Senior Product Manager at @databoxhq | AI Product Strategy, Growth & PLG Expert Free AI Analytics Skills: https://teapawt.net/t.co/FLk8fawO6c
Joined February 2012
- Tweets260
- Following766
- Followers148
- Likes228
Ziga Potocnik retweeted
Most reporting problems aren't reporting problems.
They're 12 tools that don't agree with each other.
Ziga Potocnik retweeted
Every Monday I opened three tabs before coffee. GA4, HubSpot, a spreadsheet someone built two roles ago.
Not because I needed more data. Because I didn't trust any single view of it.
That's not a dashboard problem. That's a trust problem.
Ziga Potocnik retweeted
Databox is now an official MCP connector for Claude.
Claude is brilliant; it just didn't know your business until now.
Give it the context it was missing: hubs.li/Q04rD7zf0
Ziga Potocnik retweeted
I wanted to show you how I get the full overview of my search visibility using three Claude skills.
They help me pull SEO, GA4, and AEO together by reading through my live data through Databox MCP, and returning a structured report you can share the moment it lands.
I use the GA4 report for weekly operating rhythm, the SEO report for monthly organic review, and the AEO report for tracking AI engine visibility as it shifts.
Try it out and see what the reports uncover for you.
Free to download: hubs.li/Q04q-vRK0
Ziga Potocnik retweeted
You are forcing Claude to lie to you. 😬
I wanted to test if that was true. I asked Claude for last quarter's revenue number and fed it our Stripe billing report. It gave me a number, fast and confident.
It was wrong by 15%, according to Hubspot.
Does that mean Claude’s number was wrong? No.
They are both “revenue”. But the definition is a bit different.
I ran the same question again, this time with Claude connected to our data through Databox's MCP instead of a pasted spreadsheet.
The question remained the same, but the answer came back different, because Claude wasn’t guessing anymore.
That's the actual fix to making your analysis with AI relevant and trustworthy.
Curious how you add the semantic layer to your analysis…?
#AIanalysis #MCP #MarketingReporting
Ziga Potocnik retweeted
I built my first Monday marketing report by hand back in 2014.
12 years later, one Claude skill runs all of it and hands me a single report that is ready before our Monday marketing team meeting.
I listed the metrics our team focuses on in the image. It's basically the same GA4 data I was staring at back then, just automated now, and with an explanation and action items attached.
We've been building out a shared library of skills like this at Databox, so that more teams can automate and upgrade their reporting processes. I’ll share more on that soon.
How does your marketing team run weekly meetings?
#ClaudeSkills #MarketingOps #MarketingReporting
Ziga Potocnik retweeted
Need a skill that pulls together three data sources most people check separately, but never at the same time?
It's called SEO Performance & Visibility Report, and it writes a full SEO report from live GSC, GA4, and whatever SEO platform you've got connected (Semrush, Ahrefs, Moz, or AccuRanker), all through Databox MCP.
Want to test it on your own data? 👉🏻 hubs.li/Q04q4zSZ0
#SEO #AIskills #MarketingAnalytics #ClaudeAI
Ziga Potocnik retweeted
SEO reporting means juggling Search Console, GA4, and your SEO platform every single month.
We built a free Claude skill that pulls all of it into one structured report in under 90 seconds: rankings, backlinks, CTR opportunities, competitor visibility, conversions, and prioritized recommendations.
Works with Semrush, Ahrefs, Moz, or AccuRanker.
👉 hubs.li/Q04q4yZF0
Ziga Potocnik retweeted
GA4 shows traffic per page but it doesn't tell you which content is dying, pulling the wrong audience, or sitting on high search demand with almost no clicks.
Rick Kranz from the AI Marketing Automation Lab built a free Claude skill that does exactly what was described. It pulls lifecycle classification, ICP scoring, prioritized editorial work queue. Monthly, from your live data.
Download it for free: 👉 hubs.li/Q04nTS8c0
Ziga Potocnik retweeted
Just today 976 products 😱😱😱😱 launched in @ProductHunt
it is insane, each one has a story & a person dreaming to either become a millionaire or escape 9-5
Ziga Potocnik retweeted
We're live on @ProductHunt today! 📣
AI tools are great at reasoning. But they're disconnected from your real business data.
Databox MCP fixes that. Connect your data to Claude, ChatGPT, or n8n and just talk to it.
Ask, get a chart back & automate the rest. The answer comes to you. You don't navigate to it.
