How Creative Agencies Present a Campaign Concept to a Client (2026)

How do creative agencies present a campaign concept to a client?

Creative agencies present a campaign concept by opening with the client's own business problem restated in the client's own words, showing the strategic insight that reframes it, then revealing the creative idea as the inevitable answer to that insight, and only then showing executions. The deck is a proof of reasoning, not a gallery of work. Most agencies present two concepts rather than three, present them live before sending anything, and close by naming the single decision they need from the room.

A campaign concept is not approved in the room. It is approved in the three rooms it travels through afterwards. The live presentation is the easy one, because you are there to narrate. The concept then gets forwarded to a decision-maker who was not in the meeting and reads it with no voiceover, and weeks later somebody asks why this direction was chosen and the reasoning has to still be attached to the work. Agencies lose approved work at the second and third stage far more often than they lose it in the room.

I have sat on both sides of this. I have pitched documentary treatments to commissioners and I have watched brand teams forward a deck internally until the idea arrived at the budget holder stripped of everything that made it make sense. The pattern is always the same: the deck was built to be performed, and nobody built the version that survives being read alone.

Tooling follows from that. Slide tools such as Figma Slides, Pitch, Canva and Keynote own the live room, and they are the right answer for the performance. Loom owns the forwarded walkthrough. Storyflow is built for the third room, the record: the concept, the strategic reasoning, the references and the rejected routes stay attached to each other on one canvas, so the answer to "why did we choose this" is still there in week six. It is early access, it has no slides and no presenter mode, and it is not the tool you present from. Use it for the thinking and the record, and present from a slide tool.

For the design-craft version of this question, see How to Present Design Concepts to Clients. For what to do when the first round lands badly, see How to Present a Concept When the Client Hates the First Round.

The Three Rooms every campaign concept has to survive

Every campaign concept is presented three times, and only the first one has you in it.

Room One: the live presentation. You are present. You control pace, order and emphasis. You can read the room and adjust. You can answer an objection the moment it appears. This room is generous, and it flatters weak decks, because your narration is doing work the slides are not. An agency that only tests its concept in Room One does not know whether the concept is good or whether the presenter is.

Room Two: the forward. Within a day, your deck is forwarded to someone who was not there. A marketing director sends it to a VP. A brand manager sends it to procurement or legal. Nobody narrates it. The recipient scrolls, forms a view in roughly ninety seconds, and replies with one line that can end the project. A deck built for Room One is mostly images and single-word slides, which is exactly the deck that dies here, because without narration it reads as thin.

Room Three: the record. Six weeks later, the campaign is in production and somebody asks why the concept went this way instead of the other route. Or a new stakeholder joins and wants to reopen the decision. Or the campaign underperforms and the post-mortem needs to establish what was decided and on what basis. If the reasoning lives only in a slide deck version 4 final v2, the honest answer is that nobody can reconstruct it, and the decision gets relitigated from scratch.

Most agency decks are engineered for Room One and abandoned in Rooms Two and Three. That is the single largest avoidable loss in agency creative, and it is a packaging problem rather than a creative one.

The fix is not one artifact that serves all three rooms, because the three rooms want different things. The fix is deciding, before you build anything, which artifact serves which room:

Agencies that consistently get work approved build all three deliberately. Agencies that lose approved work build the first and hope.

The seven-beat presentation sequence

The order below is close to universal across agencies that pitch well. It is not a template to fill in. It is a sequence of arguments where each beat earns the next.

Beat one: the problem, in their words. Open by restating the business problem using the client's own language from the brief and from the kickoff conversation. Not your reframing of it yet. Theirs. This does two things: it proves you listened, and it establishes the thing you are about to reframe. If the client does not nod here, nothing after this lands.

Beat two: what we found. The research, the audience truth, the category observation, the thing you learned that they did not already know. This is where credibility is either earned or lost. One genuinely surprising finding beats nine confirmations. If you found nothing surprising, say so plainly and lean harder on beat three.

