Not My Job
Posts
How to NOT get rejected by the growing AI;DR (AI didn’t read) crowd

How to NOT get rejected by the growing AI;DR (AI didn’t read) crowd

How to use AI for content writing without turning into a sleazy bot

Midhat Hadi & Aelia Haider
March 19, 2026

If you’re semi-regular on LinkedIn, you must’ve seen the plethora of posts about identifying AI content (so people can avoid it) and how AI slop is bringing SEO down to the ground.

The first is a hit-or-miss. The second is true.

And the sentiment is pretty valid.

People are wary of AI content, and for good reason—if you can’t be bothered to write something on your own, how can you expect someone else to read it?

For founders, it’s a conundrum: writing is not their main job, so getting AI to help them is an easy + cheap fix. But the cost is losing the AI;DR audience.

🔥 The INM trick: Never ask AI to write for you. Use it to mine the nuggets of insights buried in your research, and to polish your drafts.

🔴 Safety Note and Disclaimer: Always, always, always use anonymized data. Remove and redact all personal details before you give it to AI.

All mentions of data in this issue assumes anonymized/redacted versions of it.

Remember your content assembly line?

Psst: Here’s the link to the previous issue if you missed it or want a refresher.

Let’s break it down in two phases where you can use AI:

Important Note: The “writing” block under phase 2 does NOT refer to using AI to write, but using it as structural assistance—we’ll get to the details under the instructions!

Phase 1: Identifying Voice of Customer (VoC) and Populating Messaging Matrix

Before you create content for your customers, you’d need to find out what the ongoing conversation is even about. And for that, you’ll need to map a mini-Reddit on to your messaging matrix.

Why Reddit? It’s currently the best channel to monitor real consumer conversation. And as Redditors are aggressively anti-sales-y posts, they are very quick to weed out anyone trying to advertise.

Reddit (and even Quora) act as niche marketplaces for pain—where users aren't being polite.

This creates an environment that enforces valuable contributions to conversations: real complaints and problems—and often pain points behind problems that users may not admit in formal surveys.

There are 3 steps to getting this done:

Let’s begin!

Step 1: Prep and Prime Your AI

Choose your poison: Claude, Gemini, DeepSeek, whatever you’re comfortable with.

Your poison. The poison for you. The poison chosen specifically for you. By you.

Once you’re done, it’s time to feed the beast.

Because before you can ask AI to do anything at all, you need to provide it with the “brain matter” to analyze info for you.

So here’s what you do:

Create a single document containing every scrap of raw (anonymized and redacted) data you have:

onboarding forms,
interview transcripts,
Slack screenshots of client wins, and
even those "venting" emails from customers

💡 Why all this data? Because you need to train your AI to understand your VoC so it doesn’t have to guess how your audience speaks.

Now don’t just go ahead and drop your file to the AI. You need to give it some context.

Upload your file and use this exact prompt to prime the AI’s memory.

The Prompt:

I am uploading a document that contains raw customer research, transcripts, and product details. Do not summarize this yet. First, I want you to read through everything and acknowledge that you have processed the raw Voice of Customer (VoC) data. Confirm you are ready to help me map this to a specific messaging framework.

Once it confirms, then you’re ready to start.

Step 2: Build Your Reddit Mining Map

Now that you’ve processed your own data, you need to see what the rest of the market is saying in the wild.

You don’t have to guess; you can ask AI to generate the exact phrases that lead to high-intent vent threads on Reddit.

And here are the prompt templates to do it step-by-step:

Prompt #1: To Generate Your Industry-Specific Mining Map

You want to find the breaking points, the pricing frustrations, and the DIY failures.

The Prompt:

Act as a market researcher. I am looking for founders who are frustrated with [Problem]. Based on the VoC data we just analyzed, give me a list of 20 high-intent search strings for Reddit and niche forums. Break them into these four categories: 1. Symptom-Based (The chaos moments where they realize they need a system). 2. Competitive Displacement (People looking to fire big incumbents because of cost or fees). 3. Lightweight Failures (People struggling with DIY solutions like sheets or paper). 4. Emotional Triggers (The "burnout" or "vibe" killers).

