Agency Operations / Strategy Beginner

27 ChatGPT prompts for digital marketing agencies to generate client strategy decks

If you run a digital marketing agency, you already know the dirty little secret: the strategy deck is where most of the billable hours die. You spend eight hours on a 15-slide narrative, walk a client through it on Zoom, hear “let me think about it,” and then watch the thread go quiet. I have lived this exact arc more times than I want to count, and I built this guide because the fix is not a new slide template. The fix is a better prompt chain.

These 27 ChatGPT prompts for agency client strategy deck workflows turn a messy kickoff transcript, a Google Analytics export, and a few client emails into a tight, decision-ready 10-slide story in roughly five working days. I’ll show you the full prompt library, the deck blueprint, a 5-day rebuild workflow, and the 2026 data behind why most decks still fail.

Quick answer (TL;DR): A modern 2026 strategy deck follows a 10-slide blueprint (situation, audit, audience, positioning, goals, channel mix, calendar, budget, measurement, risk). Twenty-seven chained ChatGPT prompts cover every section. The biggest leverage points are audit (Prompts 1–5) and measurement (Prompts 20–23). The biggest mistake is shipping a deck that reads like a capability tour instead of a decision document.

Why clients ghost your strategy decks in 2026

The data is brutal. According to a Reuters Institute 2026 trends report (updated 12 January 2026), buyers are reading more but trusting less. Add the Wikipedia entry on Digital marketing (last edited 5 June 2026), which now flags that the entire discipline has shifted toward “generative engine optimization” and “answer engine optimization” - and the implication is clear. The deck you wrote in 2023 reads like a brochure from a bygone era.

The stat to remember: OpenAI’s ChatGPT reached 900 million weekly active users in February 2026 (Wikipedia: ChatGPT, 2026). Your client’s buyers are asking ChatGPT what agency to hire before they ever open your PDF. If your deck is not on-brand in that conversation, you’ve already lost.

The economic pressure is just as real. As of mid-2025, 83% of Chinese respondents, 65% of US respondents, and 54% globally report using generative AI (Wikipedia: Generative AI, 2025). Your client is using ChatGPT to pressure-test your deck in real time. If the answers line up, you get paid. If they don’t, you get a polite “we’ll circle back.”

And the deck has a second job. Wikipedia’s ChatGPT page (last edited May 2026) documents that in January 2026, OpenAI launched ChatGPT Health, and in March 2026 began showing ads to logged-in US adult users. The implication for agencies: prospects will increasingly land on your deck having already had a conversation with a chatbot about what their strategy should look like. Your prompts must produce output that survives that comparison.

The 10-slide strategy-deck blueprint for 2026

A strategy deck is a 10–15 slide document that walks a client from “here’s what we found” to “here’s what we recommend you sign off on today.” It is not a capabilities deck. It is not an agency intro. It is a decision document.

The 10 slides that work in 2026:

Cover & executive hook - one line that frames the business problem
Situation snapshot - what we know about you in 60 seconds
Audit & evidence - what the data says, in plain language
Audience & ICP - who we’re going after and why
Positioning & narrative - how we’ll frame the message
Goals & KPIs - what success looks like, in numbers
Channel mix & tactics - where the work happens
Calendar & milestones - the 90-day and 12-month rhythm
Budget & resourcing - what it costs, who does what
Measurement & risk - how we report, what we cut if it fails

Slide 11 is almost always a Q&A appendix. Slide 12 is the sign-off ask. Everything past slide 12 is your scope document, not your strategy deck.

The 27 prompts below map 1:1 to this blueprint. They are written for GPT-5.5 but work on Claude, Gemini, and Mistral with light tweaks. The format is intentionally tight: a multi-line prompt, an example output, and a pro tip.

Section 1 - Audit & situation prompts (Prompts 1–5)

A strategy deck without an audit is just an opinion. The five prompts in this section turn raw client inputs (transcripts, exports, brand briefs) into a defensible “here’s what we know” story.

Prompt 1 - Kickoff call distiller

Purpose / context: Turn a 60-minute kickoff call transcript into a structured “situation snapshot” that becomes slide 2 of your deck. This is the prompt I run first, before I touch the slide.

The prompt:

You are a senior strategy lead at a digital marketing agency. I'll paste the
raw transcript of a 60-minute kickoff call with a new client. Your job is to
produce a "Situation Snapshot" for slide 2 of our strategy deck.

Extract and structure the following in clean markdown:

1. COMPANY FACTS - 3-5 bullets covering what the company does, who it sells
   to, and the founding moment that matters for the story.
2. CURRENT STATE - what marketing is happening today, channels, team, and
   tooling. Be specific. If they mention HubSpot, name HubSpot.
3. THE PROBLEM THEY HIRED US TO SOLVE - one sentence in the client's own
   words, then a paraphrase we can defend in the deck.
4. INTERNAL CONSTRAINTS - budget band, team size, legal/compliance issues,
   politics, or any "we tried that, it didn't work" history.
5. EXTERNAL CONSTRAINTS - competitors they fear, market trends they care
   about, regulatory issues, or seasonality.
6. SUCCESS DEFINITION - how the client will judge us in 90 days and in 12
   months. Quote them where possible.
7. OPEN QUESTIONS - anything ambiguous, missing, or contradictory that we
   need to resolve before we write another slide.

Tone: neutral, evidence-led, no agency hype. If the client said something
self-contradictory, flag it under OPEN QUESTIONS. Output under 400 words.

TRANSCRIPT:
"""
[PASTE TRANSCRIPT HERE]
"""

Example output (abbreviated):

COMPANY FACTS

B2B SaaS, project management for architecture firms, 38 employees, founded 2021.

Founder previously ran a 12-person firm; built the product for her old pain.

THE PROBLEM THEY HIRED US TO SOLVE

Client words: “We have 2,000 signups a month and almost no paid users.”

Paraphrase: Top-of-funnel is healthy, mid-funnel conversion is broken.

Pro tip: Run this prompt on the transcript the same day as the call. Memory fades after 48 hours and you’ll start inventing details. I keep a Notion database of these snapshots - one row per client, this prompt is the cell.

Prompt 2 - Google Analytics 4 narrative generator

Purpose / context: Convert a raw GA4 acquisition + engagement export into a 3-bullet story that lands on slide 3. Stops you from over-promising on a channel that’s actually flat.