Beat three: the insight. A single sentence that reframes the problem so the answer becomes obvious. This is the hinge of the entire presentation. An insight is not a statistic and not a description of the audience. It is a tension: people want X but do Y, because Z. If your insight can be swapped into a competitor's category without changing, it is not an insight.

Beat four: the idea. The creative concept, stated in one line, before any execution is shown. Say it out loud, let it sit, and do not rush to the visuals. If the idea needs three slides of setup to be understood, it is not one idea.

Beat five: what it looks like. Now the work. Three to five executions maximum for the hero concept, chosen to show range rather than volume: one that shows scale, one that shows craft, one that shows how it survives a small format. Resist the instinct to show everything the team made.

Beat six: how it runs. Channels, phasing, what happens in week one versus month three, and what the client's team has to do. This beat is where agencies most often go thin, and it is the beat that procurement and the budget holder actually read. A brilliant idea with no operational answer reads as expensive.

Beat seven: the decision. Name the single decision you need today. Not "thoughts?" and not a list of six open questions. One decision, stated plainly: "We need agreement on the territory today so we can build the production plan. Everything else can move." Presentations that end with "any questions" get feedback. Presentations that end with a named decision get decisions.

How many concepts to present, and why three is usually wrong

The default is three. The default is usually wrong.

Three routes signals that the agency does not have a point of view. It also invites the worst outcome in agency creative, which is the composite: the client takes the headline from route one, the art direction from route two and the tone from route three, and the result is worse than any of the three and nobody owns it.

Two is the strongest number for most concept presentations. One is the recommendation, developed properly, with the reasoning visible. The second is genuinely different in strategy rather than in decoration, developed enough to be real. Two routes force a choice between positions rather than a shopping trip between executions.

Present one route when you have a strong existing relationship, the brief is narrow, and the agency is confident enough to stake the meeting on it. This is the highest-trust move and it reads as conviction when it works and as arrogance when it does not.

Present three or more only when the brief is genuinely exploratory, the client has explicitly asked for territories rather than a recommendation, or you are in a competitive pitch where breadth is being scored.

Whatever the number, rank them. Say which one you would run and why. An unranked set of routes hands the agency's judgment back to the client, which is the one thing the client is paying not to have to do.

What goes in a campaign concept deck

A concept deck is usually fifteen to twenty-five slides. Longer than that and Room Two stops reading. The spine:

• Title and context. Client, project, date, version. Boring and load-bearing, because this deck will be forwarded.
• The brief as we understood it. One slide. In their words.
• What we found. Two to four slides of research or audience work, with sources named.
• The insight. One slide. One sentence. Nothing else on it.
• The idea. One slide. The concept line and a single image.
• Why this works. One slide connecting the idea back to the business problem from slide two. This is the slide that survives the forward.
• Executions. Three to five per concept. Hero format first.
• Range. How the idea extends: social, out of home, retail, whatever the channel mix is. Two to three slides.
• How it runs. Phasing, channels, dependencies, what the client owns. Two slides.
• Measurement. What success looks like and how it is read. One slide, and it should be specific.
• What we need. The decision, the timeline, the next milestone. One slide.
• Appendix. Everything you might be asked for and nothing you plan to present.

The appendix is doing more work than agencies give it credit for. It is where the rejected routes, the full research, the competitive audit and the detailed costs live. In the room you never open it. In Room Two, it is the reason the deck reads as substantial rather than thin.

Try it on a board

The appendix is the half that survives the forward

Rejected routes, full research and detailed costs belong somewhere retrievable rather than buried in slide 41. Hold them on a board where each one carries the reason it was rejected.

See how it works→Browse templates

Marketing Campaign template →

What to withhold, and why

Good concept presentations are edited hard, and the editing is strategic rather than aesthetic.

Withhold the rejected routes from the main deck. Showing the work you killed invites the client to resurrect it. Keep them in the appendix or, better, in the record where they are labelled as rejected with the reason attached. Rejected routes are enormously valuable six weeks later and actively harmful in minute twelve.

