Campaign learnings in a spreadsheet - a real problem?

conversionninja

I've spent months running campaigns, and my brain is basically a sticky note covered in coffee stains. Every success, every flop - I log it all in a monster spreadsheet. then when a new campaign rolls around, I'm down on my knees sifting through rows of data, trying to spot patterns. sometimes I ask Claude or ChatGPT for creative ideas, but their suggestions feel so generic - like they've never even met my audience.

Existing tools? They're obsessed with churning out creatives faster or automating launches. Nobody's closing the loop. Nobody's saying, 'Hey, based on your past wins, try this angle for your skincare crowd.' So I'm toying with building something that actually learns from my own campaign history - no manual data entry required.

But before I dive in, I need to know:

Is this just my personal obsession, or do other marketers feel this pain too?
If a tool could automatically structure all your past campaign learnings and then serve up tailored creative recommendations for your actual audience - would that make you sit up?
Is this worth building, or am I chasing shadows?

No sales pitch. Just trying to see if the world needs this. Thanks for reading.

adcraft

Honestly, the bigger problem here isn't scattered notes-it's that most people are tracking vanity metrics and calling it "learning." If your tool just parses the same garbage I'm already ignoring, it's a hard pass from me.

The whole "automatically pull performance data" thing sounds fine in theory, but the real challenge is creative context. Like you said, sometimes an ad works because your competitor pissed off their audience that week, not because of any audience preference your model is trying to surface. If the tool can't distinguish signal from noise, you're just digitising my filtering problem, not solving it.

Data organisation alone saves hours, sure. But if the extraction layer isn't smart enough to flag correlation vs causation, then all you've built is a prettier spreadsheet. Worth building if you focus on the logic that rejects bad patterns, not just the one that repeats them.

budgetburner

people miss the point. The recommendation engine isn't the problem - it's getting clean data in the first place. Spent months structuring my own 'learnings' and roughly eight out of ten hours went into tagging creatives consistently so patterns would actually surface. Solve the data capture mess and the rest sorts itself.

funnelhacker

You're dead right about the core pain-nobody needs another dashboard that just shows you what happened. the real value is in the "what do I do next" decision based on actual performance signals, not generic best practices or GPT fluff.

I've burned serious budget on campaigns where the only thing holding me back was knowing which creative variant to push and which to kill. Most tools are either execution robots or vanity metric graveyards. they don't tell you "pause that keyword, shift budget to this ad group because its CPA dropped 15% after the last landing page change." That's the closed-loop insight that actually moves ROAS.

If I were validating this, I'd talk to 10-15 people who manage ad spend themselves-founders, solo buyers, in-house performance marketers. Ask them to walk you through their last "what do I test next" decision. If they all open a spreadsheet or admit they're guessing, you've got gold. if half say "my agency handles it," dig deeper. Does the agency give them actionable creative recommendations or just a report they ignore?

The real hard part isn't the "which ad won" part-that's trivial. it's connecting performance data back to specific creative elements. "Image A beat Image B" is noise. "Lifestyle shots with a subject looking at the product convert 22% better than product-only for cold audiences" is a revenue multiplier. That requires structured tagging, not just a dashboard.

for me, the biggest wasted spend comes from not knowing why a creative failed. You fix that, you unlock serious scale.

weekendhustle

Honestly, I built something really similar back in 2018 and it was a tough sell. You'd think people would throw money at it - the benefit is so clear - and they say it sounds incredible, but actually getting paying customers? properly hard work 😅

Ended up turning it into an internal tool and using it as a differentiator in proposals instead. Still found it useful, just not as a standalone product. Worth bearing in mind when you're validating demand.

ranktracker

Been running campaigns for years and honestly, what you're describing is exactly the thing that keeps me from losing my mind - so no, it's definitely not just you. The gap between historical data and generic AI suggestions is massive, especially when you're juggling multiple ad accounts or testing across different audience segments. I've watched people leave serious money on the table because they start every campaign from scratch instead of building on what their own data already tells them. Most tools treat your historical performance like it doesn't exist, which is insane considering that's your entire competitive advantage.

Your tool would tackle a real bottleneck, not a nice-to-have. The fact that nobody's nailed it yet suggests either it's harder to build than it looks, or there's no clear monetisation path - but the demand is absolutely there. My advice: validate hard with people managing multiple accounts or high-volume campaigns before you start building. Solo operators will probably stick with spreadsheets. Also, think carefully about whether you're pulling raw data directly from ad platforms or still relying on exports - that manual data entry piece is what kills adoption for most people.

ctrjunkie

We tried something similar not long ago. Honestly, the real headache wasn't the creative itself - it was tying any success back to shifts in audience targeting or the offer. Everyone gets obsessed with the visual, but that's rarely where the leverage is.

funnelhacker

Hell, I'd bet my last bid - what was the single metric that killed it? From the buy side, it's usually latency in the bid stream or lack of post-bid transparency, not just CPM. What feedback did you actually log?

weekendhustle

Honestly, this hits hard 😅. I've seen the same thing-even as a PR person, trying to get budget for something that doesn't promise a shiny new revenue stream is like pulling teeth. Everyone's been burned by empty promises, so trust is in short supply. It's exhausting constantly having to prove value instead of just getting to do good work.