Concept Feedback Before Final Files: Get Useful Client Comments

Key Takeaways

• Storyflow is our pick for concept feedback: the brief, two or three directions, the plan and the written decision sit on one board the client opens by link with no account, before anyone produces a final file.
• Our editorial rule: every concept share asks exactly three questions. Which line of the brief does each direction answer, and which does it miss? Which direction do we develop, and what do you take from the others? What would stop you approving the finished version?
• Concept feedback decides the idea while changes cost hours; final approval checks finished files when changes cost days. Mixing them is where most rework comes from.
• In Storyflow, show two or three directions, one wall each, at rough fidelity on purpose. One gives the client nothing to choose; four or more hands them your job.
• A vague comment is a question you have not asked yet. "Make it pop" becomes "which element should the eye land on first?"
• Record the decision in a note on the board, with the date, the decider, the chosen direction, what was borrowed and the dealbreakers, then export a dated PDF. Resolving a comment thread deletes it, so a comment is never the record.
• Storyflow is not a proofing tool: no frame-level approval, no version compare and no approve button per file. Final files go to a proofing tool or a written approval against a named file.

How to Get Useful Concept Feedback: 8 Steps

Most studios share creative work too late. The first time the client sees anything is a polished draft, so every comment is about polish, and the real disagreement (about the idea) surfaces after the files are built. These eight steps move the important decision to the cheapest moment. Storyflow is our product, so weigh the tool choice with that in mind; the steps work with any shared surface that holds the brief beside the work.

1. Put the brief on the Storyflow board first, top left. Objective, audience, the single-minded proposition, mandatories and channels. Every comment gets judged against it, so it has to be on screen.
2. Build two or three directions, one wall each on the Storyflow board. A name, a one-line thesis, a moodboard and four to six rough applications per direction. Rough on purpose, labelled as rough.
3. Add the plan beside the directions. What happens after the decision: development, the next review date, final files, launch. A client who can see the plan knows what this decision unlocks.
4. Write the three questions in a note beside the directions. Brief fit, choice, dealbreakers. They tell the client what you need from them, and they stop "thoughts?" from being the whole request.
5. Present it live for 20 to 30 minutes, then leave the link. Walk the brief, then each direction, in presentation mode. Hold your recommendation until the client has reacted to all of them.
6. Give the client 2 working days, and one decider. Stakeholders can add input; one named person merges it and answers the three questions.
7. Turn vague comments into questions within a day. Reply on the board, in the round's notes table, with a follow-up question for each comment that is a feeling rather than a decision.
8. Write the decision in a note on the Storyflow board and export a dated PDF. Date, decider, direction, what is borrowed from the others, dealbreakers, next step. Development starts from that note, not from memory.

If you need the round structure after the concept is chosen (rounds, deadlines and the approval sentence at final), the creative approval workflow template covers it. This guide stays at the concept stage, where the decision is still cheap.

A fictional eyewear campaign's moodboard, laid out in walls. One direction per wall is what lets a client compare routes side by side instead of scrolling between files.

Why Feedback Is Cheap Early and Expensive Late

Creative work gets more expensive to change at every stage, and the client's comments get less useful. At concept stage a client can say "the second one, but warmer", and the studio redoes a moodboard. At final files the same sentence means rebuilding every asset in every format.

The industry knows the cost

The UK's Institute of Practitioners in Advertising published research with BetterBriefs in October 2021, from more than 1,700 marketers and agency staff in over 70 countries: 80% of marketers thought they wrote good briefs and 10% of agencies agreed, and respondents estimated that a third of marketing budget is wasted because of poor briefs (IPA, BetterBriefs). The follow-up on creative ideas, published in March 2025, reported that creative development now typically takes five rounds to reach a signed-off idea, up from three in IPA research in 2007, and that only 10% of respondents said ideas are always judged against clearly defined criteria (IPA, obstacles to creative ideas).

Rework does not come from clients who change their minds. It comes from decisions that were never asked for at the stage where they were cheap.

Diverge, then converge, on purpose

The UK Design Council's Double Diamond, its model of the design process, puts a "develop" phase of generating different answers before a "deliver" phase that tests and narrows them. Concept feedback is the moment between the two: you have explored, and the client helps you choose. Sharing final files first skips the choosing, so the client chooses later, by rejecting.

