How do you price this if there are no prices?

Fixed price per engagement, in writing, before anything starts. We don't publish a rate card because a brand identity and a content engine aren't the same job, and a made-up number is one we'd have to walk back. What we will do is save you the discovery theatre: tell us your budget range early in the first call and we'll say in that call whether it's in range. The quote follows shortly after, and it doesn't move unless the scope does — if we underestimated the work, that's ours to eat, not a change order.

It depends on the job, and we'd rather scope it than guess at it on a web page. A landing page or a campaign set is a short piece of work; a brand identity, a product interface or a search rebuild is a longer one. Real dates go in the written scope, and if we're going to miss one you hear it before the date, not after.

Who actually does the work?

You'll have the names before you sign, not after. We're deliberately small — the people who designed and built the AADDYY platform are the people on your account, and we bring in a specialist by name when a job genuinely needs one, telling you which parts they own. Nobody sits between you and the person making the thing. If what you're really asking is how many hands are on this: ask on the first call and you'll get the actual number.

Is this just AI output with a markup on top?

No. The tools do drafts, variants, resizes and production grunt work. A human sets direction, edits, and signs off on everything that reaches you. Where a deliverable is machine-generated end to end — bulk alt text, transcript cleanup, image resizing — we say so in the quote, and it is priced like the commodity it is.

You do. Ownership transfers as invoices are settled rather than in a lump at the end — so if the engagement stops partway you own what you paid for and there's nothing to chase. Files, repos, domains and accounts are in your name from day one. We'll ask before showing any of it publicly, and no is a complete answer that costs you nothing.

Either side can end a project engagement on reasonable notice, set out in the agreement. You settle for what was delivered, everything produced to that point is already sitting in your accounts, and there's no termination fee or clause that makes leaving more expensive than staying. If you'd rather bring the work in-house, say so early — training the person taking over is part of the engagement, not an upsell.

Is there a minimum engagement?

No minimum for project work — one landing page is a real job and we'll quote it as one. Partner engagements run for an agreed term, because the compounding only shows up with enough runway to be worth your money and our attention.

AADDYY Studio · Done for you

Same tools. Our hands.

Studio is the team that built the AADDYY platform — 50 AI tools, an npm SDK and a public MCP server you can install before you reply to this page. Design, engineering, search and social, run on that tooling, on an account in your name.

An agency bills people to do what our tools already do.

Drafts, variants, resizes, alt text, transcripts and first cuts run through the tools, so you are not billed hours for them. What you are paying for is judgement — strategy, art direction, code review, and the call on what actually ships. The quote itemises which lines are machine work priced as a commodity and which are people, so you can argue with the split instead of guessing at it.

Every engagement lives on an AADDYY account in your name. You see the tools we used, the outputs they produced and what each run cost. Nothing is a black box you pay extra to look inside.

The distance between brief and first draft is short because drafting is automated. That is the entire trick. Ask us to open the pipeline on the first call and we will.

Design

Identity, product interfaces, and campaign creative that ships as a system rather than a folder of exports.

Brand Identity System

Identity that ships as code, not a PDF.

We generate mark, lockup and palette directions in volume on our own image tooling, then cut hard. What survives becomes a working system — tokens, type scale, spacing, contrast rules — instead of a long book nobody opens. You leave with a Figma library and a token file your engineers can import the same day.

For — Companies whose product outgrew their first logo

Product & Interface Design

Every state designed, not just the happy path.

We design the flow, the screens, and the states most decks skip. Every component ships with a spec: variants, breakpoints, focus and hover behaviour, motion timing, and the token it binds to. Engineers build from the spec instead of guessing from a static frame.

For — Product teams where engineering outruns design

Campaign & Content Creative

The last variant costs what the first did.

We build the template system first — grid, type lockups, crop rules, motion beats — then run volume through it: static sets, ad variants, short-form cutdowns, every ratio a platform asks for. Our image, product-photo and video tools sit behind an API, so scale is a script rather than a night of manual exports. A human sets the direction and approves the final cut.

For — Brands running always-on social and paid with few hands

Engineering

Production web front-ends, mobile clients, and the AI systems behind them — metered, queued, and handed over with a runbook.

Web Development

Next.js front-ends that render on the server.

We build marketing sites, dashboards, and logged-in product surfaces on the Next.js App Router, with a typed API layer behind them instead of fetch calls scattered through components. Every page ships server-rendered metadata, JSON-LD, and a canonical URL — a page search engines and agents can't read isn't finished. Design tokens live in code, so the second page costs less to build than the first.

For — Teams whose site now has logged-in surfaces the template can't hold

App Development

One React Native codebase, plus the API behind it.

Cross-platform apps from a single React Native codebase, plus the API they talk to — we take both sides or the seams leak. We design the client around a submit-poll-download job contract, so work that takes minutes survives a locked phone and a dropped connection. Store submission, release signing, and crash reporting are in scope, not phase two.

For — Web products that need a real mobile client

AI Integration & Automation

Custom agents, MCP servers, and the plumbing underneath.

We wire AI into systems you already run: agents that do a specific job, MCP servers so your internal tools appear inside Claude and Cursor, and automations that retire a recurring manual process. Every integration ships with a metering layer — cost per call, per user, written to a ledger — because unmetered AI spend is how these projects get cancelled. Prompts, model choice and price live in config, so tuning is a row update, not a deploy.

For — Teams with a manual process and unused API keys

Search

Rank in Google, get cited by the assistants, and keep publishing after launch.

Answer Engine Optimization

Get cited inside ChatGPT, Perplexity and AI Overviews.