The prompt:

You are a marketing analyst. I'll paste a GA4 export covering the last 90
days (acquisition, engagement, conversions). Produce a 3-bullet narrative I
can drop into slide 3 of a strategy deck.

Rules:
- Each bullet must lead with the channel, then the number, then the
  interpretation. Example: "Organic search drove 41% of sessions, but the
  conversion rate is half of paid social - suggesting bottom-of-funnel
  intent is missing."
- One bullish bullet, one bearish bullet, one "watch this" bullet.
- No more than 90 words total.
- Flag any data anomaly (a channel that spiked one week and disappeared,
  a campaign with high clicks and zero conversions, etc.).
- If the export is missing a critical view, say so.

GA4 EXPORT:
"""
[PASTE CSV OR TABLE HERE]
"""

Example output:

Bullish. Branded search drove 22% of conversions at a 6.4% conversion rate - these are people who already know us, and they convert 3x the site average.

Bearish. Paid social produced 38% of sessions but only 9% of conversions. We are paying for attention we cannot convert.

Watch. Direct traffic jumped 40% in week 6, which lines up with a PR mention. We should double down on earned media.

Pro tip: Always paste raw numbers, not summaries. The whole point of this prompt is to catch the story you didn’t notice in the data. I run this prompt twice - once with the full export, once with just the top-10 traffic sources - to see if the model is consistent.

Prompt 3 - Competitor teardown

Purpose / context: Feed in three competitor URLs plus a positioning hypothesis. Get back a clean “what they’re saying vs. what they’re doing” table that becomes the spine of your positioning slide.

The prompt:

You are a competitive intelligence analyst. I'll give you three competitor
URLs (homepage, pricing page, and one product page) plus a one-line
hypothesis about why clients choose them. Your job:

1. Summarize each competitor's positioning in 1 sentence (their homepage
   hero text, paraphrased).
2. List the 3 claims they make most often (e.g., "fastest", "easiest",
   "for enterprise").
3. Identify 2 weaknesses in their positioning that we can exploit.
4. Compare their pricing model and packaging tiers.
5. Tell me the one message they're all repeating that we should NOT copy.

Format: a markdown table with one column per competitor. Output under 500
words.

COMPETITOR A: [URL]
COMPETITOR B: [URL]
COMPETITOR C: [URL]
OUR HYPOTHESIS: [1 sentence]

Example output (abbreviated):

	Competitor A	Competitor B	Competitor C
Positioning	”The fastest way to ship"	"Built for enterprise compliance"	"For makers, not managers”
Repeating	”10x faster"	"SOC 2 + HIPAA"	"No code required”

Pro tip: Use this prompt on the same day you build the positioning slide. If you can’t fill the “weakness” column honestly, your positioning hasn’t been stress-tested. I keep a separate Notion DB of competitor teardowns indexed by industry.

Prompt 4 - Brand voice & tone extractor

Purpose / context: Pull a brand voice profile out of an existing site, blog, or social archive. Used to write the positioning slide without making it sound like every other agency deck.

The prompt:

You are a brand strategist. I'll paste a sample of existing copy (website
homepage, 5 recent blog posts, and the last 30 LinkedIn posts). Produce a
"Brand Voice Profile" with the following:

1. VOICE ATTRIBUTES - pick 3 from this list and justify with a quoted
   example: confident, playful, clinical, direct, witty, warm, irreverent,
   authoritative, plainspoken, academic.
2. VOCABULARY TO KEEP - 5 words/phrases the brand uses a lot.
3. VOCABULARY TO AVOID - 5 words/phrases that would feel off-brand.
4. SENTENCE RHYTHM - short and punchy, mixed, or long and explanatory.
5. POINT OF VIEW - first person ("we"), second person ("you"), or third
   person ("the team"). With examples.
6. TONE SHIFTS - any moments where the brand breaks its own voice
   (humor, sarcasm, vulnerability) and whether they land.

Output: a tight 200-word profile. If the brand voice is inconsistent,
say so - that's the most useful finding.

COPY SAMPLES:
"""
[PASTE HERE]
"""

Example output:

VOICE ATTRIBUTES: plainspoken, witty, warm. KEEP: “small but mighty”, “ship it”, “good enough is the enemy”. AVOID: “synergy”, “leverage”, “best-in-class”. SENTENCE RHYTHM: mixed, but prefers short. Average sentence: 12 words. TONE SHIFTS: Drops the “we” voice in case studies and switches to a customer quote - this is the strongest moment in the copy.

Pro tip: If the brand voice is inconsistent, that’s a finding, not a failure. Put the inconsistency on the deck. “Your homepage says ‘enterprise.’ Your Twitter says ‘tinkerers.’ Pick one. Here’s our recommendation.”

Prompt 5 - Quick-win detector

Purpose / context: Identify 3–5 changes that can ship in under 30 days and show measurable lift. Becomes the “early wins” section of your strategy deck and sets the tone for the relationship.

The prompt:

You are a fractional CMO. I'll paste the situation snapshot from Prompt 1,
the GA4 narrative from Prompt 2, and the competitor teardown from Prompt 3.
Produce a "Quick Wins" list for our strategy deck.

Rules:
- 3 to 5 wins maximum.
- Each win must be shippable in 30 days or less, with the current team.
- Each win must be measurable. Name the metric, the baseline, and a
  realistic 30-day target.
- No "rebuild the website" - too big.
- No "post more on social" - too vague. Specify the platform, the format,
  and the cadence.
- For each win, name the owner (client-side or agency-side).
- Flag any quick win that conflicts with the bigger strategy in slide 7.

Output as a numbered list. Under 350 words.

SITUATION SNAPSHOT:
[PASTE]

GA4 NARRATIVE:
[PASTE]

COMPETITOR TEARDOWN:
[PASTE]

Example output:

Pause branded search budget for 14 days. The conversion rate is 6.4% but the cost-per-click has crept up 20%. Baseline CPL: $42. Target: $32. Owner: agency media buyer.

Rewrite the homepage hero. Current hero is “Welcome to Acme.” The competitor teardown shows the buyer expects a benefit-led hero. Ship a new hero in 7 days. Owner: agency copy lead + client designer.

Add a single intent popup on the pricing page. Trigger: exit-intent, copy: “Not sure which plan? Talk to a human.” Baseline: 0.3% popup conversion. Target: 1.2%. Owner: agency dev.