Withhold execution detail on the non-recommended route. If route two is rendered as beautifully as route one, you have not made a recommendation, you have made two recommendations.

Withhold the production cost until the idea is agreed. Not to hide it. Because a number shown next to an unapproved idea becomes the lens through which the idea is judged, and creative reviewed on cost gets smaller.

Do not withhold the operational burden on the client's team. Agencies routinely soften this and it is the single most common cause of an approved campaign stalling in month two. Say plainly what the client has to supply and by when.

The four objections that kill good work, and how agencies answer them

"It doesn't feel like us." This is almost never about the work. It is about the client not having been walked from their brand to this idea. The answer is not to defend the idea. It is to go back to beat three and re-establish the insight, then re-derive the idea from it. If you built the record properly, you can show the path rather than assert it. Agencies that answer this objection with taste arguments lose. Agencies that answer it with the derivation win.

"Can we see it another way?" Sometimes a real request, often a stall from someone who cannot approve alone. Answer it by asking which part is not working: the strategy, the idea, or the execution. Those three have completely different remedies and the client usually has not separated them. Naming the layer converts a vague retreat into a scoped revision.

"What did the research say?" Usually asked by the person who has to defend the spend upwards. Have the research in the appendix, cite it by name, and offer the full version. This objection is friendly and agencies frequently misread it as hostile.

"Let's take it away and discuss internally." This is Room Two announcing itself. The single highest-value move in the entire presentation is what you do in the next sixty seconds. Offer to record a five-minute walkthrough for whoever was not in the room, and offer to join the internal discussion. Agencies that let the deck travel alone lose control of the narrative at precisely the moment the actual decision gets made.

Live, async, or hybrid: choosing the format

Live and in person is still the strongest format for a first concept presentation with a new client or on a high-stakes campaign. You get the room's reaction in real time, and reaction is data.

Live and remote is the working default. It costs you peripheral vision: you cannot see the person who has stopped nodding. Compensate by asking for cameras on for the reveal beat and by explicitly checking in after beat four.

Async is right for iterative rounds, for global teams across time zones, and for revisions on an already-approved territory. It is wrong for a first concept reveal, because the reveal is a performance and async removes the pacing that makes a reveal land.

Hybrid is what most good agencies actually do: present live, then immediately send an async package built for Room Two, containing the deck plus a short recorded walkthrough. The walkthrough is the single cheapest insurance policy in agency creative. Five minutes of your voice attached to the deck means the version that reaches the budget holder still has an advocate in it.

What agencies use to build and hold the presentation

Match the tool to the room rather than looking for one tool that does everything, because nothing does all three well.

For Room One, the live presentation. Figma Slides if the team already lives in Figma and the work is design-heavy. Pitch if you present often and want viewer analytics telling you who actually opened it. Canva if speed matters more than craft and non-designers need to edit. Keynote if the presentation is a performance and you want the best transitions available. All four are strong. The choice is mostly about where your team already works.

For Room Two, the forward. Loom for the recorded walkthrough. It is the standard for a reason: it takes five minutes, and it puts a narrator back into a deck that would otherwise travel silent. Pair it with the deck as a PDF, because PDFs open everywhere and links to editable files trigger permission friction at exactly the wrong moment.

For Room Three, the record. This is the gap in most agency stacks. The requirement is a place where the brief, the research, the insight, the chosen idea, the executions and the rejected routes with their reasons all sit in one retrievable space that a person can open in week six and understand. Miro and Mural do this if the team maintains them, and in practice workshop boards decay into unlabelled sticky notes within a fortnight. Notion does it if somebody owns the structure, and it flattens visual work into documents. Milanote holds visual references beautifully and does not hold reasoning.

Storyflow is built for this third room specifically. The campaign brief, the research, the concept board, the references and the rejected routes live on one canvas, and the AI reads across the whole board rather than one card, so asking why a route was chosen returns an answer assembled from what is actually on the board. That is the job it does, and it is the reason it belongs in an agency stack.