Concepts are judged by the brief, or by taste

The same 2025 IPA report found that only 23% of agencies and 15% of marketers say the brief is always used when ideas are evaluated, and that 89% of agencies and 84% of marketers say personal opinion has a big impact on decisions. Putting the brief on the same screen as the work is the cheapest fix there is. On a Storyflow board, the brief sits at the top left of the canvas, so it is the first thing the client sees and the thing every question points back to. If the brief is thin, fix it before the concepts: how to brief a moodboard so the client approves it covers what to agree, and the ChatGPT creative brief workflow drafts one onto the board from notes.

A fictional ski brand's design exploration, with palettes, type pairings and imagery in columns. At concept stage, this is the fidelity a client should see: enough to choose, rough enough to change.

The Storyflow 3-Question Test

"Let us know your thoughts" gets thoughts. A client who is not a designer answers an open question with the only thing they can judge without help: taste. Three specific questions turn the same ten minutes into three decisions.

The test, defined

The Storyflow 3-Question Test: every concept share on a Storyflow board asks the client exactly three questions, written beside the directions: which line of the brief does each direction answer best, and which does it miss; which one direction do we develop, and what would you take from the others; and what would stop you approving the finished version. This is our editorial recommendation, built from how concept reviews fail, not from a study. The number is three because a concept stage has three decisions to make: does it fit, which one, and what must never happen.

Question 1: brief fit

"Which line of the brief does each direction answer best, and which does it miss?" This anchors the review in the brief rather than in preference, which is what the IPA research says is usually missing. It also gives the client language for a dislike: "it misses 'warm and approachable'" is a note you can act on; "I don't love it" is not.

Question 2: the choice

"Which one direction should we develop, and what would you take from the others?" One direction, said out loud. The second half catches the borrowing that happens anyway ("the second, with the photography from the first") and makes it a decision rather than a surprise in development.

Question 3: the dealbreaker

"What would stop you approving the finished version?" This is the question studios skip most easily, and it is the one that saves the final round. It surfaces the hidden criteria: the chair who hates green, the retailer who requires the logo top left, the box that only prints in one colour, the legal line that must fit on a 4:5 post. Each answer becomes a constraint on development, written on the board.

Why three, and why not more

Three questions fit in one message a client can answer in ten minutes and one meeting agenda. A fourth question usually splits one of the three ("which colour?" is part of question 2), and long questionnaires get short answers. If you need detail (copy tone, casting, budget), ask it after the decision, in development, where it belongs.

How to put the test on the board

In Storyflow, write the three questions in a note beside the direction walls, with a three-row table under it: the questions in the first column and an empty Answer column the studio fills. A view-only link opens the board with no account, so the client can read everything; to comment or draw, they sign in, and can answer each question in a comment beside the note. Clients who will not sign in answer by email. Either way, the studio copies the answers into the table, where they stay.

A brand personality framework with sliders between opposing traits. Pinning the brief's adjectives this way gives question 1 something precise to point at.

Concept Feedback vs Final Approval

Concept feedback decides which idea to develop while the work is still rough and cheap to change; final approval confirms that finished files are correct and ready to release. In Storyflow, concept feedback happens on the board beside the brief, and final approval happens against named files in a proofing tool or a written approval line.

Concept feedback asks "is this the right thing?" Final approval asks "is this thing right?" The questions sound alike and need different material, different tools and different people in the room.

What "approved" means at each stage

At concept stage, approved means: develop this direction, with these borrowings, inside these constraints. It does not mean every image on the moodboard will appear in the final work, and the decision note should say so. At final, approved means: these named files may be released as they are. Write both meanings on the board before the first share, so nobody signs one thinking it is the other.

Trap 1: approving a concept as if it were final

A client who says yes to a moodboard and then sees the developed comps may feel the studio changed something, because the moodboard's stock photo of a model in a field is not the shoot they will get. Label every concept image with its job ("the light, not the location") and write "concept, not final artwork" on every direction wall.

Trap 2: reopening the concept at final

The other direction: a final-round comment that says "actually, could it feel more like the second route?" That is a new concept decision at final-file prices. Point at the decision note, price the change, and let the client choose whether to pay for it. The note's date and decider make this a conversation about a decision, not about memory.

When the two can merge

Small jobs, such as a single social post or a one-off poster, can go straight from concept to final in one share, if the client sees the brief and two options and the job takes less than a day to redo. The three questions still apply; the record is just shorter.

