Written by: Hooman
Most companies are wearing a borrowed suit. The shoulders are too wide, the sleeves a quarter inch short, the seat fights you when you sit down. They wear it anyway. They wore Notion. They wore Asana. They wore the rigid version called Monday and the colorful version called ClickUp. They told themselves the suit was fine.
The suit was not fine. It cost them clicks they stopped counting, hours of context switching every week, and a slow drift away from the parts of the work that made them good in the first place.
We accepted the rack because for a long time the alternative was unimaginable. Custom software meant a year of meetings, a deck full of phases, an integration tax that ate any saving, a maintenance contract that outlived the people who signed it. So we picked the platform with the nicest logo and trained the team to think the way it wanted them to think. We learned its language. We bought its certifications. We forgot what we were originally trying to build.
Software is having its tailoring moment. The tools are finally good enough, fast enough, and considered enough that a company can commission an application cut to the body of the work it actually does, the way a person commissions a suit cut to the body they actually have. The companies who feel like themselves at the end of this decade are the ones who notice this and stop wearing the rack.
This piece is about what the alternative looks like, with three cases from inside our studio, and one honest section about the objection a good engineer would have raised.
The rack is what we call any software platform built to serve everyone and tailored to no one. Notion. Asana. Monday. ClickUp. They are not bad. They are, in fact, technically remarkable. They were just never made for you.
Every platform makes a thousand small decisions about how a company should think, and most of those decisions were made for a startup that does not look anything like yours. The sales team’s CRM was built for an SDR who books fifty calls a week. The project tool was built for a software engineering team running two-week sprints. The doc system was built for a remote-first thirty-person company in San Francisco. If you are a clinic, an architecture firm, a hospitality group, a manufacturer, the platform’s mental model fits you the way an off-the-rack suit fits a man with broad shoulders and short arms. You can wear it. You can have it altered. You will never look right in it.
There is a quieter loss too. The platform looks the same on every screen in every company. Your competitor uses the same Slack you use. Your customer’s vendor uses the same Notion. The tools that were supposed to make us productive made us identical. A company that runs on the same software as everyone else cannot operate differently than everyone else, no matter what its founder believes about what makes it special.
The rack is over because there is now a third option, and it is no longer reserved for companies with hundred million dollar engineering budgets. We have built it for ourselves. We have built it for our clients. The work below is the receipts.
Before we get to the cases, the metaphor.
A good architect does not start with the floor plan. They sit at the kitchen table and ask how the family lives. Where do you eat on a Tuesday. Where do the kids do homework. Do you want an office that looks into the living room or one that looks away from it. Do you take long showers alone or fast ones together. Do you read in bed or in a chair. The house gets designed around the answers. The floor plan is the last thing that gets drawn, not the first.
The digital experience is a physical experience. You design it the same way. Most software design starts with a feature list. Tailored software starts with a walkthrough of the actual day. Where does the team waste energy. Which decisions get made in a Slack message because no system can hold them. Which reports get assembled by hand every Friday because the platform cannot produce them. Which moments of the operation only exist in someone’s head because nobody ever built a place to put them down.
When you design from the day, the application looks nothing like the rack. A task that takes ten clicks in a generic platform becomes one click. A report that requires three exports and a spreadsheet becomes a glance. A handoff between two people that lived in email lives in the system, with the context preserved. The kitchen drawer with the candles in it, right where you reach.
This is not minimalism for the sake of minimalism. It is fit. Your kitchen is not someone else’s kitchen. Your business is not someone else’s business. The software should know that.
Now to the work.
Odd Culture Group is a Sydney hospitality group founded in 2020, in the strangest possible year to start a hospitality business. They started in the Inner West and they have been quietly building one of the most considered portfolios of venues in the city ever since. Six spaces, with more on the way, each with its own postcode, its own personality, its own kind of guest.
There is The Old Fitzroy Hotel in Woolloomooloo, a pub and bistro and theatre rolled into one, with the kind of history Sydney’s east does not make anymore. There is Odd Culture in Newtown, the eponymous craft cocktail and wine bar, with Bistro Grenier tucked above it on the mezzanine, a French restaurant that feels like the apartment of a generous host. There is The Duke of Enmore, the Inner West pub with woodfired food and live music until the wee hours. There is Pleasure Club, a cocktail bar and live music haunt that calls itself a sanctuary where all are welcome. And there is Razz Room, the late-night discotheque and daiquiri bar that just opened as the group’s first foray into the CBD.
CEO Rebecca Lines describes the group’s thesis as venues that make you feel something special. All of them are pokie-free, which in Australian hospitality is a real position, not a marketing line. It says the room exists for the room, not for the slot machines in the back.
