Ines Halloran
Design director · 16 years: typography, agency identity work, in-house product, now fractional for Series A to C teams
Brand and product design direction
A composite written by avva as a worked example. Not a real person.
Before your AI advises you on brand and product design direction, it reads a reference model avva wrote for this field: the decisions it holds, the reasons behind them, and where it would push back.
Use this model for
- Rebrand or refresh
- Redesign scope
- Design system timing
- Accessibility calls
- Research before build
- Visual direction
Add this link to Claude, ChatGPT, or Cursor. Your AI then checks Ines’s decisions before it answers you. The answer is still your AI’s.
https://On a phone? Email the link to yourself and add it on a computer.
What your AI sends is used for the answer and not stored.
This is a reference model, so it answers in full to everyone, with or without an account. Without an account your AI gets a smaller part of a real expert’s model, and all of it once you sign in.
See Ines’s model answer a question
One question this reference model answers, and the published decision it rests on. Your situation gets an answer of its own, written by your AI.
Example inside your AI app: an exchange avva wrote for a reference model
We're a 20-person Series A fintech. Our marketing site looks dated next to our two biggest competitors and the board keeps bringing it up. Do we redesign the brand or the site first?
Checked Ines’s Decision Model
Neither, in the way the board means it. Keep the name and the mark. Rewrite the homepage headline so it says what you do, for whom and the number you move; replace every illustration with a real screenshot; and move the site's type and color closer to the category leaders, not further away.
Why. In finance, looking like a bank is what "safe" looks like, and that trust is free if you borrow it. What makes the competitors look newer is that they show their product and say one sharp sentence. Difference belongs in the sentence; a new logo spends the one design budget you have where no buyer is looking.
What would change this. If your buyer has changed — you now sell to CFOs where you sold to analysts — the mark may be telling the new buyer the product is for somebody else, and then the brand goes first. That is a change of audience, not of taste.
Outside this model. Nothing here about what the redesign should cost or how to run an agency relationship. This model covers what to change, what to copy on purpose, and what to leave alone.
Copy an example prompt4 patterns with a gap to fill, once you are connected
Before you commit
A full pass over a plan or a draft, with the changes it needs.
Use Ines Halloran's Decision Model on avva before we lock this in. What we're deciding: <paste your plan, doc, or diff here> Run their review, follow the steps it returns, and give me the verdict and any required changes before you continue.Choosing between options
A choice between two options, their stop rules checked first.
Use Ines Halloran's Decision Model on avva to compare these. Option A: <describe it> Option B: <describe it> Context: <what constrains the choice — redesign scope> Check their stop rules first, then compare the options under their judgment, and tell me which one their model favors and why.A quick gate
A quick look for anything they would refuse outright.
Run this past the stop rules in Ines Halloran's Decision Model on avva before I send it. <paste the draft, plan, or change> Tell me only what they would refuse or flag, and why. If nothing trips, say so.Before you have an approach
The questions they ask before an approach gets locked in.
I'm starting on rebrand or refresh. Ask Ines Halloran's Decision Model on avva what they would want settled first. Situation: <two or three lines> Give me their questions, and flag the ones my situation does not answer yet.
Patterns with a gap to fill. None is an answer this model gave. The worked answer above is avva’s too — this is a reference model, and no real person’s decisions are on this page.
Connect Ines’s Decision Model to your AI
Three steps. The first works now, with no key or account; the last makes your AI check Ines’s model without being asked.
The link works in AI apps that can add a remote MCP server (MCP is the open standard for plugging in outside tools). You set it up once.
A connector is an address your AI reads from: one line, nothing to install.
https://avva.chat/ mcp/ design- direction- ines- halloran - Open Claude on the web or in the desktop app and go to Customize → Connectors.
- Press + next to Connectors and choose Add custom connector.
- Paste the address above and name it avva-design-direction-ines-halloran.
- When Claude asks to allow avva, say yes.
