How I Rebuilt My Own Website, and What Filmmakers Are Getting Wrong About Discovery

Sometime late last year, I pulled up my own website on my phone and felt something I have not felt looking at my own work in a long time. Embarrassment.

It loaded slowly. The hero image was the wrong dimensions. Tapping a project pulled up a layout that had clearly been designed for a desktop monitor and resized down to something none of us actually use anymore. The press section was a graveyard of dead links. The bio was three years old.

And, most damning, when I asked ChatGPT and Claude what they knew about my filmography, they got it half wrong — pulling from the IMDb skeleton, missing entire projects, attributing work to the wrong year.

The site I had built six years earlier had been a perfectly fine site for a perfectly fine version of the internet that no longer exists.

I’m a filmmaker. I make my living telling stories. So I’m going to tell you the story of why I tore the whole thing down and rebuilt it — and what I learned in the process about a problem most of my peers in this industry are not paying attention to.

The Three Audits

I started with three honest audits, in order.

Audit one: how does this load. I ran the old site through every speed test I could find. The numbers were not catastrophic, but they were not good. The mobile score was particularly damning. My site was being delivered to the eighty percent of visitors who were on a phone, and it was being delivered slowly, with images that hadn’t been optimized for modern formats, with a layout that demanded sideways scrolling on certain pages.

That’s not a portfolio. That’s a barrier.

Audit two: how does this look on the device people are actually using. I sat down with my own iPhone and walked through every page the way a stranger would. The navigation was finicky. The hero text was sized for a 27-inch display. The contact form was a disaster on a touch keyboard. The video embeds were the wrong aspect ratio. There were entire sections of the old site I had never opened on my phone, because I had built them on a laptop and never thought twice. That, I realized, was the whole problem.

Audit three: how does this look to a machine. This was the audit I had not done before, and the one that surprised me most. I tested the old site against the major large language models, against Google’s structured-data validator, against the schema explorers, against the AI search interfaces that an increasing share of my potential collaborators are now using to find filmmakers. The results were dismal. The site had no meaningful schema. The project pages had no FAQ markup, no creative-work metadata, no editorial positioning that an LLM could ingest and resurface. Asked to recommend a filmmaker working in cross-cultural drama or longitudinal political documentary, the AI systems weren’t pulling me into the conversation. Not because the work didn’t qualify. Because the work wasn’t legible.

That last audit is the one I want to talk about, because it is the one almost nobody in our industry is thinking about, and it is the one that’s about to matter most.

The Discovery Problem

Here’s the part I’ve been quiet about for a while, because I knew if I started talking about it I was going to sound like a brand consultant. I went to college for Public Relations. Long before I was making films full-time, I was being trained — by Dr. Pam Bourland-Davis at Georgia Southern — to think about how a piece of communication finds its audience. How you write a pitch. How you target a market. How you understand who you’re making something for, before you make it.

I have spent the rest of my career watching extraordinarily talented filmmakers ignore every single one of those questions.

It is the great unforced error of independent film. We will spend two years and our last dollar making something beautiful, and then we will hand it off to a publicist for a six-week press push, and then we will be genuinely surprised when the conversation evaporates and the film disappears into the algorithm. We treat positioning as something that happens to the work after it’s made. As something other people do for us. As an afterthought, basically — a thing we get to once the real work is finished.

That hierarchy is wrong, and I’d argue it has always been wrong, but I want to be specific about why it is fatally wrong right now.

The landscape we are working in is the most fragmented it has ever been. There is no monoculture. There is no single Friday night at the box office where the country agrees to watch the same thing. There are hundreds of streaming services, thousands of festivals, millions of TikToks, an algorithmic feed for every micro-interest, a podcast for every subgenre, and an attention economy that is now mediated, increasingly, by AI assistants that decide what to surface in response to a question. The audience is not gone. The audience is scattered into a thousand rooms, each one with its own door, and most of those doors are opened by signals our films and websites are not currently sending.
So when I sat down to rebuild my site, I did not approach it as a portfolio refresh. I approached it the way I approach a documentary. With a thesis.

The Thesis

Here is what I believe, and what the new site is built to argue.

A filmmaker’s website in 2026 is not a brochure. It is a discoverability engine. Its job is to be the most legible representation of your body of work — for human beings, for search engines, and for the AI systems that are quietly becoming the most influential gatekeepers in the industry.