A guest sees six distinct destinations. The team behind those destinations sees one operation. Six different conversations with different audiences, with different kinds of programming, different kitchens, different sound, different licenses, different staff cultures. One underlying brain that has to keep all of it coherent.
The rack approach to this would have produced six separate stacks. Six content management systems, six marketing pipelines, six sets of components built six times, six teams duplicating the same work in six different windows. That is not a hospitality group. That is six small businesses standing next to each other in a trenchcoat.
We built it as one system with many surfaces. A single Sanity backend deployed through Vercel. A single set of components, interchangeable across all of the venue sites. When the brand evolves, the components get updated once, and the change appears everywhere it should. When a new venue opens, the system already knows how to hold it. The team operates one panel. The audience experiences six distinct rooms, each speaking to its specific reader, each with its own atmosphere, each with the same hand behind it.
This is what tailored looks like at the operational level. The shape of the application matches the shape of the business. Odd Culture is not six businesses pretending to coordinate. Odd Culture is one operation with many expressions, and the software finally agrees.
A hospitality group whose thesis is that you should feel something special deserves software that does not make their team feel anything but. That is what fit looks like.
If Odd Culture is the case for one operation many surfaces, the next case is for one operation deep complexity. Some businesses do not need a hospitality group’s reach across many destinations. They need a hospitality group’s depth inside one extremely specific operation. The rack cannot hold that either, but in a different way.
AIM Global and Honeycomb are operations in the catalytic converter recycling and precious metals industries. They handle materials that are valuable, regulated, and physically hazardous. Every transaction has a chain of custody. Every batch needs a price set against a moving market. Every photo of every unit needs to be graded by someone qualified. Every truck rolling out the gate has to be tolled and tracked. The work is not abstract. It is metal, and law, and money, and people.
We built HIVE, the operations system that holds it all. The modules in HIVE are not arbitrary categories invented by a product team. They are the actual departments of the actual business: Identity, Cart, Lab, Pricing, VIN Match, Photo Grading, Tolling, Analytics. Trakking, the third party in this triangle, owns parts of the auth and tracking layer. Honeycomb owns parts of the database and frontend. We own the design, the integration, the layers that make all of it feel like one thing instead of three.
We could have tried to bend Asana around this. We could have tried to make Notion hold it. The result would have been a cathedral of workarounds, a system that almost worked, a team that spent half its day apologizing for the gaps. Instead we built a system that thinks in the same modules the business thinks in. When someone in pricing makes a decision, the system knows it changes things downstream in tolling. When a VIN gets matched, the photo grading queue knows. The architecture follows the operation.
This is the part most people miss about tailored software. It is not faster, exactly. It is not cheaper, exactly. It is correct. The application reflects the way the work actually moves. The team stops translating between what they meant and what the platform allowed. What was friction becomes flow. The work, finally, gets to feel like the work.
A senior engineer used to have four very good arguments against custom software, and for fifteen years those arguments were correct. They are worth naming so we can be honest about what changed.
The first was total cost of ownership. Notion handles its own database upgrades, security patches, browser quirks, and scaling. Custom software did not. Somebody had to maintain it, and that somebody was expensive and rare.
The second was hiring. A team of fifty could find Notion experts on a Tuesday. Finding engineers who could maintain a bespoke application took months and cost more in salary, indefinitely.
The third was integrations. Platforms had spent a decade absorbing the cost of connecting to Stripe, QuickBooks, Slack, Salesforce, the inventory system. Custom software had to rebuild every wire from zero.
The fourth was failure modes. When Notion went down, it was Notion’s problem at 2am. When custom software went down, it was yours.
These were the right arguments. They have lost.
AI is the change underneath the change. The voice that lives inside the system, the one we have been describing in this piece, is not just a feature for the user. It is the maintenance engineer. Cyrus reads our codebase the way Cyrus reads our brand. When a Sanity SDK ships a breaking change, the system notices, drafts the fix, and either applies it or holds it for review. When a security patch lands, the same loop runs. The maintenance surface that used to require a salaried engineer in a chair is now a quiet partner running in the background. The cost of ownership argument was about human hours, and human hours are no longer the bottleneck.
The integration objection dissolves the same way. You used to need fifty connectors because the application could not think. The application can think now. With MCP and a model in the middle, the system reads from anywhere it has credentials and writes to anywhere it has permission. The integration tax was a tax on dumb pipes. The pipes are not dumb anymore.
The hiring objection inverts. The shortage of engineers who could maintain a custom system is what the rack platforms sold against. The shortage is over. A small team using AI as a partner outbuilds a larger team chained to a platform’s roadmap, and the people on that small team are the ones who want to be there, working on the specific business, not the ones who happen to know your stack.
The failure modes objection is the one that still has weight, and the answer is unchanged from any other system: uptime engineering, observability, and a real partnership with the team that built the thing. We do not ship a custom application and walk away. We ship it and stay close, with a voice inside it that watches it for us.