Missing on a work account? Ask your admin to allow it. Claude Code and Cursor take a local config instead.
The link is added on the web or in the desktop app. Email this line to myself
No key in this line, and none needed: a reference model answers in full to everyone.
Check it worked. Send this in the chat once the link is added.
Ask Ines Halloran's Decision Model on avva what kinds of decision it covers, and summarize it for me in three lines.Then try it on a decision you are making this week: copy an example prompt and fill in the gap.
Not legal, medical, financial, or other professional advice. Terms of Use
About this model
Open any of these to see what this model is built from, and how it is put together.
How this model is groundedWhat it holds, and what Ines weighs first
This model holds 60 decisions, 16 forces, 9 criteria, and 14 stop rules.
What Ines weighs first
avva set this order. 3 of 9 criteria are shown, the ones the model puts first. The other 6 reach your AI too, along with what each one looks for, the stop rules, and what changes Ines’s answer.
Ines’s usual answer changes under 8 named conditions. Those conditions and the 14 stop rules reach your AI with the rest.
See examples of the decisions in it5 of 60, written by avva for this model
The first 5 in the record, in published order. They are not the ones nearest to your situation, and not a summary of the rest. Your AI does not copy a decision it reads here: it checks what decided that case against your own facts, and weighs it beside which pressures tended to settle Ines’s cases. Your AI receives all 60 with no key, ordered against your situation.
What decided itA stranger finishes the task without help · It shows the actual product
Situation: Project-management SaaS, 15 people, activation at 22 percent. The growth team wanted a 7-step product tour with tooltips on first login.
Decision: No tour, and no redesign of the empty states or first-run this quarter. Put real product screens and a thirty-second recording on the signup page so people know what they are signing up for; expect fewer signups and a higher activation rate, and redesign first-run only if activation stays under 30 percent.
Why: A 22 percent activation rate in a project tool is mostly people who signed up for something they had not seen. Tours are dismissed within two steps; showing the actual product before the signup form moves the explaining to where it filters, and a first-run redesign measured on a mis-sold audience measures the audience.
What decided itBuyers already know what this category looks like · The difference has to be said, not drawn
Situation: Series A fintech, 20 people, 300 paying businesses. The board said the marketing site "looks dated" next to two competitors and asked for a rebrand — new logo, new palette, a custom illustration system — before a Series B roadshow in ten weeks.
Decision: No rebrand and no new logo. Rewrite the homepage headline, replace every illustration with real product screenshots, and move the site's type and color closer to what the category leaders use, not further away.
Why: In finance a buyer reads "looks like a bank" as "safe", and that trust is free if you borrow it. The competitors look newer because they show their product and say one sharp sentence; their marks have nothing to do with it. Difference belongs in the sentence.
What decided itThe team is one designer
Situation: Series A with budget for one of two for six months: a branding agency or a senior in-house designer. The product interface needed an overhaul and the site was dated.
Decision: Neither first. A senior freelance product designer for three months on the product interface, then hire in-house with the freelancer's screens as the brief. No agency near the product.
Why: An agency delivers a brand book and leaves; a product interface needs someone in standup every week. Three months with a senior freelancer tells you which role to hire.
What decided itA stranger finishes the task without help · The team is one designer
Situation: Series A, one product, two designers and eleven engineers shipping weekly. The new design lead wanted to spend the quarter building a component library and a token set before redesigning onboarding, where 60 percent of signups drop.
Decision: No design system this quarter. Redesign onboarding first, keep one page of screenshots of every shipped component, and extract a component only when it has been rebuilt for the third time.
Why: A library built before the product has repeated itself encodes guesses. Three copies is the evidence that a component is real; two is a coincidence, and the onboarding drop is costing signups every day the library is being drawn.
What decided itIt has to feel instant · Daily users work by hand position
Situation: Analytics dashboard, 50 people. Before an enterprise launch the design team wanted page transitions, animated chart entry and a branded loading sequence to make the product "feel premium".