Pro tip: Resist the urge to put a 6-month initiative here. Quick wins should fit in a single sprint, with a single owner, and a single metric. The deck sells the speed, not the ambition.

Section 2 - Goals & KPIs prompts (Prompts 6–9)

A strategy deck without goals is a wish list. The four prompts in this section force the client to commit to numbers, attribution windows, and a single primary KPI.

Prompt 6 - Goal ladder builder

Purpose / context: Convert fuzzy goals (“grow the business”) into a tiered goal ladder: business outcome → marketing outcome → channel outcome. Becomes slide 6.

The prompt:

You are a strategy lead. I'll give you a client's stated goal in their own
words. Your job is to ladder that goal into 3 tiers.

TIER 1: BUSINESS OUTCOME - what the CEO cares about (revenue, gross
margin, valuation, headcount). Must be in dollars or people.
TIER 2: MARKETING OUTCOME - what marketing directly moves (qualified
pipeline, MQLs, conversion rate, brand awareness index). Must be a number.
TIER 3: CHANNEL OUTCOME - what each channel contributes (clicks,
impressions, signups, demo requests, etc.). Must be a number we can pull
from a dashboard.

For each tier, give me 1 to 3 candidates. For each candidate, give me a
realistic 12-month target and a stretch 12-month target. Flag any tier
where we don't have a clean measurement today - that's a data gap to
close.

Output as a markdown table. Under 250 words.

CLIENT'S STATED GOAL:
"""
[PASTE]
"""

Example output:

Tier	Candidate	12-mo target	Stretch	Data gap?
1. Business	ARR	$4.2M	$5.1M	none
2. Marketing	Qualified pipeline	$1.8M	$2.4M	clean
3. Channel	MQLs (paid)	320/mo	460/mo	none

Pro tip: If the client can’t give you a Tier 1 number, the deck can’t move forward. Push for it in the kickoff, not in the strategy review.

Prompt 7 - KPI dictionary

Purpose / context: Produce a one-page “what we mean by X” document that prevents the most common client-agency argument: “you said leads, we said pipeline.” Becomes a hidden appendix of the deck.

The prompt:

You are a marketing operations consultant. I'll give you the 6 to 10
metrics our agency is committing to track for a client. Produce a "KPI
Dictionary" with the following columns:

- METRIC NAME (plain English)
- DEFINITION (in one sentence, no jargon)
- FORMULA (the math)
- DATA SOURCE (GA4, HubSpot, Salesforce, etc.)
- REPORTING CADENCE (daily, weekly, monthly, quarterly)
- OWNER (client or agency)
- COMMON MISUNDERSTANDING (1 sentence - the thing clients usually get
  wrong about this metric)

If any two metrics are mathematically redundant (e.g., MQLs and MQL-to-SQL
rate), flag it. We only ship metrics that earn their slot.

METRICS WE'RE COMMITTING TO:
[LIST]

Pro tip: The “common misunderstanding” column is the most valuable part. It pre-empts 80% of the “but you said…” emails. I send this dictionary to the client before the strategy review and ask them to mark anything they disagree with.

Prompt 8 - North Star Metric challenger

Purpose / context: A North Star Metric is the single number that best captures the value the product delivers to the customer. Use this prompt to pressure-test the client’s choice before it goes on slide 6.

The prompt:

I'm going to give you a candidate North Star Metric for a client. Your job
is to challenge it.

A good North Star Metric has 4 properties:
1. It reflects customer value delivered (not vanity).
2. It leads revenue (not lags it).
3. The team can move it (it's actionable).
4. It's a count or rate, not a percentage that hides volume.

CANDIDATE METRIC: [NAME]
CONTEXT: [1-2 sentences on the business]
BASELINE: [current value]

Reply with:
- VERDICT: keep, refine, or replace.
- If refine, give me the new metric and the 1-sentence reasoning.
- If replace, give me 2 alternative North Stars and the trade-off.
- One warning sign that this metric is being gamed internally.

Pro tip: AARRR (Acquisition, Activation, Retention, Revenue, Referral) - the Pirate Metrics framework coined by Dave McClure - is a useful sanity check, but the North Star should sit above the AARRR funnel, not inside it.

Prompt 9 - 90-day success scorecard

Purpose / context: Build the “what does success look like in 90 days” page of the deck. This is what protects the agency if the engagement goes sideways.

The prompt:

You are a client services director. I'll paste the goal ladder from Prompt
6 and the KPI dictionary from Prompt 7. Produce a 90-day scorecard for
slide 6 of our strategy deck.

Format: a 3-column markdown table - METRIC, BASELINE, 90-DAY TARGET.
Then a 4th section titled "WHAT WE WILL NOT MEASURE IN 90 DAYS" with 2-3
metrics that are too noisy or too slow to judge the engagement.

Rules:
- Every metric must have a baseline from the client's data (or a
  flagged "no baseline, will establish in week 1").
- 90-day targets should be conservative, not aspirational. We'd rather
  over-deliver.
- If a metric is up to interpretation, write the interpretation in
  parentheses.

GOAL LADDER:
[PASTE]

KPI DICTIONARY:
[PASTE]

Pro tip: The “what we will NOT measure” section is the most underused move in agency decks. It saves you in the Q&A when a client asks, “but what about brand awareness?” You say, “Great question - and here’s why we deliberately deferred it to month 6.”

Section 3 - Channel & tactic prompts (Prompts 10–14)

The channel mix is where most decks go soft. “We’ll do SEO, paid, content, email, social, and PR” is not a strategy. The five prompts in this section force trade-offs.

Prompt 10 - Channel mix matrix

Purpose / context: Map every potential channel against the goal ladder and the budget, then rank by expected ROI. Becomes the spine of slide 7.

The prompt:

You are a media planner. I'll give you the goal ladder (from Prompt 6) and
a monthly budget range. Produce a Channel Mix Matrix.

For each channel, score it on 4 dimensions (1-5 scale):

- FIT - how well it serves the audience and the goal.
- COST - relative cost to reach the audience (1 = cheap, 5 = expensive).
- SPEED - how fast we can see signal (1 = months, 5 = days).
- CONTROL - how much we can iterate and optimize (1 = low, 5 = high).

Then give me a TOTAL (sum of 4 scores) and a VERDICT (lead, support,
test, kill).

