A 142-page prompt, 18 emails, and one solopreneur's first Black Friday campaign

Today, Covarrubias has earned AI certifications and leverages the technology for her business, including creating a 100+ page super-prompt and custom GPTs to help organize her thoughts, plan, write, and run seamless autonomous marketing campaigns with ActiveCampaign. 

The result? 

More personalized suggestions, workflows, and increased motivation for Covarrubias to keep her email campaigns running smoothly.

What goes into a 142-page brand document for AI

What this all looks like in practice for Covarrubias is a ​“super prompt” she built with her business coach, Amber, to help her create content for marketing campaigns for her business.

Here’s what the document entails, designed to help AI tools like Claude and other custom GPTs deeply know and understand Covarrubias’ brand background, capabilities, voice, and style. In some cases, the doc is knowledge for a Custom GPT. In others, it informs Claude skills. 

Her superprompt includes:

• Methodologies and frameworks used in her coaching practice.
• A detailed breakdown of Covarrubias’ ideal customer: first-generation women and femmes (i.e., first to go to college, first to start a business, daughters of immigrants).
• Her constraints, including marketing tactics she’d never use.
• History of Covarrubias’ podcast and business background—plus transcripts from her show.
• Writing samples, including emails that resonated with her audience.
• And more—this is a living document that Covarrubias updates and adds to on a regular basis.

Let’s dive into what it actually looks like:

“It’s a lot of context,” Covarrubias explains. ​“It’s still not perfect, but [the output] gets me about 75–80% there when I’m writing.”

Covarrubias said as someone with a neurodiverse brain—she has ADHD—not starting with a completely blank page has been incredibly helpful to actually move projects along, including marketing campaigns.

“Once I built this document, it helped with my ever-changing ADHD mind. The biggest thing the doc does is it gives me guardrails to my wild mind so I don’t make things up as I go,” she says.

“I go to my cowriting tool GPT and I say, ​‘OK, I have this campaign I’m working on. Let’s write the first email together,’” she shares. ​“And it gives me a starting point and I change things, remove words, add storytelling—I get it to 100%.”

How a 20-minute prompt saves hours on the backend 

As Covarrubias’ super prompt demonstrates, AI is only as powerful as the information you give it to work with.

While AI can definitely serve as a time-saver, Covarrubias believes that most people aren’t laying enough groundwork before expecting business-changing results.

Instead, because of the impressive amount of groundwork Covarrubias has laid, she creates a solid foundation. She says she sometimes spends 20 minutes or more to create a super detailed prompt, but the output is so much better than if she had shared less information, saving time and energy in the long run.

“The one thing I often say is ​‘the result of your delegation is as good as the quality of your instructions,’” she shares. ​“I keep that in mind when I’m working and planning campaigns and doing all the things to make sure that I’m investing enough in the front to make sure that I don’t have to spend a lot of time in the back.”

Case study: Using Claude for Covarrubias’ Black Friday, Cyber Monday campaign

From super prompt to campaign timeline

In the fall of 2025, Covarrubias decided to run her first-ever Black Friday, Cyber Monday marketing campaign, tied in with her birthday and the anniversary of her podcast.

She was skeptical because there’s so much noise in people’s inboxes during the sales-heavy week, but she still wanted to experiment and turned to AI, specifically, Claude, which she believes serves more as a philosopher and is better for thinking and evaluating than other tools, with her idea.

“I re-shared my super prompt (the 100+ page brand background document) and explained I wanted to do a campaign for a specific version of my client,” Covarrubias says. She wanted to do a Black Friday / Cyber Monday sale for her course ​“Unmute Yourself,” positioned as a self-improvement challenge.

Here’s the full language of the prompt she shared to help get her campaign planning started:

“Because my prompt includes all my tech, AI knows that I work with ActiveCampaign, that my checkout cart is ThriveCart, that my course tool is MemberVault,” she says. ​“It already knows. I don’t have to explain it.”

Claude even offered up a list of tags to add to ActiveCampaign in the process, which Covarrubias says saved a ton of time brainstorming and figuring out workflows. Covarrubias noted while Claude’s recommendation was a great start, she still needed to make tweaks, a reminder that AI is just a tool and can’t fully replace your human brain.

Pam and her AI thought partner then refined the campaign together. Once they had the timeline and the foundation, it was writing time.

One week, 18 emails, and a mid-campaign pep talk

Covarrubias and Claude spent about a week writing a total of 18 emails, designed to be sent over the course of three weeks, and sent at exactly the best time because she relies on Predictive Send, a powerful feature that maps the optimal send time for each recipient.

By week two, Covarrubias felt burnt out and told Claude, ​“Hey, I don’t think I want to continue the campaign. Let’s end it here.”

Claude jumped into partner mode, and encouraged Covarrubias to keep going.

The tool motivated her by saying: ​“You have one week left. You’re more than halfway through. Let’s just push through it and remember that one of the things that is important in your super prompt is to have data to make decisions later. If you don’t complete this project, we’re not going to have enough data. And you would never tell your clients to leave something undone, would you, Pam?”

This little bit of tough love was what Covarrubias needed to hear, and she pushed through to complete her campaign.

The insight she might have missed without AI

The campaign delivered wins. It also had numbers worth taking a hard look at.

When the campaign was complete, Covarrubias went into ActiveCampaign to grab the stats from her efforts and fed them into Claude to help evaluate what went well and what could be improved upon for future campaigns. 

Covarrubias asked Claude: ​“Analyze which specific campaigns had higher click rates to replicate successful elements.”

Claude provided Covarrubias with a helpful analysis that allowed her to look at the data in a fresh, new way she hadn’t considered before. She had been super focused on the number of unsubscribes, but Claude provided information that helped her to better understand the data from her emails.