- Cookie3
- Posts
- we went through it. we're back
we went through it. we're back
everybody and their sister is using AI, here’s how you can use it with KOL intelligence

hey there,
We're back. Did you miss us? Because I missed you. Like genuinely, writing into the void while we were gone was not it.
But we're here now. It's been a minute, and it hasn't all been smooth sailing. We went through some rough patches. But we worked through it, we're still here, and honestly? We're coming back stronger than before.
Cookie Pro is coming. Very soon. We're not going to say too much yet, but just know we've been building something, and it's worth the wait. Stay close.
Now. On to why we're actually in your inbox today. In case you haven't noticed, everybody and their sister is using AI right now. Your colleagues are using it. Your competitors are using it. Someone's nan is probably prompting Claude as we speak. And if you're a crypto marketer, there's a very specific way you should be using it that most people are sleeping on, pairing it with KOL Intel to make your reporting and strategy actually mean something.
That's exactly what this edition is about. We're breaking down the exact workflow we use internally at Cookie, the prompts included, so you can steal it wholesale
Let's get into it.
First, why post counts will gaslight you
Tracking & Evaluating KOL Mentions
Every crypto marketer needs to track their mentions. Usually, it falls into two buckets:
Paid KOLs – are they hitting the KPIs you set?
Organic mentions – who's talking about you unprompted, what are they saying, and how is it moving your mindshare?
The challenge isn't finding posts. It's cutting through the noise and pairing volume with the right metrics.
When I pulled Cookie's organic KOL data for November vs October, posts were up across every tier. Looks incredible, right? Except impressions absolutely tanked. If I'd only looked at one metric, I'd have either thrown a party or had a breakdown. The real story was in smart engagement, and the real story was actually good.
KOL Score | Posts | Smart Engagement | Impressions |
≥ 3000 | +12.5% | +26.7% | −23.7% |
≥ 4000 | +21.5% | +35.0% | −40.0% |
≥ 6000 | +22.4% | +7.8% | −52.6% |
Reach dropped. But the audience that actually engaged? Higher quality, better aligned, and growing faster than posting volume in most tiers. That's not a bad story, that's a great story hidden inside a scary-looking impressions number.
You can only see this when you layer metrics together. One number alone will lie to your face every time.
The actual workflow (steal this)
Step 1: Search. Open KOL Intel, put in your keywords.
For Cookie we use: Cookie.fun, Cookie3, Cookie DAO, cookie (AND) campaign, $COOKIE. Set your timeframe, decide if you want replies included (reminder: threads count as replies on X), and run two separate searches, one for paid KOLs to check KPI delivery, one for organic to spot what's actually moving.

Step 2: Filter.
Minimum follower count. Minimum KOL Intel Score. Cut the coordinated small-account noise before it wastes your time. You want signal. Not 47 accounts with 200 followers all posting the same thing at 3am.

Step 3: Export and let AI do the heavy lifting.
Export the CSV, throw it into your AI tool of choice. Upload multiple months at once. One conversation, all the comparisons. You're welcome.

The prompts that actually get you somewhere
These are word-for-word what I use. Not vibes. Actual prompts.
"Find top 10 accounts driving the most total smart engagement and impressions for [brand] within [timeframe]."
→ These are your relationship-building targets. If someone's organically driving reach for you, they go on a watchlist immediately.

"Calculate total posts, smart engagement, and reach by KOL Intel Score brackets: 3000–4000, 4000–5000, 5000–6000, 6000+. Present in a table. Then calculate MoM change by the same brackets."
→ Gives you the leadership KPI (what % are top-tier KOLs driving?) AND a read on whether your messaging is actually working.
"Within those brackets, which group has the best smart engagement/post and impressions/post ratio?"
→ Sometimes your content-market fit lives in mid-tier KOL audiences, not the mega accounts. This is how you find out.
"In brackets 3000–4500, 4500–6000, and 6000+, find me the top 5 accounts in each by smart engagement and impressions."
→ Your power-KOL list. Have your SMM engage with them regularly. Open a direct comms line. Let them hear things before announcements go live. They stay warm, you get the lift.
The actual point
KOL Intel's value isn't just reading the dashboard. It's what you do when you layer metrics, filter properly, and interrogate the numbers instead of just reading them top to bottom.
You go from "we got mentioned 200 times this month" to "here's exactly who is moving our mindshare, here's why, and here's what we're doing about it next month." That's the difference between a tracking tool and a strategy engine.
One more thing worth knowing: Cookie AI is currently in early testing and it's basically this entire workflow inside one interface, no CSV exports, no tab-switching. Same playbook, fewer steps. Watch that space. We’ll be dropping it soon.
As always, stay cookish!