Channels to consider (add any I'm missing):
- Organic search (SEO)
- Paid search (SEM)
- Paid social (Meta, LinkedIn, TikTok)
- Organic social
- Email (lifecycle)
- Content (blog, podcast, video)
- PR / earned media
- Partnerships / affiliates
- Events (in-person)
- Community (Discord, Slack, forum)
- Outbound (cold email, LinkedIn)

GOAL LADDER:
[PASTE]

BUDGET: $[X] / month

Output as a markdown table.

Example output:

Channel	Fit	Cost	Speed	Control	Total	Verdict
Paid LinkedIn	5	5	4	5	19	Lead
SEO	5	3	1	4	13	Support
Cold email	4	2	5	5	16	Lead

Pro tip: If every channel scores 16+, the matrix is broken. Push the agency team to actually rank. A good strategy deck kills at least 2 channels.

Prompt 11 - Tactic expansion

Purpose / context: For each “lead” channel from Prompt 10, generate 5 specific tactics with measurable outputs. Stops the deck from saying “run paid social” without saying how.

The prompt:

For the channel "[CHANNEL]" generate 5 specific tactics we can ship in
the next 90 days. For each tactic, include:

- TACTIC NAME
- ONE-LINE DESCRIPTION
- ASSET(S) NEEDED (creative, copy, landing page, tool, etc.)
- ESTIMATED HOURS (agency + client)
- METRIC WE'LL MOVE
- FIRST 7-DAY CHECKPOINT (what we'll look at to know if it's working)

Rules:
- No vague tactics like "increase engagement" or "build brand."
- Each tactic must name a specific platform, format, or campaign type.
- Include at least one "test" tactic (low budget, high learning) and at
  least one "scale" tactic (higher budget, lower learning curve).
- Output as a numbered list. Under 500 words.

Pro tip: Tactics that take more than 8 agency hours need a sub-prompt. I run this once for the high-level view, then again for each tactic that needs a creative brief.

Prompt 12 - Content pillar generator

Purpose / context: Generate the 3–5 content pillars that the editorial calendar will hang off. Used by the agency content team and the client for thought leadership.

The prompt:

You are a content strategist. I'll give you the audience profile, the
positioning hypothesis, and the channel mix. Generate 4 content pillars
for the next 6 months.

For each pillar, include:
- PILLAR NAME (3 words max)
- AUDIENCE INTENT it serves (the question it answers)
- THE 3 SUB-TOPICS we'd cover in month 1
- THE FORMAT we'd lead with (written, video, audio, interactive)
- THE INTERNAL SME we'd quote
- ONE COMPETITOR who already does this well (and one anti-example)

The pillars must collectively cover 80% of the buying journey. The pillar
that doesn't get traffic is the one we double down on, not kill.

AUDIENCE PROFILE:
[PASTE]

POSITIONING HYPOTHESIS:
[PASTE]

CHANNEL MIX:
[PASTE FROM PROMPT 10]

Pro tip: The StoryBrand framework (Donald Miller) is a useful gut-check: every pillar should map to a “character has a problem, guide gives a plan, guide calls to action, failure is avoided, success is achieved” arc. If a pillar doesn’t, cut it.

Prompt 13 - Funnel stage mapper

Purpose / context: For each channel and each tactic, name the funnel stage and the metric that proves the stage is working. Stops the deck from claiming “brand awareness” with zero measurement.

The prompt:

Map every tactic from the channel mix (Prompt 10) and the content
pillars (Prompt 12) to one or more of these funnel stages:

1. AWARENESS - prospect becomes aware of the problem and the brand.
2. CONSIDERATION - prospect compares us to alternatives.
3. CONVERSION - prospect takes a revenue action.
4. RETENTION - existing customer stays and expands.
5. ADVOCACY - customer refers or creates content.

For each tactic, give me:
- STAGE (one of the 5 above)
- ENTRY METRIC (the signal the tactic is reaching the right people)
- EXIT METRIC (the signal the tactic is moving them down the funnel)
- IF THIS BREAKS, NEXT MOVE (the contingency tactic)

If a tactic covers more than 2 stages, it's probably too vague - flag
it.

Output as a markdown table.

Pro tip: A common failure mode is over-investing in awareness with no consideration engine. The Bowtie Funnel framework is useful here: it’s not a single funnel, it’s a “funnel” for acquisition and a separate funnel for retention, joined at the conversion event. Most agencies miss the second half.

Prompt 14 - Differentiation pressure test

Purpose / context: Stress-test the channel mix by trying to copy it with a smaller budget. If a competitor can replicate the strategy in 30 days, it’s not a strategy - it’s a tactic.

The prompt:

You are a cynical competitor strategist. Look at our channel mix and
proposed tactics (paste below). Then answer:

1. If a competitor with 1/3 of our budget tried to copy this strategy,
   what would they do first, and how long would it take?
2. Which tactic in this mix is the hardest to copy and why?
3. Which tactic looks impressive but is probably the easiest to copy?
4. What is the single piece of intellectual property (proprietary data,
   unique process, exclusive partnership) that would make this strategy
   defensible for 12+ months?
5. What is the most likely failure mode in the first 90 days, and how
   would we detect it early?

Output as 5 numbered answers. Be harsh. We need this to survive a board
meeting.

CHANNEL MIX:
[PASTE]

TACTICS:
[PASTE FROM PROMPT 11]

Pro tip: Run this prompt on the strategy before the client sees it. If the answers are vague, the strategy is vague. I have a rule: if the model can name the failure mode, we ship. If it can’t, we don’t.

Section 4 - Calendar & budget prompts (Prompts 15–19)

A strategy without dates and dollars is a thought experiment. The five prompts in this section turn the channel mix and tactic list into a working plan.

Prompt 15 - 90-day rollout calendar

Purpose / context: Build a week-by-week calendar for the first 90 days. Becomes the body of slide 8.

The prompt:

You are a project manager at a digital marketing agency. Build a 90-day
rollout calendar from the channel mix and tactic list below.