Prompt #2: To Do a Competitor Intelligence Sweep

This is how you find the gaps in the big players.

Use the AI to generate a comparison map so you can see exactly where your competitors are dropping the ball.

The Prompt:

I need to find where people are complaining about [Competitor Name]. Generate 10 specific search queries for Reddit that will find:

Direct alternatives for [Competitor Name].
Comparisons between [Competitor Name] and [Other Competitor].
Threads asking if [Competitor Name] is worth the cost.
Specific "Am I the only one who hates..." threads regarding [Common Industry Practice].

Now that you have all the beautifully targeted search queries, copy + paste them into a search engine followed by site:reddit.com.

Note: You can do the same for Quora. It’s not as authentic as Reddit (there are a lot of advertising posts and comments), but you can still find some good stuff.

Step 3: Populate Your Messaging Matrix (Table)

Take the data you get from the Reddit mining step, and feed it to your AI along with other relevant info (transcripts, internal notes, etc.).

And give it a specific sorting task for your Messaging Table with this prompt:

The Prompt:

I am providing a blank Messaging Table and raw research data. I want you to act as a Research Clerk and fill the table by listening and extract every specific instance:

Where a user mentions a 'breaking point' or 'time-waste' and put it into Pain Points.
From Comparison threads, find where they realized the old way was broken and put it into Pivot Point.
From Testimonials, fill out the Product/ Feature Description.
From Testimonials, identify the Endgame and Emotional Payoff and put them into their respective cells.
From Sales transcripts, pull out and fill Biggest Objections and Objection Handling.
From Forums/Reddit, find threads where people complain about [Competitor’s] gatekeeping or high-friction sales process and put them into Objection Handling.

Constraint: Use their exact, quirky quotes. If you can't find a row, leave it blank. Do not use ‘marketing speak’.” Use their exact words—do not summarize. For every entry, include the 'Who' (their role).alternatives for [Competitor Name].

You can copy from the empty messaging table here. Also, replace “breaking point” and “time-waste” with any other filter words you want to use.

This will populate your table with customer data instantly, without you having to spend hours doing it on your own.

You get the quirky phrases directly from the audience that make relevant copy without the manual labor.

Now, if you want to take a step further and understand where your Messaging Matrix rows match with different levels of awareness, here’s a quick table guide for you.

Understanding awareness stages is vital when you’re creating a landing page that’s talking to different groups at the same time (like employers and employees).

💡 P.S. Spend 60 seconds verifying that the quote or link actually exists in your raw data dump before you hit publish, because AI tends to hallucinate at times.

Phase 2: Translating Strategy to Content and Polishing it Up

Your messaging table is ready. But your content draft is just a blank doc in front of you.

How do you translate one to the other?

You need to connect the two (and you also need to write what your customers are dying to hear!). Here are the steps for it:

Step 1: Generate Your Headline

Use a Mad Libs formula to turn a row from your matrix into a scroll-stopping headline.

The Formula: "If you’re [Target Audience] dealing with [Pain Point], here is how to get [Endgame] without [Biggest Objection]."

The Prompt:

Using my Messaging Table, generate 5 headlines for my target audience using this exact formula: 'How to [Endgame] without [Objection] even if you are currently [Pain Point].' Use the exact quirky words from [Unique words and phrases].

Note: You can replace everything highlighted in the prompt above with any of the headline writing formulas we discussed in the last issue.

Step 2: Structure Your Outline

Once you have your headline, don't ask the AI to "write a blog post." That’s how you get fluff.

Instead, tell it to build a structure using the specific rows of your Messaging Table. This ensures the content is structurally sound before you even start editing.

Use the AI to map your Matrix Rows to a standard Problem Agitation Solution framework.