What to Share at Concept Stage

Clients give better feedback on less. A concept share has one job, to get a direction chosen, and every item on the board should serve that job. For identity work, the brand identity presentation page shows the same board for logo territories.

Show directions, not drafts

A draft is one answer, half finished. A direction is a complete idea, roughly made. Clients respond to drafts by fixing them and to directions by choosing, and choosing is what you need. Give each direction a two or three word name and a one-line thesis, and check that every direction answers the same brief. What a creative concept is sets out the difference between a concept and its executions.

Rough on purpose

Concept work should look like concept work. A comprehensive layout, the traditional comp, is a proposed design presented to a client on the way to approval, and hand-rendered comps help clients focus on the design idea instead of nitpicking the finish. Use borrowed images, grey type where copy is not written, and the word ROUGH in a corner. A mood board exists to communicate a direction before final work, and it does that best when nobody mistakes it for the final work. How to present a mood board to a client covers labelling images by their job.

Two or three directions, and no decoys

One direction gives the client nothing to compare, so they compare it to what is in their head. Four or more turns the client into the creative director. Two or three is the useful range. Never add a weak third direction to make your favourite look better: the decoy effect, documented by Huber, Payne and Puto in 1982, shows that a worse third option can shift choices, and clients sometimes pick the decoy or see through it. Every direction you show should be one you would be glad to make.

Hold your recommendation until the end

Show every direction before you say which one you would choose. The anchoring effect, described by Tversky and Kahneman in 1974, pulls judgments toward the first reference point offered, so a recommendation stated up front colours everything after it. Give your recommendation, with its reason in the brief's words, after the client has reacted to all of them. How agencies present campaign concepts covers the argument around a recommendation, and presenting design concepts to get a decision covers the meeting.

Put the plan on the board

The plan beside the directions answers the question every client has and rarely asks: what happens if I say yes? A five-row table (development, next review, final files, approval, launch) is enough. In Storyflow, it sits on a wall to the right of the directions as a table with date cells, rows kept in date order. Agreeing the review window at kick-off, as part of creative client onboarding, makes the two-day deadline feel normal rather than pushy.

Leave these out

• Final copy, unless the concept is a line of copy.
• Options for every detail ("three blues"). Details are for development.
• Your process: the 80 references you collected. Show the 15 that make the case.
• Anything you would not be happy to produce.

An interior moodboard with named swatches. Labelled colours and materials give a client something specific to answer question 1 about.

How to Stop Vague Comments

Vague comments are not bad clients. They are honest reactions from someone who has not been given words for what they see. The fix is to reply with the question that turns the reaction into a decision, the same day.

Ten vague comments and the question to ask back

Ask for reactions, not solutions

Clients who propose solutions ("move the logo to the left") are usually pointing at a problem ("I cannot see who it is from"). Thank them for the solution and ask for the problem, then solve it your way. The Nielsen Norman Group's guide to design critiques makes the same point for internal reviews: agree on the problem the design must solve before critiquing, restate the goals at the start, and treat directives such as "I would have done it differently" as comments to reframe against those goals.

Why clients comment on the small things

C. Northcote Parkinson's law of triviality, from 1957, describes a committee that approves a nuclear reactor in minutes and argues about a bicycle shed for an hour, because everyone feels qualified to judge the shed. A client who comments only on the font size has often not found a way into the idea. Question 1 gives them that way in.

One voice, merged before it reaches you

Five stakeholders' comments sent separately give you five contradictions to resolve and no authority to resolve them. Ask the client to name one decider who merges the comments and answers the three questions. On a Storyflow board, give the decider's answers their own table, and treat other comments as input to it.

Reply on the board, not in the inbox

Write each follow-up question in the same row as the comment it answers, in the round's notes table on the board. Email threads split, and a question asked in one thread is answered in another. In Storyflow, the table keeps comment, question and answer in one row, and a blank Answer cell shows what is still open.

A product moodboard where every image connects to the hero shot. When a client says "make it pop", a board like this lets you ask which single image the eye should land on first.

How to Record the Decision

A decision nobody wrote down is a decision you will have again, at final-file prices. The record takes five minutes and saves the hardest conversation in the project.

The decision note

Write it in a note, not a comment

In Storyflow, write the decision in a note at the top of the board, not in a comment. A signed-in client can comment, but resolving a comment thread deletes the whole thread, and an approval in a deleted thread is no approval. A note stays where you put it, and anyone opening the board by link sees it first.