For each week (W1 through W13), give me:
- THE FOCUS (one sentence)
- THE TACTICS we're shipping that week
- THE ASSETS that need to be ready by the Friday before
- THE REVIEW MEETING (date, attendees, agenda)
- THE HANDOFF to the client (what they own starting that week)
- A GO/NO-GO CRITERION for whether we proceed to the next phase

Calendar constraints:
- W1-W2: setup, tracking, contracts, kickoff
- W3-W6: launch phase 1
- W7-W10: iterate based on phase 1 data
- W11-W13: phase 2 plan + first quarterly review

CHANNEL MIX:
[PASTE]

TACTICS:
[PASTE]

Pro tip: Block “innovation” weeks into the calendar. The Media Strategy 360 framework (planning across paid/owned/earned in 90-day cycles) only works if you reserve at least 1 week per cycle for unexpected experiments.

Prompt 16 - Monthly budget allocator

Purpose / context: Convert the total monthly budget into a per-channel allocation with named line items. Becomes slide 9.

The prompt:

You are a finance lead at a digital marketing agency. I'll give you the
total monthly budget and the channel mix. Produce a monthly budget
allocator in markdown table form.

For each channel:
- LINE ITEM (e.g., "LinkedIn Sponsored Content", "SEO blog writing",
  "HubSpot retainer")
- AMOUNT (in USD)
- % OF TOTAL
- VENDOR OR TOOL (if any)
- INVOICE CADENCE (monthly, quarterly, one-time)
- WHO APPROVES (client-side approver name and role)

Then a 4th section: "VARIANCE ALLOWANCE" - the % of budget we can shift
between channels in-month without client approval. Default is 10%.

TOTAL MONTHLY BUDGET: $[X]
CHANNEL MIX:
[PASTE]

Pro tip: If a single line item is more than 40% of the budget, add a “kill criteria” note: the condition under which the agency can recommend pausing it. This pre-empts the “why are we still spending $20k on LinkedIn?” question at the quarterly review.

Prompt 17 - Resourcing matrix

Purpose / context: Name the humans on the agency side and the client side for every tactic. Catches the most common reason agency engagements fail: nobody is on the hook.

The prompt:

You are an agency operations lead. I'll paste the 90-day calendar and
the monthly budget. Produce a Resourcing Matrix.

Two sides:
- AGENCY SIDE: which agency role owns what (Strategist, Media Buyer,
  SEO Lead, Designer, Copywriter, Developer, PM).
- CLIENT SIDE: which client role owns what (Marketing Lead, Sales
  Lead, Product Lead, Executive Sponsor, Legal/Compliance).

For every line in the calendar, name the agency owner AND the client
owner. If a line has no client owner, flag it as "AT RISK" - agency
work will stall.

Add a RACI column:
- R = Responsible (does the work)
- A = Accountable (signs off)
- C = Consulted (gives input)
- I = Informed (gets the update)

90-DAY CALENDAR:
[PASTE]

BUDGET:
[PASTE]

Pro tip: “AT RISK” is the most important word in this matrix. If a tactic has no client owner, you will do the work but it will be ignored. Better to find out in the deck review than in the monthly retro.

Prompt 18 - Dependency map

Purpose / context: Identify what needs to happen first, what can happen in parallel, and what blocks what. Saves you from shipping a 90-day plan that’s secretly a 9-month plan.

The prompt:

You are a delivery architect. Look at the 90-day calendar and the
resourcing matrix. Produce a Dependency Map.

For every tactic or deliverable, identify:
- WHAT IT BLOCKS (which other tactics can't start until this is done)
- WHAT BLOCKS IT (which other tactic must finish first)
- THE CRITICAL PATH (the longest chain of dependencies - this is the
  real delivery date)
- THE FLOAT (how many days a tactic can slip without delaying
  critical path)
- THE SINGLE POINT OF FAILURE (one person or one asset that, if
  delayed, delays everything)

Output as a markdown table, sorted by critical path position. Under 300
words.

Pro tip: If the critical path runs through a single client approver, the strategy is fragile. Push the client to add a backup approver in the kickoff.

Prompt 19 - Scope creep firewall

Purpose / context: Pre-write the responses to the requests the client is going to make that aren’t in the original scope. Becomes the “out of scope” page of the deck.

The prompt:

You are a client services director. Look at the strategy and budget
below. List the 5 most likely out-of-scope requests the client will make
in the first 90 days. For each:

- THE REQUEST (in the client's voice)
- THE REASON IT'S NOT IN SCOPE (1 sentence)
- THE FIXED-PRICE QUOTE we'd offer if they say yes
- THE TRADE-OFF (what we'd stop doing to free up the hours)

This is for our internal use, not the deck. Be specific.

STRATEGY:
[PASTE]

BUDGET:
[PASTE]

Pro tip: When the client inevitably asks for “just one more landing page” at month 2, you pull out this matrix. It saves 30 hours of awkward negotiation.

Section 5 - Measurement & reporting prompts (Prompts 20–23)

Reporting is where agencies bleed margin. The four prompts in this section build a measurement system that runs on autopilot and produces a deck-ready update.

Prompt 20 - Reporting dashboard architect

Purpose / context: Pick the right tools and the right charts for the client’s KPIs. Becomes the appendix of the strategy deck and the source of the monthly report.

The prompt:

You are a marketing operations architect. I'll give you the KPI
dictionary, the channel mix, and the budget. Design a Reporting Dashboard
spec.

For each tier, recommend:
- THE TOOL (e.g., Google Looker Studio, HubSpot, Airtable, Notion,
  ClickUp, Slack for alerts, Otter for call summaries, Loom for video
  recaps, Pitch for decks)
- THE CHART TYPE (line, bar, stacked, funnel, table, scorecard)
- THE CADENCE (live, daily, weekly, monthly, quarterly)
- THE DATA SOURCE (the source of truth for that metric)
- THE OWNER (who fixes it when it breaks)
- THE COST (tool license, agency hours to maintain)

Two principles:
1. The number of charts should not exceed the number of KPIs. No
   vanity charts.
2. If two tools can produce the same chart, we pick the one the client
   already pays for.

KPI DICTIONARY:
[PASTE]

CHANNEL MIX:
[PASTE]

Pro tip: For a 5–10 person agency, the cheapest viable stack in 2026 is: Google Looker Studio for charts, Airtable for campaign tracking, Notion for docs and strategy, ClickUp for the agency PM, Slack for client comms, and Pitch or Google Slides for decks. Resist the urge to add a new tool per client.

Prompt 21 - Monthly report generator

Purpose / context: Generate the actual monthly report text from raw numbers. This is the prompt that saves 6+ hours of human writing time per month.