Decision: No page transitions, no animated chart entry, no loading sequence, and no skeleton screens either. Keep the last numbers on screen, marked as updating, until the new ones arrive; the only motion is under 150 milliseconds, where state changes.
Why: Analysts open this dashboard forty times a day. Chart entry is waiting for something they knew was coming, and a skeleton screen is a loading sequence in gray: it wipes the numbers they were reading. In a work tool, premium means the previous answer stays put until the next one is ready.
How this composite takes up a decision7 steps in order, written by avva from its decisions
Step 1Trace the proposed fix back to the failure: read a rebrand, a tour, a dark mode or a refresh as a symptom and find what is actually losing people — a headline that says nothing, a signup page that hides the product, a screen where sessions stall.
When: A fix, not a failure, arrives as the request
Unless: The request comes with churn reasons attached
Step 2Pull the evidence already in hand before commissioning any: last month's session recordings, the support queue, churn-interview reasons, recordings of sales calls. A recording of people finishing the task outranks anyone's reaction to the page.
When: A live product exists to watch
Unless: No product is in anyone's hands yet
Step 3Size a redesign by where sessions concentrate or stall: restyle the handful of screens people live in (eight of 140 carried most sessions), fix the three places sessions stall, and let the stylesheet carry new type and color everywhere else.
When: A redesign is proposed for a product with live users and many screens
Step 4Order the work by what costs users now: the onboarding drop or the requested feature ships first; the component library and the visual refresh follow; the logo comes last.
When: A user-facing failure and a tidy-up are both queued
Unless: The tidy-up is itself what users hit, such as one control rebuilt five times with tickets naming it
Step 5Argue taste last: the 7:1 body-text floor holds even where conventions are waived, and once the options clear it the owner picks the same day, because the argument costs more than either color.
When: The options differ only in how they look and neither is a category signal
Step 6Answer a refusal with a named substitute and the budget or energy attached: the campaign budget goes to the headline and a sixty-second recording, the CMO's energy to the one-line category sentence, the CEO gets one change that leaves the evidence alone.
When: A senior person's idea is declined
Step 7Close with what would reopen the call and what stays shipped meanwhile: a measured threshold (redesign first-run only if activation stays under 30 percent) or repeat evidence (revisit the name after launch with the recordings of every sales call).
When: The call defers or declines a change that stays plausible
Technical methodologyHow the 60 decisions that say what won and what lost are read
- 60written decisions, the material avva composed this model from
- 16forces those decisions turned on, the pressures Ines decides under
- 196contests read off 60 decisions: every force that settled a case is set against every force that lost there, so a case naming two of each counts four times
- In groupswhich forces tended to settle cases over which, read off those contests
Nobody ranks these forces by hand. Where any order between them comes from is on For AI agents.
How this composite decidesA summary, written from the decisions in it
Sixteen years in brand and product design: trained as a typographer, a decade of identity work at agencies, then design lead on a consumer-finance product, now fractional for Series A to C teams. She has retired more design systems than she has launched and argues that visual difference is the most expensive way to be forgettable.
What to redesign, what to copy on purpose, and what to leave alone: rebrands, renames and logo changes, how far to borrow the look your market already trusts, redesign scope from onboarding flows to dashboards, when a design system has earned its place, contrast and legibility floors, research that watches instead of polls, screenshots versus illustration, dark mode, motion, and agency versus in-house.
The summary covers what the model holds.
What this isA composite avva wrote, and what that changes
- Written by avva. A reference model, a worked example so you can see what a published model looks like and connect an AI to one. Marked Demo, here and on its card.
- Not a real person. The name is invented and the record is a composite, authored in one consistent voice. No real expert’s positions were copied.
- Same format. Your AI receives it exactly as it would receive a real expert’s model, cases and stop rules included.
- Steps aside. When a real expert publishes in this domain, we take this model off the shelf by hand.