Where Storyflow loses, plainly: there is no slide mode, no presenter view and no speaker notes, so you cannot present from it. It is early access, so the template library is smaller than Milanote's and the integrations list is shorter than Notion's. There is no offline mode. It is the record and the thinking, not the performance. An agency that tries to replace its slide tool with it will have a bad meeting.

For the tool comparison in depth, see The Best Tools for Presenting Design Concepts to Clients and The Best AI Tools for Agencies for Campaign Planning.

The forty-eight hours after the room

The presentation is roughly forty percent of the outcome. The forty-eight hours afterwards are most of the rest.

Within two hours: send a short written summary. Three bullets on what was presented, the decision requested, and the next milestone. No attachments yet. This is the anchor document, and it is what gets quoted in the internal thread you are not part of.

Within one day: send the Room Two package. The deck as a PDF, a five-minute recorded walkthrough, and one line naming who to contact with questions. Name a specific person, because "let us know" routes to nobody.

Within two days: follow up with a question rather than a nudge. "Did the phasing on slide eighteen answer the retail question?" restarts a conversation. "Just checking in" does not.

Throughout: keep the record current. Every objection raised and answered goes into the record while it is fresh. Six weeks later, that log is what lets you answer a reopened decision in four minutes rather than four hours.

How agency presentations differ by client type

Startups and founder-led brands. The founder is usually in the room and usually decides in the room. Present fewer routes, be more opinionated, and expect the conversation to go straight to execution. Room Two barely exists. Room Three matters enormously, because founder-led brands change direction fast and the record is what stops the same debate recurring monthly.

Mid-market marketing teams. The person in the room is rarely the person who signs. This is the environment Room Two was named for. Build the forwarded artifact with the same care as the presentation itself.

Enterprise and regulated categories. Legal and compliance will read the deck without you. Every claim needs a source in the appendix. Expect the concept to be approved in principle and then re-approved in detail. The record is not optional here, it is a deliverable.

Public sector and non-profit. Committees, published criteria, and often a formal scoring rubric. Structure the deck to the rubric explicitly, even where that hurts the narrative flow. This is the one context where matching the client's format beats presenting well.

Mistakes that read as junior

• Opening with the agency credentials. Nobody in the room booked the meeting to learn about you.
• Showing executions before the idea. The work then has to explain itself, which it cannot.
• Presenting three equal routes with no recommendation.
• A deck of images with no words, forwarded to someone who was not there.
• Answering "it doesn't feel like us" by defending the craft instead of re-deriving the idea.
• Ending on "any questions."
• Letting the deck travel without a narrator.
• Treating the rejected routes as waste rather than as the most valuable part of the record.
• Quoting production cost before the idea is agreed.
• No version number on a deck that will be forwarded four times.

The Bottom Line

Agencies that present well are not better performers. They are better packagers. They understand that the concept has to survive three rooms, and they build a different artifact for each one: a paced live presentation, a self-explaining forwarded package, and a durable record of the reasoning.

The sequence is fixed and it works: their problem, what we found, the insight, the idea, the work, how it runs, the decision. Two routes, ranked, with a stated recommendation. Rejected work kept out of the room and kept in the record. A named decision instead of "any questions." A recorded walkthrough sent the same day, because the meeting where your campaign is actually approved is one you are not invited to.

Build the deck for the room you are in, and build the record for the room you will not be in.

Present from Figma Slides, Pitch, Canva or Keynote. Send the forward with Loom. Hold the record where the reasoning stays attached to the work: Storyflow keeps the brief, the research, the chosen concept and the rejected routes on one canvas, and its AI reads the whole board, so in week six the question "why did we go this way" has an answer that is still there. It will not present for you, and it is not trying to.

Author

Justkay is a documentary filmmaker and the founder of Storyflow. He has pitched documentary treatments to broadcasters and commissioners, and has spent a decade watching creative work get approved, forwarded, and quietly reinterpreted on the way to the person who signs.