An agency whose portfolio lives inside other people’s channels: Marketcolor
A London technical content agency working for finance, life sciences and technology — where the site had to prove thirteen years of craft without being able to show most of it.
The work is scripts, animation, reports and explainer video — and almost none of it is Marketcolor’s to publish.
Marketcolor writes and produces for Nasdaq, Deutsche Bank, BNY, SWIFT, Google, IBM, the European Commission and around thirty other institutions. The output is scriptwriting, storyboard design, post-production, copywriting, print and digital reports, eLearning series.
All of it ships under the client’s name, inside the client’s channels, often behind their login. An agency site in this position cannot run the usual case-study format — there is frequently no screenshot, no live link and no permission to describe the outcome in numbers.
The audience does not help either. A marketing lead at a bank is not browsing for inspiration; they are checking whether this agency has done something serious for someone serious, in their sector, recently. Three questions, and a conventional portfolio grid answers none of them quickly.
There was also a maintenance problem waiting. Agency sites decay because new work has to be designed in, so it stops being added — and a content agency with a stale site is arguing against itself.
Publish the ledger, not the portfolio
If every commission is one row — client, one sentence of scope, a date — then the site is a record rather than a showcase. It answers all three questions in a scan, and adding work becomes writing a line rather than designing a page.
The client’s name is doing the persuading. Everything else on the row is just proving it happened.
In this market, credibility transfers. ‘Scriptwriting, design and post-production for the Nasdaq Eqlipse animated explainer’ does more work than any amount of process narrative, because the reader already knows what Nasdaq’s standard is.
So the row was designed to carry exactly three things: who it was for, what was actually made, and when. No adjectives, no results claim, no case-study link — and consequently nothing that needs legal approval before it can be published.
One page, one reverse-chronological ledger, and a component that holds the standard as it grows.
The site opens by stating what the agency is and where it operates — a technical content agency in London, working 09:00 GST to 16:00 EST, for finance, life sciences and technology. Two sentences that qualify the reader before they scroll.
Below it the commission log runs continuously back to 2012. Long engagements carry a range rather than a point in time, and a commission that produced several pieces collapses into a numbered set inside its own row, so one relationship stays one entry instead of flooding the timeline.
The showreel and the agency’s own animated history sit at the top as the one place where the work can be shown rather than described, and contact is a phone number, an email address and a credentials deck — no form.
A commission is one row, and the row is the component
Client, one sentence of scope, a date. Nothing else.
Fixing the shape of the entry is what makes the page maintainable. Adding a piece of work is writing three fields, which anybody in the studio can do correctly, so the site keeps pace with the business instead of falling six months behind it.
It is also the quality mechanism. The craft lives in the type, the rhythm and the spacing of a repeated element — so it is inherited by every future entry rather than negotiated per project. There is no layout for a non-designer to get wrong.
Reverse chronology is the argument
Newest commission first, running back to 2012.
For most content sites date order is a lazy default. Here it is the point: the visitor’s real question is whether this agency is currently working, and at what level. The top of the page answers it before anything is read.
The depth answers the second question by itself. A log that keeps going for thirteen years is a claim about longevity that no About page can make as convincingly.
Multi-piece commissions collapse into one entry
A numbered set inside the row rather than eight separate rows.
A single engagement that produced twenty-five videos would otherwise bury a year of other work. Nesting keeps the timeline readable and keeps the unit of the page a relationship rather than a file.
The counter is shown plainly — a set of eight reads as a bigger engagement, which is information the visitor wants anyway.
Qualify the reader in the first two sentences
What we make, where we are, which sectors, which hours.
Naming the sectors filters honestly: the wrong client leaves quickly, which is the correct outcome for an agency this specialised. Stating the working window is a practical detail that matters enormously to a New York or Gulf client and appears on almost no agency site.
It buys the ledger its context. Without those two lines a list of banks is just a list of banks.
Contact is a phone number, not a funnel
Call, email, or download the credentials deck.
A contact form implies a queue and a qualification process. At this end of the market the person getting in touch already has a budget and a deadline, and making them wait for a reply is a way to lose them.
The credentials deck exists for the other route — the reader who has to circulate something internally before anyone picks up a phone.
A site that is a record of work rather than a presentation of it.
The commission log carries hundreds of entries across roughly thirty named institutions and thirteen years, and it stays current because publishing new work costs a sentence.
The standard holds structurally rather than editorially: whatever gets added inherits the same typography, spacing and rhythm, so the page cannot be eroded by the people who have to keep it alive.
What I would keep from this one
The constraint turned out to be the brief. Not being able to show the work forced a format that suits this audience far better than a case-study grid would have — a buyer at a bank wants evidence, not narrative, and evidence is a list.
Marketcolor has been redesigned since I worked on it, and the ledger is still the spine of the site. That is the part I would defend anywhere: on any site somebody else has to maintain, put the quality in a repeated component and let the content be plain.