
Dan Moen
342 posts

Dan Moen retuiteado

✅Building a Marketing Dashboard That Drives Decisions
@danmoen
#MarketingTips
#MarketingStrategy pikeandvine.com/building-a-mar…
English

@MoeCanDoIt Use AI to help you develop brand guidelines and custom CSS classes it can use when it's building. Start every thread by making it read through both and then summarize its understanding. This helps with UI consistency, but not UX. That's on you.
English
Dan Moen retuiteado

5 Interesting Facts on Treasuries from Torsten Slok at Apollo:
1. For every $5 the government collects in taxes, a dollar goes to paying interest on debt.
2. $10 trillion of government debt will mature over the next year, which is 33% of all debt outstanding.
3. The share of T-bills outstanding has increased to 22%, and 85% of Treasury gross issuance is T-bills.
4. The average federal net interest expense per day, including weekends, is now $3.5 billion.
5. Foreign ownership of Treasuries has declined to 25% of the total outstanding, down from 33% a decade ago.

English
Dan Moen retuiteado

@anu_sthree This moment will pass *very* quickly, and a quilt is a brilliant idea for a useful memento.
English
Dan Moen retuiteado

@JustenEcom Ah, makes sense. So it gave you creative direction and you executed the design elsewhere. Looks amazing, nice work!
English

I tried to upgrade my banner with AI… and got a reality check.
For fun, I gave ChatGPT and Claude the exact same input.
The results were… not even close:
ChatGPT: total disaster 🤝
Claude: actual “optimization” (like you’d expect)
After a bit of fine-tuning, the Claude version is now my banner.

English


@BridgetMWillard As a Seahawks fan and huge Pete Carroll fan I hope he either goes to a team ready to compete for a championship or retires.
English

Are you overwhelmed by new content while your top-performing assets gather dust?
Stop producing more. Strategically promote your proven work. Audiences value consistency and quality over volume.
Reuse. Reclaim. Reengage. 🚀
#ContentMarketing #Strategy
English

@zara_ferna94287 Any examples you think are awesome? Sephora Beauty Chat is the best of the big brand offerings I've seen so far.
English

@zara_ferna94287 My goal this week was just to try all of the best examples out there. More a learning exercise than seeking a specific solution. Since most of them suck, I built one to see why it's hard. Turns out it's not actually hard.
English

@danmoen The Oura ring chat bot was able to detect an issue with my ring and initiate a shipment of a replacement in like 2 minutes 😅
English

@saadullah45 All of that makes sense!
I would probably add a 3rd call up front at 100 articles:
1. Pick up to n categories (nano)
2. Pick up to 3 articles from chosen categories (mini)
3. Write the answer (premium)
English

Smart setup! If your Hugo content fits the context window, skipping a vector DB is a great way to keep latency low.
Where RAG would level this up:
1. Semantic Matching: The 'File Picker' relies on keywords. RAG understands meaning, so if a user asks about 'rates' but your file is pricing.md, it won’t miss the connection.
2. Scalability: As you hit 100+ articles, sending a huge list of filenames to the first LLM becomes slow/expensive. A vector DB handles that search instantly.
English