Assistants answer from structured, machine-readable sources — not from your homepage prose. We rebuild the site into that shape: llms.txt and the surrounding discovery files, a linked JSON-LD entity graph, per-agent crawl directives, and one canonical answer sentence per page reused across metadata and schema. Then we watch which assistants cite you, and which ones cite a competitor instead.

For — Brands losing clicks to AI summaries of their own category

SEO

Rankings that survive the AI Overview.

We audit what a crawler actually sees: title and meta status, heading structure, viewport, schema coverage, internal linking, referring domains. Then we fix it and write the pages that earn the position. Ranking still sends the highest-intent traffic, and assistants build their answers on top of what already ranks — so this is the floor, not the ceiling.

For — Sites with real pages and flat organic traffic

Content Engine

A publishing pipeline, not a freelancer roster.

Two lanes, one pipeline. Editorial pieces get a named angle, a real research pass and a human edit; programmatic pages get generated from your own dataset — catalog, locations, comparisons — each with its own canonical, schema and answer capsule. Both run on a schedule, dedup against what you have already published, and ping the index the moment they go live.

For — Teams that need publishing volume without repeating themselves

Social

Short-form volume where attention is, and founder-voice writing where the buyers are.

Instagram Page Management

A designed grid, and reels that keep coming.

We build the visual system first — grid logic, colour, type, and the post templates you'll actually repeat — then produce against it. The grid is a designed object, not an archive. Reels carry reach, carousels carry saves, stories carry the people already following you, and we staff them as three separate jobs rather than one content pile.

For — Consumer brands whose product photographs well

LinkedIn Management

Founder voice on the profile. Proof on the page.

Two properties, two jobs. The personal profile carries reach and trust — first person, in the founder voice, about decisions and numbers they actually own. The company page carries proof, hiring and the artefacts a buyer forwards internally. We run both, plus the half of LinkedIn nobody staffs: commenting under the accounts your buyers already read.

For — B2B founders selling to a named account list

TikTok Management

Volume, and hooks rewritten until one lands.

TikTok rewards two things a retainer agency is structurally bad at: volume and speed. We post often, cut every video with several different first-two-seconds, and read the retention curve instead of the like count. A trend gets an entry call while it is still a trend — not in next month calendar.

For — Brands with a face, a demo, or a strong opinion

Or take the whole surface

Studio Partner

Four vendors' worth of surface. One roadmap, one invoice, one name to call.

We take the whole surface: brand system, website, search footprint, social calendar, and the AI systems that keep all four fed. The point isn't consolidation, it's propagation — when you change the positioning on a Tuesday, the site copy, the ranking pages and the social queue all move together, because the same people own the words and the deploy. Four vendors take a quarter to agree on that. You get a named studio lead, one roadmap, one standing call, one invoice, and a workspace you can open any day of the week to see what ran, what it produced and what it cost.

The studio runs on aaddyy.com — the same public tool catalog, MCP server and SDK anyone can sign up for. When an engagement needs a tool that doesn't exist, we build it and run it on your account. Whether it ever reaches the public catalog is your decision, written into the agreement, and the default is no. You can inspect the whole machine before you hire the team.

What a partner engagement includes

A named studio lead in your Slack, not an account manager
One roadmap across design, engineering, search and social
A steady shipping cadence, reviewed on a rhythm we agree
Brand system, production website, and the CMS your team runs it on
Keyword map, publishing calendar, and the pages that fill it
A private AADDYY workspace loaded with your voice, prompts and assets
Every account, repo and file in your name from day one

For — Companies with a number to hit and no in-house marketing function

Why this is different

You aren't billed hours for the machine's half

Drafts, variants, resizes, alt text and first cuts run through the tools. People do strategy, art direction, code review and the call on what ships. The quote itemises which lines are machine work priced as a commodity and which are people, so you can argue with the split instead of guessing at it.

The work happens in an account you can open

Every engagement runs on an AADDYY workspace in your name. The tools we used, the outputs they produced and what each run cost are visible to you on any day of the engagement — not summarised in a monthly deck.

Ownership transfers as you pay, and leaving is cheap

Repos, domains and accounts are opened in your name on day one, and ownership transfers as invoices are settled rather than all at the end. Either side can end a project engagement on reasonable notice, with no termination fee and no notice period dressed up as a retainer.

You can inspect the machine before you hire the team

We show the thing we built: the tool catalog, the SDK on npm, the MCP server, the discovery files, the token file. Install any of it before you write to us. On the first call we open the pipeline live rather than reading you case studies.

How an engagement runs

Brief

A short call. Bring the problem, not a spec. Before we say anything we read what already exists — the site, the deck, the analytics, the files from the last agency.

Scope

A written scope: what ships, who makes it, what it costs, and the dates. If you've asked for more than you need, that's the call where we say so.

Build

Work starts on an AADDYY account opened in your name. You see real output early — not a mood board, the actual thing in draft form.

Review

Scheduled rounds of revisions, agreed upfront, with one named reviewer on your side. Anything beyond that gets quoted, never assumed and never absorbed into a mystery line item.

Handoff

Source files, repos and accounts are already yours, so handoff is a walkthrough, not a migration. Then we either stop or move to a standing cadence. Your call — there is no renewal clause.

If it goes wrong

A date we're going to miss gets flagged before the date, with a new one and the reason — you never find out by the deadline passing. Work you reject inside the agreed rounds gets redone at no cost. And if you decide early on that this isn't working, you can stop, settle for what shipped, and keep it. That goes in the agreement, not a goodwill gesture.

Tell us what needs making.

A paragraph is enough to start. No deck, no procurement form, no discovery fee, no NDA before we will talk. And rather than reading you case studies, we will open the machine on the first call — the repos, the pipeline, the actual output, running live.

Questions people actually ask

Bring the hard one.

The project nobody on your team has had time to start.