A mid-size e-commerce marketing manager cut her weekly AI-related work time from 8–10 hours to under 2 by using structured ChatGPT prompts for campaign briefs and a custom brand-voice system prompt for weekly reports. Results achieved within 30 days using the AI Survival Kit for Marketing Managers.

Marketing Manager Published

From 10 Hours to 2:
How a Marketing Manager
Reclaimed Her Week With AI

Sarah K. was falling behind on campaign briefs and weekly reports — a slow creep that was starting to show up in her performance reviews. She fixed it in a weekend.

75%
reduction in brief writing time
30 days
to see measurable results
1 weekend
to read the core guide

Disclaimer: This is a composite account based on early user experiences. Names and identifying details are fictional. Individual results vary.

The Problem

She was spending 10 hours a week on tasks AI could do in 25 minutes

Sarah K. is a marketing manager at a mid-size e-commerce brand in Sydney. She manages a team of two and owns everything from campaign briefs to paid social reporting. On paper, her role is strategic. In practice, she was spending most of her week writing.

Campaign briefs took two hours each — sometimes more, if the brief needed to cover multiple channels. Weekly performance report narratives (the bit that explains the numbers) took 90 minutes every Friday. And because she also owned the monthly content calendar, that added another two hours every few weeks.

She knew her colleagues were using AI. In meetings, they'd mention "running it through ChatGPT" with a casualness she didn't share. She'd tried ChatGPT once, typed in a vague prompt, got back something generic, and closed the tab.

Then came the performance review. Her manager didn't flag her output — she was producing good work. The comment was about speed and about "leveraging new tools the business has invested in." It was diplomatic, but the message was clear: other people in comparable roles were moving faster.

Sarah didn't need to become a prompt engineer. She needed to stop starting from a blank page.

What She Did

One weekend. One guide. A system that runs itself.

Sarah bought the AI Survival Kit for Marketing Managers on a Friday evening. She read the core guide — 40 pages — over Saturday morning. She didn't take notes. She just read it.

Saturday afternoon she worked through the campaign brief module. The kit includes a structured prompt template: a five-field input (product, audience, channel, objective, tone) that gives the AI enough context to produce output you can actually use. She tested it on a brief she'd already written the previous week. The AI version needed editing. It didn't need rewriting.

Sunday she built a system prompt for her brand voice — a persistent instruction set she pastes at the start of every ChatGPT session. It includes her brand's tone descriptors, three example sentences in the right register, and a list of terms the brand avoids. Once it's in, every output sounds like the brand, not like ChatGPT.

By the following Wednesday she had used these two tools on three live briefs and two report sections. She hadn't told her manager yet. She wanted to be sure first.

3 Prompts She Used

Prompt 1 — Campaign Brief

You are a senior marketing strategist. Write a campaign brief for the following: Product: [DTC skincare brand, hero product is a vitamin C serum, $68] Audience: Women 28–42, health-conscious, mid-market income, based in Australia Channel: Meta (Facebook + Instagram), email to existing customer list Objective: Drive 200 units sold in 14 days; secondary goal is growing the email list Tone: Warm, credible, not clinical. Think Goop without the woo-woo. Format the brief with: Overview, Audience Insight, Core Message, Channel Breakdown, and Success Metrics. Keep each section to 3–5 bullet points.

Why it works: The five-field structure gives the AI the minimum context it needs to make real decisions — not generic ones. The tone instruction ("Goop without the woo-woo") is specific enough to produce differentiated copy. Without the tone instruction, you get corporate beige.

Prompt 2 — Weekly Report Narrative

Write the performance narrative section for a weekly marketing report. Use this data summary: - Meta ads: CPM down 12% vs last week, CTR up 0.4%, ROAS 2.8 (target 3.0) - Email: Open rate 34% (above industry avg), click rate 3.2%, 12 unsubscribes - Revenue: $42,000 for the week, 8% below target of $45,500 - Key driver of miss: Hero product out of stock for 3 days mid-week Write 3 short paragraphs: (1) Overall performance summary, (2) What worked and why, (3) Key action for next week. Tone: direct, data-led, no jargon. Written for a non-marketing CEO.
Note: No actual revenue figures or client data go into ChatGPT. Sarah used representative numbers for this prompt. For anything commercially sensitive, describe the trend rather than the exact figure: "Revenue was below target by approximately 8%, driven by a stock outage mid-week."

Why it works: Giving the AI structured data points (not a spreadsheet) and defining the audience ("non-marketing CEO") shapes both the content and the register. The output needs light editing, not a rewrite.

Prompt 3 — Brand Voice System Prompt

You are a senior marketing writer for [Brand Name], an Australian DTC skincare brand. Tone: Warm, confident, and credible. We educate without lecturing. We're human, not clinical. Voice notes: - Write in second person ("you") when addressing the customer - Short sentences. Never more than two clauses. - We never use: "innovative", "game-changing", "revolutionary", "cutting-edge", "seamless", "holistic" - We do use: "real results", "made for your skin", "backed by dermatologists" Audience: Women 28–42, health-conscious, time-poor. They're smart and they can smell a sales pitch. Be direct. Keep this instruction active for the entire conversation. Every piece of copy you write should match this brand voice.

Why it works: This goes in first, before any actual task. It anchors every subsequent output to the brand. The "words we avoid" list is particularly powerful — it's the fastest way to stop ChatGPT defaulting to buzzword-heavy corporate copy.

What Changed

Within 30 days of reading the guide

Campaign briefs: 2 hours each

25 minutes, including review

Weekly report narrative: 90 minutes

15 minutes to draft, 10 to edit

Starting from a blank page every time

Brand system prompt ready in ChatGPT

Performance review flag on tool adoption

Training her junior team on AI workflows

The biggest unexpected change: Sarah started training her junior team member on the same workflows. She built a shared Notion page with the brand system prompt and three brief templates. New team members now hit the ground running. That wasn't something she'd planned — it just happened because the system was easy enough to hand off.

She updated her LinkedIn to include "AI workflow integration" in her skills section. Her next quarterly review mentioned it directly — this time as a positive.

"I was embarrassed to admit I hadn't started using AI yet. Now I'm the person my whole team asks."

— Sarah K., Marketing Manager, e-commerce brand, Sydney

Common Questions

AI for marketing managers

Can AI actually write a usable campaign brief? +

Yes — with the right prompt structure. Generic prompts produce generic output. The key is giving the AI a structured brief that includes the product, target audience, channel, objective, and tone. A well-structured prompt produces a campaign brief that needs editing, not rewriting. The Marketing Managers kit includes a tested template for exactly this.

How do I stop ChatGPT from producing generic marketing copy? +

The fix is a system prompt that encodes your brand voice before you ask for any output. You tell the AI your brand's tone, the words you avoid, example phrases you like, and your audience persona. Once that's in place, every output reflects your brand — not ChatGPT's default corporate register.

Is it safe to put client campaign information into ChatGPT? +

OpenAI's default settings use conversation data to train models unless you opt out. You can disable this in Settings → Data Controls. For anything commercially sensitive, use Claude (which has stronger enterprise privacy defaults) or disable training data usage. The AI Survival Kit covers this in the data safety module.

Get the Same System

AI Survival Kit for Marketing Managers

30 copy-paste prompts. A tested campaign brief template. A brand voice system prompt builder. Everything in the story above — and more.

See the Marketing Managers Kit →

From $47 · Instant PDF · 30-day money-back guarantee