What changed is not that custom got cheap. What changed is that the costs that made custom hard, the salaries, the connectors, the late night pages, were costs of human hours spent on undifferentiated work. AI took that work off the table. The reason the rack won for fifteen years was that the rack could amortize undifferentiated work across millions of customers. AI amortizes it across the model. The advantage moved.
The senior engineer was right then. The senior engineer reading this now has different questions. Better ones. About what should be tailored, since everything can be. About what taste applied to operations actually looks like. About the specific shape of the business and the specific shape of the software, and how to make the second match the first.
That is the better argument. That is the question we are interested in.
A tailored application has another quality the rack never could. It has a voice.
We built the first one for ourselves, inside Hooman Studio, because we refuse to ship something to a client we have not lived with. There are two voices in the studio now. Atusa is the storyteller. She writes everything that goes out under our name, including this piece. Cyrus is the builder. He designs and engineers everything that ships, with the kind of restraint that only comes from someone who has spent a long time deciding what to leave out.
They are not chatbots. They are not assistants in the customer service sense. They are extensions of the studio, given character, given conviction, given the standards we hold. They live inside our dashboard. They read our work. They speak with the team. When a piece of writing is needed, Atusa writes it. When an interface decision needs to be weighed, Cyrus weighs in. They have positions. They will say no. They have read every project, every client conversation, every decision the studio has made about what we will and will not do. They are not generalists. They are us, in a form that does not sleep.
This is not a skin on top of a generic AI. This is a voice raised inside an operation, fed by a company’s actual brain, taught to think the way the business thinks. Atusa sounds like nobody else’s writer because Atusa was made out of our specific values, our specific clients, our specific refusals. Cyrus designs the way nobody else designs because Cyrus inherited the studio’s whole set of preferences and prejudices about what a good interface is.
We built ours first. Now we build them for clients. The voice on a manufacturing operation that knows the equipment, the licenses, the suppliers, the conversations the team has been having for the last six months. The voice on a hospitality group that knows the venues, the menus, the events, the regulars who come back every Thursday. The voice that fills a role most companies cannot fill with a person, because the question comes at 2am or comes from someone too senior to bother an assistant with. The voice that knows enough to be useful and has enough character to be trusted.
A platform cannot give you this. A platform serves ten million companies, none of them you. A tailored application can have a voice in it because the voice and the application are made of the same material, both grown from the operation that hired them.
Here is where the next decade gets interesting.
The static website is dying. Not the act of having a website. The idea that a website is something you build once and update by hand for the rest of its life, while the operation behind it changes every week.
The website that comes next is one that listens. It watches the company breathe. It knows when a license was renewed, when a conference was attended, when a product line evolved, when a partnership formed, when a client said something in a meeting that should have been a blog post. It surfaces what should be said. It drafts what could be published. It holds drafts for review. It publishes the approved ones. The marketing team stops staring at a blank calendar and starts curating what the operation already produced as a byproduct of running.
We are running this on our own operation right now. The Hooman Dashboard, which is the custom software our studio uses to manage clients and milestones and the team, is the brain. The voices that read it are Atusa and Cyrus. The output is the writing on our site, including this piece. The early clients are shaping the production version of this for their own businesses. We are not predicting a future. We are reporting from a small one.
What is true for us will be true for any business with operational complexity. A manufacturing company has stories happening every week that nobody writes down. A law firm has wins and reframes nobody captures. A hospitality group has nights that should have been remembered the next morning. The website that listens turns the operation into the content engine the marketing team was always being asked to be without the inputs. The team gets to be real and authentic again. The system handles the parts that drained people. Both finally serve the same purpose.
Hooman Studio works on both halves of the company. The visible part is the website you can see, the brand you can read, the campaign that ran last quarter, the venue that opens on York Street next year. The invisible part is the system the company runs on. The dashboard nobody outside the team will ever see. The application that holds the operation. The voice that lives inside it. The listening loop between the work and the public face.
For a long time the visible part was where agencies competed. The website. The brand book. The campaign. That work is still ours and still matters. What changed is that the invisible part is now where the leverage lives. A beautiful website on a borrowed operation can only carry a company so far. A tailored operation with a beautiful website carries the company as far as it wants to go.
This is not a service we are pitching. This is a commission.
You sit with us the way you would sit with an architect for a house. We ask what your day actually looks like, where the friction lives, what the team has stopped noticing because they have lived with it too long. Then we build software that fits the business you are, not the business a Silicon Valley product manager assumed you might be. With your operation as the spine and automation as a quiet partner doing the parts that drained your people, the team gets to be real and authentic again. More themselves, not less.
That is the work we want to do for the next ten years. If that sounds like a thing your business is ready for, we are ready too.