Support us today 👇 [link in comments]
Ziga Potocnik retweeted
Dashboards are just pretty graveyards for data. 🪦
You open it at 9 AM.
Stare at a red line.
Go back to Excel to figure out why.
That is not analytics. That is a screensaver with extra steps.
Meet @databoxHQ MCP.
It connects your business data directly to Claude, ChatGPT, or Cursor.
No dashboard to build.
No SQL to write.
No "let me ask the data team and get back to you."
👇 You just... ask.
"What happened to my CAC last week?"
"Which Meta campaign actually drove revenue?"
"Why did traffic drop on Tuesday?"
Your AI does the math. Using your real metrics. Your definitions. Your history.
And it gives you an answer you can act on. Inside your chat.
The part that broke my brain:
✅ 130+ data sources (Google Ads, HubSpot, Shopify, Meta).
✅ Interactive charts rendered inside Claude/ChatGPT.
✅ Push data from anywhere → AI can see it instantly.
✅ Automate reports with n8n or Make.
Connect in under 60 seconds.
Then stop staring at red lines.
Start asking why they are red.
👉 Chat with your business data here:
databox.com/mcp
@databoxHQ MCP is live on ProductHunt right now:
producthunt.com/products/dat…
♻️ Repost if you have ever ignored a dashboard.
—
#Analytics #AI #Claude #ChatGPT #DataScience #DataboxMCP
Ziga Potocnik retweeted
How does getting your performance numbers directly from a Slack chat sound? 🤩
Guess what, there is free n8n template that makes this possible.
You @mention the bot, ask a performance question, and get an answer from your actual Databox data in seconds. 🎉
It's free, requires no code, and it only takes 15min to set up.
Grab it here: hubs.li/Q04htJlV0
Ziga Potocnik retweeted
If you haven't merged your paid ads and website data yet, you're missing insights neither source can show alone.😬
There's a free n8n template that fixes that by pulling data from both sources through Databox, running AI analysis across both, and emailing you a cross-channel report with recommendations every morning.
It's basically the briefing your analyst would build if you asked nicely. 🙊
Grab it here: hubs.li/Q04htrZq0
Ziga Potocnik retweeted
I once calculated how many hours my team lost to "data that lives somewhere else."
Export from this tool. Clean in Excel. Import into that dashboard. Repeat every Monday. Hope nothing breaks.
It was easily 5 hours a week. Per person.
And the answer from every analytics platform was always the same: "We don't have that integration. Submit a feature request. Wait 6 months."
So we changed the answer.
@databoxHQ just launched Custom Integrations.
Here's what it actually means:
→ Connect virtually any JSON-based API without writing a single line of code
→ Paste your API docs into ChatGPT or Claude, ask it to generate the config
→ Paste that config into Databox. Done. Live dataset in minutes.
→ Start asking questions in plain language immediately. No SQL. No dashboards to pre-build. No metrics to pre-define.
This isn't a minor feature update. It fundamentally changes what "supported integrations" means.
The output is a full structured dataset not just a metric. Queryable. Filterable.
You can explore and build on it immediately. Use it with our AI analyst or access it via MCP with any AI tool you already use.
OAuth2, complex pagination, dynamic date ranges — all handled. No configuration headaches.
What used to take a developer half a day now takes the time it takes to make a coffee.
This is the most significant update to @databoxHQ since our AI analyst launched. And no other analytics platform on the market does this in a single, code-free workflow.
If you've ever stared at a "Request Integration" button and felt your Sunday evening drain away : databox.com/custom-integrati…
And check out the product hunt launch here: producthunt.com/products/dat…
👇 Drop "DATASET" in the comments.
♻️ Repost for the person on your team still manually exporting CSVs every Monday morning.
Ziga Potocnik retweeted
📣 If your job depends on making sense of performance data — and getting there fast — this is for you.
On April 8th, we're launching Databox AI — and it changes how all of them work with data.
Here's what we're introducing:
🔍 Genie — your AI analyst, available 24/7. Ask anything in plain English. Get answers grounded in your real business data, with the context to act on them immediately.
📝 AI Performance Summaries — automated written overviews of your metrics, trends, and goal progress. Less time reporting. More time deciding.
🔗 Databox MCP — connect your performance data to Claude, ChatGPT, and tools like n8n. Analyze, automate, and act — without leaving the tools you already use.
Analytics has moved on. Have you?
✨ REGISTER NOW for the LIVE launch event on April 8th → hubs.ly/Q048S7wZ0