Export a dated PDF

A live board keeps changing as development starts, so save the concept as it was approved. In Storyflow, export the board as a PDF the day of the decision, name it with the date ("Kiln-and-Co_concept_approved_2027-04-14.pdf") and send it to the decider with the note's text in the email. Storyflow has no version compare, so the dated PDF is the proof of what was approved.

What the decision unlocks

Development starts from the note: the chosen direction, its borrowings and its constraints. Copy the dealbreakers into the development brief, because the designers who were not in the meeting need them most. At final approval, the dealbreakers become the checklist: does the box print in one colour, does the logo read at 12 mm.

A six-stage workflow from brief to delivery with a comment beside one stage. The decision note sits where the concept stage ends, so the next stage starts from it.

The Concept Round Template to Copy

Copy this table onto the board as the concept round's notes table. The filled rows are a fictional round for Saltmarsh Studio's work on Kiln & Co, a fictional ceramics brand's rebrand and spring launch. The Question column says which of the three questions each comment answers, or "None" when it answers none of them.

Four of eleven comments answered none of the three questions, and each became useful after one follow-up. Three dealbreakers (one-colour print, 12 mm legibility, the stockist's logo position) were found at concept stage, where they cost a line on the development brief instead of a reprint.

A logo shown as one mark with variations branching off it. A dealbreaker such as "readable at 12 mm" is easiest to test when every variation sits on the same board.

Worked Example: Saltmarsh Studio and Kiln & Co

The studio, the client and the people are fictional; the method is the one this guide recommends. Saltmarsh Studio is a two-person brand studio. Kiln & Co, a ceramics maker selling at markets, through one stockist and online, has asked for a rebrand and a spring launch campaign. Ana, the founder, decides; her business partner has opinions.

The board

Top left: the brief, with the objective (look like a maker, not a gift shop), the audience (people who buy one good mug rather than six cheap ones), the proposition ("made by hand, used every day") and three adjectives: warm, handmade, honest. To its right, three direction walls: Fired (terracotta, kiln textures, a stamp-like mark), Studio Light (white space, editorial still life, a fine serif) and Hand to Table (lifestyle photography of people using the pieces, a handwritten mark). Each wall has a moodboard of 12 to 15 images, a logo territory sketch and four roughs: the box, a 4:5 post, the web homepage and the market stall sign. To the right, the plan table. Beside the directions, the three questions in a note, with a three-row answer table.

The share

Saltmarsh presented live for 25 minutes in presentation mode, wall by wall, with the brief first. They gave their recommendation (Fired) only after Ana and her partner had seen all three. Then they left the view-only link, gave two working days and named Ana as the decider. The client presentation tool page shows the same walk-through. Ana signed in and answered the three questions in comments beside the note; her partner sent comments by email. The studio copied both into the answer table and the round table the same day.

The follow-ups

The eleven comments in the template above came in over two days. The studio replied to each vague one with a follow-up question in the same row within a day. The partner's "bigger logo" turned out to be about the market stamp, a 12 mm mark pressed into clay, which became the most important dealbreaker in the project.

The decision

On 14 April, the decision note read: "Direction: Fired. Borrowed: people in frame, from Hand to Table. Dealbreakers: one-colour print on kraft; name readable at 12 mm; stockist's logo top left on shelves; one system offline and online. Not decided: final photography, copy. Next: comps on 28 April. Decider: Ana Ruiz." The studio exported the board as a dated PDF and emailed it with the note's text.

What it saved

Development started from a constraint list nobody had written in the brief. Without question 3, the one-colour print rule would likely have surfaced when the box printer saw the files, and the 12 mm stamp when the first stamp was cut. Both would have meant redrawing the mark after final files. The example is illustrative; the point is where in the process each constraint appeared.

A fictional apparel brief laid out on a board beside its references. The brief at the top left of the concept board is what question 1 points back to.

Do This in ChatGPT or Claude

You can lay out the concept board, the three questions and the round table from a chat, and they land on the Storyflow board the client opens by link. ChatGPT calls these plugins and Claude calls them connectors; both use the same Storyflow account.

Storyflow works inside ChatGPT and Claude; the board stays in your Storyflow account.

