — profile, orders, relationships — holds the left zone; the full multi-channel conversation timeline holds the center; answers and actions dock on the right. Order details deepened from a bare order number to products, prices, notes, and history, so nothing essential required leaving the conversation.
Tested it instead of decreeing it. The zones redesign went into usability testing against competing directions, won, and shipped as the production agent workspace. The detail views kept iterating through internal feedback until the depth matched what agents actually needed — products as the focus, not order numbers.
Tradeoffs
Visibility vs. depth
Keeping all three zones on screen at once caps how much depth any one zone can show. The compromise was progressive disclosure — a compact order card in the customer zone that opens to full detail in place. Getting that detail hierarchy right took more iteration than the zone structure itself.
A language is a commitment
Once zones became the shared vocabulary, every new feature had to answer to it — and information that straddled zones forced the sharpest debates. That friction was the point, but it made some decisions slower before it made all of them faster.
Outcomes
Won & shipped
Usability tests
The zones redesign won in testing against competing directions and shipped as the production workspace — product NPS rose after launch
Opinion → Zones
Layout debates
“Where does this information go?” became a question with a shared answer instead of a recurring fight
“Ian’s network effect is remarkable. Other groups learn the importance of design and how to best work together. The result is a better product by teams that understand and enjoy building together.”
Joe Ciuffo, Director of Product Marketing, Gladly