Big brands are running AI-made video ads. 5x more videos, 10x faster.
But who analyzes all of them and figures out why one worked? Not just performance, frame by frame.
We built the other AI for the first problem AI created.
Out of beta today. Try it free!
@suno and @ElevenLabs look like direct competitors. Their video-ad playbooks aren't even close.
5 public Meta ads each, watched frame by frame by https://t.co/UT6uvLI6X3: 79 opportunities for Suno, 64 for ElevenLabs.
https://t.co/F59svr9XBX
https://t.co/HMv8KqZLga
@suno and @ElevenLabs look like direct competitors. Their video-ad playbooks aren't even close.
5 public Meta ads each, watched frame by frame by https://t.co/UT6uvLI6X3: 79 opportunities for Suno, 64 for ElevenLabs.
https://t.co/F59svr9XBX
https://t.co/HMv8KqZLga
Next time you're weighing build vs. buy, ask yourself: 1️⃣ Is this where our business moat is? 2️⃣ Does it add to our IP? 3️⃣ Or are we just shifting cost from marketing to engineering?
If it's not a moat: buy, and get to ROI faster.
(2/2) Because building is the easy part. Keeping something accurate and maintained in production is where the real cost shows up. And marketing is rarely first in line for engineering time.
(1/2) At a CMO panel last week, in a room of 150+ CMOs, I asked: when it comes to AI for marketing, BUILD or BUY?
👉 The answer: For specialized solutions: BUY.
How are CMOs actually using AI today?
Last week I listened to a CMO panel in Palo Alto, with leaders from JP Morgan, Everpure, and Wrike, and their top use-cases:
1️⃣ Data and insights before content creation.
2️⃣ AI as chief of staff.
3️⃣ AI for personas.
Just watched AI-generated UGC video ads outperform human-created ones for the first time. 🤯
One agent watched the videos and sent instructions to another agent for editing. The self-optimization loop.
Super excited to push the boundaries for video ads. https://t.co/yO0zlSCkUE
960 actions suggested by AI in 1 day.🤯That's what Connecty AI delivered to a client yesterday — WITHOUT A SINGLE PROMPT.
GOOD: AI has surpassed human capability.
SCARY: who's going to implement 960 actions?
LESSON: AI can surpass humans, but can't bypass them!
Stop Building Dashboards No One Uses!
Most companies have 100+ dashboards. Most get checked once, then forgotten.
The problem? Data teams start with “what should we visualize?” instead of “what business GOALS are we supporting?”
Real value happens when you work backward from business goals - not visualization requests.
So before you open your BI tool, ask your business team
1️⃣ What are the top 3 GOALS per department this quarter?
2️⃣ Which north star metrics define success?
3️⃣ What actions do you take based on those numbers?
And then either curate a list of metrics, measures, filters, dimensions that meets those goals and empower data-savvy business teams with self-serve dashboards based on it.
🚀 Or simply generate a fully automated AI-built and AI-governed 'Day Zero Semantic Layer' for FREE today.
🧭 The First 90 Days Playbook for Data Leaders
🗺️ Phase 1 - GOALS BEFORE DASHBOARDS
- Don’t start with dashboards consolidation.
- Sit with each team and list the 2–3 decisions they make weekly (e.g., optimize spend, ship a feature, reduce churn).
- Turn those into goal statements with clear thresholds.
- Even if metrics grow later, you’ll know which numbers must be reliable first.
💡 Pro tip: Use Connecty's DayZero Semantic Layer to pre-populate a comprehensive list of your personalized business goals by each function. https://t.co/wA8s6cT0XW
🗺️ Phase 2 - AI AUTOMATED GOVERNANCE
- Pull together every existing definition - from wikis, dbt models, Google Sheets, SQLs, DAX.
- Highlight conflicts.
- Then host short workshops to align on Revenue, Active User, CAC, Churn, LTV. (Yes, it’s painful but essential.)
💡 Pro tip: Tools like Connecty’s DayZero Semantic Layer can auto-learn base semantics from query history + metadata to make alignment in minutes than weeks. https://t.co/wA8s6cT0XW
🗺️ Phase 3 - DRIFT IS NOT A BUG
- Definitions evolve. Version changes are bound to happen - handle them systematically.
- Create a workflow for definition updates, auto-versioning, and safe downstream rollout.
- Otherwise, business teams will create hidden “Google Sheet layers” of their own truth.
💡 Pro tip: Use Connecty’s Autonomous Semantic Graph for AI powered auto-updated definitions in a trusted living semantic graph, and detect or approve any dependencies downstream. https://t.co/6JeZIZIW2D
By Day 90, you’re NOT judged on architecture. You’re judged on TRUSTED decisions.
🚀Announcement! A few months ago, we blogged about our vision 'The Future of Metrics':
→ Auto-generated Metrics Layer
→ Semantics that evolve with your data
→ Meaning both humans and AI can trust
🚀 Today, we're proud to ship that vision to all our beta clients → 'Day Zero Semantic Layer'.
Day Zero Semantic Layer auto-learns your business logic from your query history, schema metadata and enriched catalog - within minutes!
No YAML. No mapping. No waiting.
✅ Query-ready semantics on Day 0
✅ Verified personalized definitions for humans & AI
✅ Built-in governance and explainability
✅ Works with any data source
Try the Future of Metrics today. https://t.co/kQ7sfrm4vJ
Without an autonomous semantic layer, your data team keeps redefining the same metrics for different teams. The result? Multiple versions of the truth - and everyone ends up chasing different goals.
Choosing the right AI Agent for Data Analytics?
Use this 15-Point Checklist ✅
https://t.co/JIZnIWROSf
Don’t settle for surface-level AI - demand true autonomy.
#AI#DataAnalytics