Matthew Gilbert

An Economist Who Wanted to Be a Designer

The credential came first and the vocation second. Matthew earned an undergraduate degree in economics from NYU Stern School of Business, wonderful preparation for nearly anything except graphic design. Early on he worked for one of his professors, a McKinsey partner, and kept the habit for good: study everything every which way you can, and bring rigor before opinion.

A design opportunity redirected him in his early twenties, and several years at Siegel+Gale, the well known corporate and consumer brand consultancy, followed. When the School of Visual Arts created an MFA design program in his native New York City, he used it to reconcile the economics degree with the work he actually wanted. Then came pure Web 1.0: a startup that got its first round of funding and never its second, and a large company whose new division was an early precursor of brand design in a digital ecosystem, around 2001 and 2002. Those colleagues scattered when the sector crumbled. One of them called back in 2004.

Building a Practice Before Anyone Agreed on the Name

When the president of Bayard Advertising asked whether he could build an employer brand practice, Matthew said yes before he knew what the words meant. He knew exactly what a consumer, corporate, or B2B brand was. This was something else. His conditions were blunt: he was confident he could build it, less confident the company had the infrastructure to support it, and certain he would be asking for money, because nobody plants a flag and watches a practice appear overnight.

The first brand client landed within about three months, and it was consumer work, ground he already knew. The employer brand piece came out of research run the way his training demanded: never to confirm an answer he had already reached, but to learn to ask the question a different way. Start with protocol. One program became many, some of them global and comprehensive, running into 2023 and 2024.

Bayard held a place among the top three or four agencies in its category for a long stretch. In 2023 it was acquired by Appcast, part of the StepStone group, a player Matthew ranks third in the space behind LinkedIn and Indeed. As the practice drifted toward more functional career site and HR tech work, he left in 2025 and built his own firm so clients would not have to assemble consultancy rigor, consumer agency craft, and recruitment marketing domain expertise from three separate vendors. Founder led throughout and deliberately boutique, so he can take on small clients. He admits the launch was terrifying. He did it anyway.

One quarrel has followed him the whole way. He has never believed employer brand was the right name for the thing he spent two decades building, and he traces the arguments and confusion that still dog the field to Simon Barrow's coining of the term in the late 1990s.

The Discipline of Solving the Right Problem

"You have to solve the right problem, and it is not always easy or obvious," Matthew says. Nobody sets out to solve the wrong one, yet people do it constantly, and usually cannot tell until it is too late. So he treats the front end of a project as the expensive part. Sometimes the right problem is finding empathy for a workforce whose conditions he has never lived, at which point the honest move is to stop, say so, and do more research.

That discipline is the distillation of hundreds of people he declines to rank: every designer, copywriter, strategist, developer, and project manager he hired, coached, or mentored over twenty years, plus colleagues in client service, technology, product, and business development. Clients taught him as much. JPMorgan Chase, Lockheed Martin, Encompass Health, HCA, Carters, United Technologies, the American Red Cross. He worked alongside talent acquisition directors, employer brand leaders, CHROs, and chief marketing officers, and in some cases directly with chief executives, as he did at Cadabra. Creative debate gets loud inside an agency, he says, and the real skill is stewarding the disagreement so nobody walks out of a long argument hating anybody.

Neither Friend Nor Foe

Matthew writes regularly on technology and refuses to treat the past decade as one story. Before ChatGPT, machine learning was largely invisible, humming quietly inside applicant tracking systems. Then came COVID. If employee value propositions had been working, he argues, there would have been no great resignation, and no such gap between what an organization implies it is and what people find on arrival.

He is more generous about what that period unlocked. The shift from a universal nine to five office to hybrid and then remote work brought people caring for children and family back into the workforce, and let an employer in a small rural area of the Northwest extend real access to opportunity far past commutable distance. Then ChatGPT arrived two and a half to three years ago and changed the terrain again. Rarely has so much structural change been compressed into so short a window, and Matthew is unimpressed by the institutions meant to help people steer through it: "I am giving that a D minus." He points to reports that China will bar companies from replacing jobs with AI, and finds it striking that neither the United States nor the EU is leading there.

His own verdict is unsentimental. AI is neither friend nor foe; it is tools. Some are extraordinary force multipliers, others cost him hours of redos and revisions. Claude is his tool of choice, and he draws the boundary precisely: he handed it a messy Word document to convert into a spreadsheet, warning that perfection was not expected, and in under a minute the result came back nearly perfect. Ask the same model which of those items is highest value or most likely to close and it will get it wrong, because it is not thinking. It is very good at guessing. That is why he pushes back on the fashionable advice to let a language model be your thought partner, and why he draws the line on forecasts at a year or two out.

He built a career on a term he still argues with, in a category most people wander into, and he has never once pretended he arrived there by accident.