The prompt:

You are a marketing analyst writing a client-facing monthly report.
I'll paste the dashboard data for the month. Produce a report in this
exact structure:

1. EXECUTIVE SUMMARY (3 sentences max, plain English, no jargon)
2. THE WINS (1-2 bullets, with the number and the why)
3. THE MISSES (1-2 bullets, with the number and the leading hypothesis)
4. THE EXPERIMENTS (what we tested, what we learned)
5. THE BUDGET BURN (actual vs. planned, with a flag if variance > 15%)
6. THE NEXT 30 DAYS (3 priorities)
7. THE ASKS (what we need from the client to hit next month's goals)

Tone: confident, not defensive. If we missed, we say so and we say
what we're doing about it. Under 600 words.

DASHBOARD DATA:
[PASTE]

Pro tip: Read the report out loud before sending. If it sounds like a press release, rewrite. Clients want to know if you missed and why, not how clever you are.

Prompt 22 - Anomaly detector

Purpose / context: Catch the data issues that human eyes miss. Becomes a weekly Slack alert in the agency’s #client-updates channel.

The prompt:

You are a data analyst. I'll paste this week's dashboard numbers (last
8 weeks). Identify any anomalies that meet at least one of these rules:

- A metric moved more than 25% week-over-week with no obvious cause.
- A channel's cost-per-acquisition changed by more than 30% from the
  trailing 4-week average.
- A conversion event happened more than 2 standard deviations from
  the trailing 8-week mean.
- A metric has been declining for 3+ consecutive weeks.
- Two metrics that usually move together have diverged.

For each anomaly, give me:
- THE METRIC
- THE CHANGE (with numbers)
- THE LEADING HYPOTHESIS (1 sentence)
- THE NEXT DATA POINT we need to confirm
- THE ACTION (investigate further, or wait one more week)

If there are no anomalies, say so - that's the answer.

8 WEEKS OF DATA:
[PASTE]

Pro tip: Anomaly detection is one of the highest-ROI uses of AI in agencies. A human will miss the third week of a slow decline. A model that sees 8 weeks at once will catch it.

Prompt 23 - Attribution model explainer

Purpose / context: The single biggest source of conflict in agency-client relationships is “your attribution is wrong.” Use this prompt to pre-write the explanation that goes in the report.

The prompt:

You are a marketing measurement consultant. I'll give you the channel
mix and the conversion event. Explain the attribution model we'd use
in 250 words or less, in plain English.

Cover:
- LAST-CLICK (what it is, why it's biased toward branded search and
  direct)
- FIRST-CLICK (what it is, why it over-credits awareness)
- LINEAR (what it is, why it's a compromise, not a model)
- POSITION-BASED (40-20-40 - middle gets 20, ends get 40 each)
- DATA-DRIVEN (algorithmic, requires enough volume to be statistically
  valid)
- MARKETING-MIX MODELING (MMM - top-down, good for high-spend, slow
  cadence)

For our specific mix and conversion event, recommend ONE primary model
and ONE secondary model. Justify the choice. Flag the model that would
look best in the deck but is the most likely to be wrong in month 1.

CHANNEL MIX:
[PASTE]

CONVERSION EVENT:
[PASTE]

Pro tip: Marketing-Mix Modeling (MMM) has gone from “Fortune 500 only” to “every agency can run it” thanks to open-source tools like Meta’s Robyn and Google’s Meridian. If your client is spending more than $50k/mo, MMM is a credible alternative to platform-attributed numbers.

Section 6 - Risk & trade-off prompts (Prompts 24–27)

The last 4 prompts protect the agency from itself. They force the strategy to acknowledge what it does not know, what could go wrong, and what we will sacrifice.

Prompt 24 - Risk register

Purpose / context: Build the “things that could kill this strategy” page. This is the slide that wins trust with sophisticated buyers.

The prompt:

You are a risk consultant. Look at the strategy below. Produce a Risk
Register.

For each risk, give me:
- THE RISK (1 sentence)
- PROBABILITY (low / medium / high)
- IMPACT (low / medium / high)
- EARLY WARNING SIGNAL (the metric or event that would tell us it's
  happening)
- MITIGATION (the specific action we'd take)
- THE OWNER (client or agency)

Cover at least:
- Channel-specific risks (e.g., iOS privacy changes, Google algo
  updates, LinkedIn CPM spikes)
- Operational risks (e.g., client team turnover, agency staff loss,
  tool outages)
- Market risks (e.g., competitor launch, macro downturn, regulatory
  change)
- Measurement risks (e.g., attribution blackout, data loss,
  platform deprecation)

If a risk has high probability AND high impact, it gets a page in the
deck, not a footnote. Be honest. We lose deals when we pretend
nothing can go wrong.

STRATEGY:
[PASTE]

Pro tip: A risk register is also a sales tool. When a buyer sees an agency acknowledge the risks upfront, the trust goes up. I’ve closed deals by saying, “Here’s the part of the deck most agencies don’t show you.”

Prompt 25 - Trade-off generator

Purpose / context: Every strategy is a set of trade-offs. Use this prompt to name the ones the deck is making implicitly.

The prompt:

You are a senior strategist. Look at the strategy below. Identify the
5 most important trade-offs being made, even if the deck doesn't
name them.

For each:
- THE TRADE-OFF (what we are choosing A over B)
- WHY WE PICKED A (the rationale)
- THE PRICE OF B (what we lose by not doing B)
- THE TRIGGER that would make us re-evaluate (e.g., "if paid CPL
  exceeds $X, we revisit SEO")
- THE EXIT COST (how hard it is to switch later)

Output as a numbered list. No hedging. The client deserves to know
what they're not getting.

STRATEGY:
[PASTE]

Pro tip: The best client question to ask after the strategy review is, “What did we decide NOT to do?” If you can’t answer it, the strategy isn’t done.

Prompt 26 - Kill criteria

Purpose / context: Pre-commit to the conditions under which the agency will recommend pausing or killing a tactic. Saves 6 months of slow failure.

The prompt:

You are a CFO advising a marketing agency. For each tactic in the
strategy below, define:

- KILL CRITERION 1 (the absolute minimum performance - if the tactic
  goes below this for 30 consecutive days, we kill it)
- KILL CRITERION 2 (the comparative benchmark - if a similar tactic
  is doing materially better, we kill this one)
- PAUSE TRIGGER (the signal that says "investigate, but don't kill
  yet")
- RE-LAUNCH CONDITION (what would need to be true for us to bring
  it back)

Be honest. Most tactics should have a kill criterion. A tactic
without a kill criterion is a tactic we'll run forever because we
don't want to admit it failed.