Install it in ChatGPT

Open the Storyflow listing in ChatGPT's plugin directory, select Install plugin, sign in to Storyflow and choose which projects ChatGPT may use, and whether it may also make changes. Availability depends on your ChatGPT plan, workspace and region, as OpenAI's plugin help explains; in a Business or Enterprise workspace an admin manages plugins, covered in OpenAI's admin controls article. In a chat, type @ and pick Storyflow, or choose it from the + menu. The ChatGPT client presentation page shows the board side, and the ChatGPT moodboard page covers the moodboard inside each direction.

Connect it in Claude

Add a custom connector with the address https://mcp.storyflow.so/mcp, following Anthropic's custom connector guide. Setup for every assistant is on the Storyflow MCP page, and the Claude client presentation and Claude moodboard pages show the same jobs from Claude.

Prompts to copy

1. "On the Kiln & Co board, add the brief at the top left as a document from this text, with headings for objective, audience, proposition, mandatories, channels and the three adjectives. [paste brief]"
2. "Make three walls to the right of the brief called Fired, Studio Light and Hand to Table. At the top of each, add a note with the direction's one-line thesis, and the word ROUGH."
3. "Add a note beside the three walls with these three questions, and a table under it with columns Question and Answer and one row per question: which line of the brief does each direction answer best, and which does it miss; which one direction should we develop, and what would you take from the others; what would stop you approving the finished version."
4. "Create a table called Concept Round with columns #, Comment as given, From, Question it answers, Follow-up asked, Answer, What it changes. Here are the client's comments: [paste]. Mark each with Q1, Q2, Q3 or None, and draft a follow-up question for every None."
5. "Add a plan table to the right of the directions with rows for development, next review, final files, approval and launch, and dates I will give you, kept in date order."
6. "Add a note at the top of the board: Direction: Fired. Borrowed: people in frame, from Hand to Table. Dealbreakers: [list]. Decider: Ana Ruiz. Date: 14 April."
7. "What changed on the Kiln & Co board in the last week, and who changed it?"

What lands on the board

Documents, notes, walls, tables, headings and links, plus colour palettes and images added to the board. The assistant reads the board, reports what changed and who changed it, and lists who is on the board. Plugin actions run on your ChatGPT plan and use no Storyflow AI credits.

What to check by hand

• Every follow-up question: the assistant drafts them from the comment's words; you know the client.
• The Q1, Q2, Q3 labels: read each against the comment. A dealbreaker disguised as a preference is easy to mislabel.
• The decision note: read it back to the decider before you export the PDF.
• Images: a reference attached in the ChatGPT chat does not reach Storyflow; the plugin gives you an upload link for the board. AI image generation is not available through the plugin.
• Comments: the plugin cannot read board comments, so paste client comments into the round table. It can add one comment card in your name.
• Allowance: a collaborator you invite joins free with 25 plugin or connector actions a week; Plus, Pro and Max have no weekly limit.

The in-app AI is separate

Inside Storyflow, on paid plans, the built-in AI has its own workflows, such as Brief my campaign, Build a moodboard for my project, Review what's on this board and What's missing here?, and it can design deliverables from the board, such as a one-pager or a designed PDF. It reads the current board plus up to five documents and three other boards you attach, and it uses Storyflow AI credits. AI image generation for concept roughs is on Pro and Max, in the app. A client invited free has no in-app AI.

Where to Share Concepts: Storyflow vs Other Kinds of Tool

Storyflow is listed first on one stated basis: the job here is concept feedback, which needs the brief, the directions, the plan and the written decision on one screen, and Storyflow holds them on one board the client opens by link with no account. Final files are a different job, and other kinds of tool do it better.

A proofing tool

A proofing tool is the right place for final approval: comments pinned to a spot on a file, versions compared side by side and a recorded approve or reject per version. That precision is a problem at concept stage, because a client handed a file and a pin tool comments on the file. Choose a proofing tool once the direction is chosen and the files are near final, and link the proof from the board. Creative proofing tools for small studios lists the features to check, and the creative approval workflow page shows how the board and the proof fit together.

A slide deck

A deck tells a story in a fixed order, which is ideal for a live pitch. Its limit at concept feedback is that the brief is on slide 2 and the directions are on slides 8 to 20, so the client never sees them together, and a PDF of a deck gets comments by page number. Choose a deck for a formal pitch meeting with a running order, as in the agency pitch process, and keep the board as the place the decision is recorded; the campaign concept presentation page shows both together. Storyflow decks present fullscreen and export as PDF, but they do not export to other slide formats.