The Prompt:

I have my headline: [Insert Generated Headline]. Now, I want you to outline a short-form post , based on the PAS (Problem, Agitation, Solution) framework, using these specific parts from my Messaging Table:

The Opening: Start with the Pivot Point to show you understand the moment they realized the old way was broken.
The Conflict: Agitate the Underlying Problems—the stuff they only admit in private.
The Solution: Describe our approach using the Product/Feature Descriptions but focus on the Emotional Payoff .
The Close: Use the Objection Handling to clear their final doubt and give them the Next Step .

Constraint: Use the Unique Words & Phrases throughout so it sounds like a human, not a bot.

✍️ When your outline is ready, start fleshing it out in your own voice. Remember to write like you speak (or how your brand speaks), and you’ll be fine.

Step 3: Conduct No-Fluff Structural Audits

Once you have a draft, do not ask the AI to “make it better” or “improve the flow.”

That is how you get generic corporate sludge.

Instead, run these three technical sweeps:

1. The "Bottom Line" Audit

This forces the AI to look at your draft through the eyes of a busy, cynical founder.

The Prompt:

Read this draft. For every paragraph, tell me 'So what?' from the perspective of a founder who has 5 minutes before their next meeting. If a sentence doesn't provide immediate value, highlight it for deletion.

2. The "Hard Brick" Audit

Vague writing is the mark of a bot. This sweep replaces “fluff” with “research-backed claims.”

The Prompt:

Identify every vague word like ‘many,’ ‘fast,’ or ‘efficient.’ Highlight these spots and tell me to insert a hard number, a specific stat, or a real-world example from my research file.”

3. The "Asset Depletion" Audit

Ensure you aren't leaving your best "proofs" on the cutting room floor.

First Prompt:

Look at Objection Handling in my table. Now look at my draft. Tell me where I missed an opportunity to insert a testimonial, a logo, or a case study result that would handle a reader's doubt.

Second Prompt:

Look at Product/Feature Descriptions. Now look at the draft. Tell me which specific section of our UI we should circle-back and bolt-on as a guided screenshot to prove this feature actually exists.

4. Big 3 Integrity Check

A longer headline isn’t always a bad thing!

The Prompt:

Read the headline.

Does it explicitly state:

The Action (What),
The Target Audience (Who), and
The Endgame (Why)?

If any of these are missing, rewrite it to be clearer, even if it gets longer.

Now read the content piece and do your own line-editing to polish the final draft.

And voila! You have your new content piece ready to go that’s not just rooted in real industry-based language but is also something targeted to the specific levels of awareness of your customers.

In other words, this is what they need from you in their own language! ✨

See you in your inbox next week, where we’re discussing repurposing the content you create!

Cheers!

Hey hey hey! 👋

Are you procrastinating taking another look at your content’s state? 😬

Or wish someone else could just tell you what’s wrong and what to do next? 🥺

Good news!

We’re offering a content audit for SaaS founders who’d like some help in this case.

Our audit deliverables include:

a detailed pdf report on the results of content gaps and recommendations, and
a 1:1 meeting for a walkthrough (so if you have any questions, ask away on the live call!)

How to NOT get rejected by the growing AI;DR (AI didn’t read) crowd

How to use AI for content writing without turning into a sleazy bot

Phase 1: Identifying Voice of Customer (VoC) and Populating Messaging Matrix

Step 1: Prep and Prime Your AI

Step 2: Build Your Reddit Mining Map

Prompt #1: To Generate Your Industry-Specific Mining Map

Prompt #2: To Do a Competitor Intelligence Sweep

Step 3: Populate Your Messaging Matrix (Table)

Phase 2: Translating Strategy to Content and Polishing it Up

Step 1: Generate Your Headline

Step 2: Structure Your Outline

Step 3: Conduct No-Fluff Structural Audits

1. The "Bottom Line" Audit

2. The "Hard Brick" Audit

3. The "Asset Depletion" Audit

4. Big 3 Integrity Check

Reply