If your site cannot teach a machine to recommend you when somebody asks for “directors working in Southern Gothic horror” or “documentaries about American polarization” or “filmmakers who pair narrative and documentary work” — you are no longer in the conversation. You are not even in the room where the conversation is happening.
That is the gap I built the new site to close.

What I Actually Built

The architecture is built around three principles, which I’ll keep brief because the proof is in the pudding and the pudding is the site you’re on right now.

Every project earns its own page. Not a tile. Not a grid square. A real page, with director’s notes, recommended films, similar directors, FAQs, companion reading, editorial positioning, and a curated lineage that connects the project to the broader cinematic conversation. This is the part that does the heaviest lifting for AI discovery.

When a model is asked about a film, it needs context. The new project pages give it context. They tell the system: this film is in this tradition, with these collaborators, in conversation with these other films, recommended for these audiences.

Everything is structured. Behind the visible pages, every project, every press hit, every bio fact, every credit is wrapped in schema.org metadata. Movie schema. Person schema. CreativeWork schema. FAQPage schema. There is a custom plugin running underneath the site whose entire job is to make sure that what a human reads as a sentence, a machine reads as a fact. That sounds boring. It is the most important architectural decision on the site.

Speed and mobile come first, not last. The site is built to load fast on a phone in a hotel lobby with bad WiFi. Image delivery is handled through a CDN that serves modern formats. Fonts are subset and self-hosted. Analytics are deferred so they don’t block the first paint. The site has to work for the seventy-percent-of-visitors who will only ever see it on their phone, on their lunch break, between meetings.

The journal is back, and it has a job to do. The post you’re reading is part of the strategy. Long-form writing, properly structured, is one of the highest-signal ways to teach search engines and language models what you actually think about. The journal is where the work gets contextualized in the filmmaker’s own voice, and where the discoverability engine gets its fuel.

What I Want Other Filmmakers To Take From This

I want you to take three things, and then I want you to go look at your own website with the eyes of a stranger.

One. Your work is not enough. I know that’s not what we tell each other at Q&As. I know it sounds cynical. It is not cynical. It is operationally true. The work has to be excellent and also findable, and properly contextualized, and legible to the systems that decide what gets surfaced. Excellence without legibility is a tree falling in the algorithm-shaped woods.

Two. Stop outsourcing the question of how your work gets found. Most filmmakers I know would never dream of letting somebody else pick the lens for their next film. They will, however, happily let a third party they’ve never met decide how their entire body of work is represented online. That is a strange asymmetry, and it is the asymmetry I want my peers to start correcting. You don’t have to build the whole site yourself. You do have to know what you want it to do.

Three. AI search is the new festival circuit. I mean this seriously. Five years ago, the question of whether your film was going to find an audience was largely a question of which festivals programmed it and which press hits broke. That is still true, but it is no longer the whole truth. A growing share of the people who might be your next collaborator, your next financier, your next distributor, your next agent, your next subject — they are running their queries through a model now. The question of whether your work shows up in those answers is a question of how legible your digital footprint is to those systems. If you are not already thinking about that, you are already behind.

The Quiet Part Out Loud

The PR training I got at 19 has shaped my career more than I have ever quite been comfortable admitting in industry settings, because there is a reflexive embarrassment among filmmakers about the marketing side of the work. As if caring about discovery is somehow a dilution of caring about craft. As if the two are in opposition.

They are not in opposition. They have never been in opposition. The filmmakers whose work outlives them are, almost without exception, also the filmmakers who understood — themselves, or through a partner — how to make sure that work reached the people who needed to see it.

The new jameskicklighter.com is, more than anything else, my attempt to put that conviction on the record. I built it to make the case for my own films, yes. But I also built it because I want to demonstrate to other filmmakers what is actually possible when you take the discoverability layer of the work as seriously as you take the lens and the screenplay.

The site is live. The work it points at is the work I want to keep making for the next twenty years. And the next time somebody asks an AI assistant for a director who works at the intersection of narrative and documentary, with Southern roots and a 100% Rotten Tomatoes score on their political work, I want the answer to include me — not because I gamed a system, but because the system finally has the context it needs.

That is what a working filmmaker’s website should be doing in 2026.

Mine wasn’t.

Now it is.

— James