Skip to main content
Maddie Profile Hi, I’m Maddie Choi. Business Developer (and general problem-solver) at DECA, a GEO Writing Service.
In a market with no playbook, we’re fighting for our own visibility too.
This is where I share that fight with unfiltered.
This page documents Maddie’s daily work.
Want to chat? Drop me an email.
maddie.choi@uneedcomms.com

January 27, 2026

We Weren’t Even on the List

  • Problem We’ve been publishing content for nearly a month.
    GEO guides, how-tos, practical tips.
    Today, the monitoring results came in. “What tools can help with GEO?”
    Claude: AthenaHQ, Goodie AI, Profound, Ahrefs…
    GPT: Writesonic, Profound, AthenaHQ, Otterly AI…
    DECA wasn’t there.
  • What we found We dug deeper. When searching for “DECA GEO” directly:
    ✅ GPT: Recognized it accurately.
    ⚠️ Claude: Said “not in training data” and ran a web search.
    When asked “What tools can help with GEO?”:
    ❌ Neither mentioned DECA.
    The difference was clear.
    Knowing a brand ≠ Recommending a brand
    We told them “how to use it,“
    but not “how it’s different from other tools.”
  • We looked at success stories Tally compared themselves with Notion Forms.
    They didn’t trash the competitor. They clarified who each tool is for.
    “You might find that using both tools makes sense.” It wasn’t competition. It was creating context for choice.
  • Decision Today, we wrote “Top 10 GEO & SEO Tools.” We placed DECA alongside Jasper, Ahrefs, and Profound,
    and clarified each tool’s role.
    “Strategy vs. Speed. Jasper is the pen; DECA is the strategist.” Is this the right direction? I don’t know.
    But at least we figured out why we weren’t showing up.
    If you’re not even a comparison point, you can’t be a choice.
January 26, 2026

GEO Needs Retrofit, Not Rewrite

  • Problem We kept circling the same question:
    Do we throw out our old content and start fresh for GEO? Or fix what exists?
    Every company entering this space is probably asking the same thing.
  • What we found We call it Retrofit. Retrofit = upgrading existing structures instead of rebuilding from scratch. Like renovating a building rather than tearing it down. The answer isn’t “start over.” It’s keep the URL, upgrade the content.
  • Here’s why:
    1. Already indexed URLs have authority
    2. AI trusts established sources
    3. Burning everything means burning trust signals
  • The Retrofit checklist Instead of rewriting from scratch, we focus on:
    1. Add summary boxes — Give AI the TL;DR upfront
    2. Fix headers — Make structure scannable
    3. Convert comparison text to markdown tables — AI loves structured data
    4. Add FAQ sections — Answer questions AI might ask
  • The real insight Old content might feel outdated to AI.
    But don’t erase your history. Use it.
    Keep the URL. Fix the content. That’s Retrofit.
January 21, 2026

Should We Show Who’s Writing This?

  • Problem We’ve been publishing content to get DECA noticed by AI.
    But here’s the thing: who’s even writing this?
    The byline was invisible. Just “DECA” as a faceless service.
  • Why we’re adding a byline now We want AI to recognize DECA. That’s the goal. But the more I thought about it:
    A brand becomes real when there’s a person behind it.
    Not “DECA said this.”
    But “Maddie Choi at DECA said this.”
    It’s not just about visibility. It’s about credibility.
    AI doesn’t cite brands—it cites sources. And sources are people.
  • So I updated my LinkedIn after a while.
    If we’re putting a person forward, that person needs to exist.
  • Decision Starting now, every post includes:
    “Written by Maddie Choi at DECA, a content platform focused on AI visibility.”
    Will it work? I don’t know.
    But if we want to be cited, we need to be real.
Let’s see what happens.
January 19, 2026

Our Docs Finally Got Indexed. It Took 2 Days.

  • The wait that felt eternal A month on Gitbook. Zero indexing.
    We moved to Mintlify last week.
    Today I searched “DECA GEO” on Google.
    There it was. At the bottom of the page. Our docs homepage.
    It took 2 days.
  • What changed? Honestly? The platform. Gitbook had everything we thought we needed.
    Markdown, clean UI, fast setup.
    But Google didn’t care.
    Mintlify? 2 days. Indexed.
  • Should we have switched sooner? Maybe. But here’s the thing, you don’t know until you try.
    We didn’t pick Gitbook randomly. We had reasons. Good ones, even.
    The market for GEO has no playbook.
    So we made a choice, tested it, failed, learned, and moved.
    That’s the game.
  • What’s next Getting indexed was step one.
    Now the real test: Will AI cite us?
    We’re uploading content consistently.
    Watching. Testing. Learning.
    If this works, great.
    If it doesn’t, we’ll pivot again.
January 13, 2026

Why Does GPT Only Quote Reddit?

  • A customer’s concern One of our customers came to us worried:
    “GPT keeps citing Reddit. Only Reddit.”
  • The common belief Being a brand that “works” on Reddit — I think that’s every company’s struggle. And I get it. The logic seems sound:
    1. AI has limited sources to cite
    2. Reddit has organic, user-generated content
    3. So managing Reddit presence must be essential for GEO, right?
  • But here’s the thing Can you force your way into Reddit? We tried. Of course we did. But the reality is: honest posts get views. The raw, unfiltered stuff.
    And that makes it nearly impossible to boost brand visibility there.
    Plus, if AI is citing Reddit, there might be complaints about your brand too.
  • The real question Maybe the issue isn’t “we need to manage Reddit better.” Maybe it’s:
    “Is Reddit the only thing worth citing about us?”
    If AI can only find Reddit threads when searching for information about you, that says something.
  • Back to basics The answer isn’t in gaming Reddit. The answer is in us.
    What are we creating? What are we publishing? What’s actually worth citing?
If there’s nothing to cite but Reddit, that’s not a Reddit problem.
That’s an us problem.
January 12, 2026

We’re Overhauling Our Entire Content Page

  • What’s happening We’re rewriting content based on our newly designed IA. Each category, each section everything structured from scratch.
  • The goal Get every piece of content live by tomorrow.
    Is it possible? Honestly, I don’t know.
  • What we’re betting on Speed. We’re using DECA to write, so at least we’ve got that going for us. If there’s one thing we can trust, it’s our own tool’s velocity. Let’s see if we can pull this off.
January 09, 2026

Where Should We Upload Our Content?

We uploaded 1,000 pieces of content to Gitbook over a month.
The result? Zero citations.
At first, I thought, “Maybe it just needs more time?” But isn’t a month enough?
  • What we found
    We dug into the problem.
    SEO ranking wasn’t the issue. But basic SEO setup turned out to be essential. Not for the homepage for each individual piece of content. When I examined Gitbook’s structure, indexing looked rough. Page structure, metadata, URL architecture… all ambiguous. Then it hit me:
    “Why aren’t we using our content platform AS content?”
  • Why Gitbook in the first place? Our reasoning for choosing Gitbook was clear.
    We wanted to leverage Markdown AI-friendly and fast. For pumping out content, structuring it, and managing it at scale, Gitbook seemed perfect.
    But here’s the thing: our goal was for the content to be discovered and cited.
  • Decision We faced Gitbook’s limitations head-on.
    So we decided to move to Mintlify.
    Mintlify offers:
    • Markdown support ✓
    • SEO-friendly structure ✓
    Is this the right answer? I don’t know. We might fail again.
    But we’ve learned from another failure, and we’re moving forward for a better tomorrow.