Output as a markdown table.

Pro tip: “Kill criteria” is the most underused word in agency decks. The Positioning by April Dunford framework is useful here: positioning only works if you commit to not being for everyone. Tactics are the same. If we won’t kill it, we don’t actually have a strategy.

Prompt 27 - Quarterly strategy review template

Purpose / context: Build the template for the QBR so the agency doesn’t have to rebuild it every 3 months. Becomes a recurring deliverable.

The prompt:

You are a client services director. Build a Quarterly Strategy Review
(QBR) template I can reuse for any client.

The template has 5 sections:

1. WHAT WE SAID WE'D DO (paste from the prior strategy deck)
2. WHAT WE ACTUALLY DID (with a 1-line status: done, partial, killed,
   deferred)
3. THE NUMBERS (the 5 metrics from the scorecard, with baseline vs.
   target vs. actual)
4. WHAT WE LEARNED (3-5 insights that changed our thinking)
5. WHAT WE'RE DOING NEXT (the 90-day plan for the next quarter)

Add a final section: "WHAT THE CLIENT IS RESPONSIBLE FOR" - the 3
things we asked them to do this quarter, with their status.

Format: a Google Slides or Pitch template structure, with the
narrative for each slide under the slide title. Under 800 words.

Pro tip: A QBR is the most underrated sales tool in an agency. Done well, it’s how you get a renewal. Done poorly, it’s how you get fired. The trick is the “what we learned” section - that’s where the agency shows it’s actually thinking.

Comparison table - prompt category vs. deck slide vs. output

This is the table I print and stick on the wall. It maps every prompt to the slide it feeds and the deliverable it produces.

#	Prompt	Category	Deck slide	Primary output	Best tool downstream
1	Kickoff call distiller	Audit	Slide 2	Situation snapshot	Notion doc
2	GA4 narrative generator	Audit	Slide 3	3-bullet data story	Google Slides
3	Competitor teardown	Audit	Slide 3	Competitor table	Notion table
4	Brand voice extractor	Audit	Slide 5	Voice profile	Notion doc
5	Quick-win detector	Audit	Slide 3	Numbered win list	Linear / ClickUp
6	Goal ladder builder	Goals	Slide 6	Tiered goal table	Notion doc
7	KPI dictionary	Goals	Appendix	KPI definitions	Airtable base
8	North Star challenger	Goals	Slide 6	Single metric with verdict	Strategy memo
9	90-day scorecard	Goals	Slide 6	Scorecard table	Looker Studio
10	Channel mix matrix	Channels	Slide 7	Scored channel table	Notion doc
11	Tactic expansion	Channels	Slide 7	5 tactics per channel	ClickUp board
12	Content pillar generator	Channels	Slide 7	4 pillar briefs	Notion calendar
13	Funnel stage mapper	Channels	Slide 7	Stage-by-stage map	Airtable base
14	Differentiation pressure test	Channels	Slide 7	5 critical answers	Strategy memo
15	90-day rollout calendar	Calendar	Slide 8	Week-by-week plan	ClickUp / Airtable
16	Monthly budget allocator	Calendar	Slide 9	Per-line budget table	Airtable base
17	Resourcing matrix	Calendar	Slide 8	RACI table	Notion doc
18	Dependency map	Calendar	Slide 8	Critical path	ClickUp Gantt
19	Scope creep firewall	Calendar	Internal	Out-of-scope quotes	Notion doc
20	Reporting dashboard architect	Measurement	Appendix	Tool & chart spec	Looker Studio
21	Monthly report generator	Measurement	Appendix	7-section report	Notion doc
22	Anomaly detector	Measurement	Internal	Weekly Slack alert	Slack + Airtable
23	Attribution model explainer	Measurement	Appendix	Plain-English model	Notion doc
24	Risk register	Risk	Slide 10	Risk + mitigation table	Notion doc
25	Trade-off generator	Risk	Slide 10	5 trade-offs	Strategy memo
26	Kill criteria	Risk	Slide 10	Per-tactic kill rule	Airtable base
27	Quarterly strategy review	Risk	Recurring	QBR template	Pitch / Google Slides

The pattern: prompts 1–5 feed the “here’s what we found” story. Prompts 6–9 turn that into a goal. Prompts 10–14 pick the channels. Prompts 15–19 schedule the work. Prompts 20–23 measure it. Prompts 24–27 protect it.

A 5-day “strategy deck rebuild” workflow

This is the calendar I run every Monday for a new client engagement. It assumes one strategist, one designer, and one project manager.

Day 1 - Audit. Run Prompts 1–5 against the kickoff call, the GA4 export, the brand samples, and the competitor list. End the day with a Notion doc titled “[Client] - Situation Snapshot v0.1.” Schedule a 30-min internal review that night.

Day 2 - Goals and channels. Run Prompts 6–14 against the snapshot. End the day with a Notion doc titled “[Client] - Strategy Outline v0.1.” The outline should have one bullet per slide in the deck. Schedule a 60-min internal review with the agency lead.

Day 3 - Calendar and budget. Run Prompts 15–19. End the day with an Airtable base titled “[Client] - Campaign Tracker v0.1” and a ClickUp project titled “[Client] - 90-Day Plan v0.1.” Run Prompt 20 to spec the reporting dashboard. End the day with the dashboard URL.

Day 4 - Measurement and risk. Run Prompts 21–27. End the day with the QBR template, the risk register, the kill criteria, and the monthly report template. Designer begins the deck. End the day with a v0.1 deck in Google Slides or Pitch.

Day 5 - Review and ship. Internal review at 10am. Client review at 2pm. Send the deck as a PDF with a Loom video (3 minutes max). Follow up within 24 hours with the budget allocator, the resourcing matrix, and the next 3 asks.

The 5-day cadence works because each day’s output is the input to the next. If you skip a day, the strategy stops being coherent. If you take 10 days, the client has already lost interest.

Common mistakes to avoid

Mistake 1: Starting with the deck. The deck is the output, not the input. Start with the audit. If the audit is bad, the deck is a costume.