A PDF by email

A PDF works for any client and is the right archive. Feedback on a PDF comes back as an email, often from several people, with no structure and no link to the page it refers to. Choose a PDF for clients who will not open links, ask the three questions in the email body, and paste the answers onto the board.

A design file shared by link

Sharing the live design file is honest and exact, and design-literate clients like it. At concept stage it pulls attention to pixels and alignments, and directions often sit on separate pages. Choose it for developed comps, after the concept decision, when exactness is the point.

A shared doc

A doc carries a written rationale better than any visual tool, and copy-led concepts (a tagline, a naming route, a manifesto) can live there. Visual directions shrink to inline images. Choose a doc for a copy-led route, and put it on the board as a document beside the visual directions.

A whiteboard app

An open whiteboard is good for co-creation with the client in the room, before there are concepts to choose. A blank canvas gives no structure for a decision: no brief position, no plan, no record. Choose one for a workshop at the discovery stage, and move the output onto a structured concept board afterwards.

A fictional studio's tracker with a Waiting on Client column. Concept reviews with a two-working-day window keep that column short.

Mistakes and Fixes

Showing final files first

The client's first look is at finished work, so every comment is about polish and the idea is never discussed. Fix: share a concept board with two or three rough directions before any final file exists.

Asking "what do you think?"

An open question gets taste. Fix: ask the three questions, written beside the work, every time.

Hiding the brief

Without the brief on screen, clients judge against what they remember, which is usually less than what they wrote. Fix: put the brief at the top left of the board, and point back to it in question 1.

Showing a decoy

A weak third direction to flatter your favourite. Clients sometimes pick it. Fix: show only directions you would be glad to make.

Leading with your recommendation

The first thing said anchors the rest. Fix: show every direction, take reactions, then recommend with a reason in the brief's words.

Collecting comments from everyone directly

Five stakeholders, five contradictions. Fix: one named decider merges input and answers the three questions; the rest is input.

Recording the decision in a comment

Resolving a comment thread on a Storyflow board deletes it. Fix: write the decision in a note at the top of the board and export a dated PDF.

Never asking for dealbreakers

The one-colour print rule appears when the printer sees the files. Fix: ask question 3 at concept stage and copy every answer into the development brief.

Treating a concept "yes" as final approval

The client approved a direction, not every image on the moodboard. Fix: write "concept, not final artwork" on every direction wall, and say in the decision note what was not decided.

Storyflow's Honest Limits for Client Feedback

Storyflow earns its place at the concept stage: the brief, directions, plan and decision on one board the client opens by link. It does not do the following, and the method above works around each.

• Not a proofing tool. No frame-level approval, no pins tied to a file version, no approve or reject button per version.
• No version compare. The board shows its current state, not the version the client approved, so the dated PDF is the record.
• Commenting needs a sign-in. A view-only link opens with no account; to comment or draw, the client signs in. Clients who will not sign in answer by email.
• Resolving a comment deletes its thread. Keep decisions in notes.
• No e-signature, proposals or invoices. The decision note is a written record, not a signed contract.
• No PPTX export or presenter notes. Boards export as PNG, JPEG or PDF; decks present fullscreen and export as PDF.
• Assistants cannot read comments. The ChatGPT plugin and the Claude connector read notes, tables and documents.
• No native desktop or mobile app. Storyflow runs in the browser and is best built on a laptop.
• Paid-only during early access. Plus is $7.99 a month billed annually ($9.99 monthly) and includes the in-app AI; pricing has the plans. Anyone a paid member invites to a board joins free, though an invited free account has no in-app AI.

For the studio side of the same board, see the client review board and the creative concept board pages; for campaign work, the campaign concept board; and for how a review board and a proofing tool split the later rounds, the design feedback and creative review pages.

The Bottom Line

Useful client comments come from asking for the right decision at the cheapest moment. Share two or three rough directions on one Storyflow board with the brief beside them, ask the three questions, turn every vague comment into a question, and write the decision in a note with a dated PDF behind it. Then let the proofing happen later, on files, where it belongs. Start with your next project: build the concept board before any final file exists, and count how many dealbreakers question 3 finds. To build it from a chat, use the ChatGPT client presentation workflow; for the moodboards inside each direction, the moodboard for clients page shows the layout.

This article was published in November 2026, with the IPA, Design Council and other sources read in November 2026.