Mistake 2: Confusing the buyer and the user. The buyer is the executive who signs the check. The user is the marketing manager who actually runs the work. The deck needs to satisfy both. The buyer wants the “what” and the “why.” The user wants the “how” and the “when.”

Mistake 3: Using 30+ slides. Past slide 12, every additional slide is a tax on the reader. Cut ruthlessly. If the strategy can’t fit in 10 slides, the strategy isn’t clear.

Mistake 4: Forgetting the existing brand voice. A strategy deck that doesn’t sound like the client is a strategy deck that won’t be adopted. Run Prompt 4 early. Use the brand voice on every slide.

Mistake 5: Pretending AI isn’t in the room. The client knows you’re using ChatGPT. They’ll know if the deck reads like a raw model output. They won’t mind if the deck is sharper because you used the model. Use the model. Edit the model. Be honest about both.

Mistake 6: Skipping the risk register. Sophisticated buyers ask, “What could go wrong?” If you don’t have an answer, you lose the deal. If you do have an answer, you win it. Prompt 24 is the highest-ROI prompt in this guide.

Mistake 7: Not including a kill criterion. A tactic without a kill criterion is a tactic that runs forever, even when it’s failing. Name the kill criterion in the deck. Make the client agree to it.

Mistake 8: Treating the deck as a one-time artifact. The strategy deck is a living document. Update it quarterly. Re-run Prompts 22 and 27 every month. The deck is the source of truth, not the kickoff deliverable.

Final word - your move

The 27 prompts above will not write a perfect strategy deck for you. They will write the first draft of every section in roughly 90 minutes, and they will get you 80% of the way to a defensible plan. The remaining 20% is judgement: which channels to kill, which trade-offs to accept, which risks to name out loud. That part is still the agency.

I built this guide because the agencies winning in 2026 are the ones using AI to ship faster and think harder, not faster or harder. If you use these prompts and the deck still flops, the problem is rarely the prompt. It’s the strategy underneath. Run Prompt 14 (the differentiation pressure test) and Prompt 24 (the risk register) again. The answer is usually there.

Bookmark this page. Share it with your strategy team. Run Prompt 1 on your next kickoff call. Then run the rest in order. Five days later, you will have a deck that survives a board meeting - and the trust that comes with it.

Now go ship something.

Reader disclosure & educational-purpose notice

This page is published by SuperFreshAI for general informational and educational purposes only. By reading it, you agree to the points below.

Editorial independence. All reviews, guides, and recommendations are written by our editorial team based on hands-on use. Some links on this site are affiliate links, and some articles are produced as partner content — both are always clearly labeled. Our editorial conclusions are never shaped by partners or affiliates.
Not professional advice. Nothing on this page constitutes legal, financial, medical, tax, or other professional advice. AI tools, pricing, and capabilities change quickly — always verify current information with the tool's official documentation before making a decision.
Educational purpose only. The content here is intended to help you learn about AI tools and workflows. It is not a guarantee of results, performance, fitness for a particular purpose, or suitability for your specific situation. Your results may vary.
No warranties. The site and its content are provided on an "as is" and "as available" basis. We make no warranties, express or implied, about accuracy, completeness, reliability, or availability. See our Terms and Privacy for the full legal terms.
Your responsibility. You are responsible for how you use the information on this page, including any decisions you make based on it. Always do your own research and consult a qualified professional when appropriate.
Affiliate & partner disclosure. When you click certain outbound links, we may earn a commission at no extra cost to you. When a piece of content is produced as partner content, it is labeled at the top of the page. See our Editorial Policy for the full standards we follow.

By continuing to read, you acknowledge that you have read and understood this notice.

27 ChatGPT prompts for digital marketing agencies to generate client strategy decks

Why clients ghost your strategy decks in 2026

The 10-slide strategy-deck blueprint for 2026

Section 1 - Audit & situation prompts (Prompts 1–5)

Prompt 1 - Kickoff call distiller

Prompt 2 - Google Analytics 4 narrative generator

Prompt 3 - Competitor teardown

Prompt 4 - Brand voice & tone extractor

Prompt 5 - Quick-win detector

Section 2 - Goals & KPIs prompts (Prompts 6–9)

Prompt 6 - Goal ladder builder

Prompt 7 - KPI dictionary

Prompt 8 - North Star Metric challenger

Prompt 9 - 90-day success scorecard

Section 3 - Channel & tactic prompts (Prompts 10–14)

Prompt 10 - Channel mix matrix

Prompt 11 - Tactic expansion

Prompt 12 - Content pillar generator

Prompt 13 - Funnel stage mapper

Prompt 14 - Differentiation pressure test

Section 4 - Calendar & budget prompts (Prompts 15–19)

Prompt 15 - 90-day rollout calendar

Prompt 16 - Monthly budget allocator

Prompt 17 - Resourcing matrix

Prompt 18 - Dependency map

Prompt 19 - Scope creep firewall

Section 5 - Measurement & reporting prompts (Prompts 20–23)

Prompt 20 - Reporting dashboard architect

Prompt 21 - Monthly report generator

Prompt 22 - Anomaly detector

Prompt 23 - Attribution model explainer

Section 6 - Risk & trade-off prompts (Prompts 24–27)

Prompt 24 - Risk register

Prompt 25 - Trade-off generator

Prompt 26 - Kill criteria

Prompt 27 - Quarterly strategy review template

Comparison table - prompt category vs. deck slide vs. output

People Also Ask - agency strategy deck FAQ (2026)

What is a strategy deck for a digital marketing agency?

How long should an agency strategy deck be?

How long does it take to build a strategy deck?

What tools should an agency use to build strategy decks in 2026?

What is the biggest mistake agencies make in strategy decks?

How do you price a strategy deck for a prospect?

How do you measure the success of a strategy deck?

Can ChatGPT actually write a good strategy deck?

What is the best ChatGPT model for strategy deck work in 2026?

How do you keep an agency strategy deck from looking like AI-generated?

A 5-day “strategy deck rebuild” workflow

Common mistakes to avoid

Final word - your move

SuperFresh AI

43 ChatGPT prompts for non-native English speakers to polish interview answers

41 ChatGPT prompts for SaaS founders in San Francisco to map local partnership opportunities

How to Detect AI-Generated Content

What Is the Best AI Tool for Writing?

AI Newsletter Writing Guide

Get practical AI insights in your inbox