I built a free Google Ads tool that catches the mistake nobody warns you about
Hey IH! I'm Michael, solo founder building SearchSavior.
Quick backstory: I kept seeing the same problem in Google Ads accounts. Advertisers add negative keywords to block wasted spend, but Google defaults every negative to exact match (the narrowest option) with no guidance on whether it should be phrase or broad instead. Worse, if you do use a broader negative, there's no warning if it accidentally blocks one of your own positive keywords. The traffic just stops and nobody notices.
So I built a tool that does two things no other negative keyword tool does:
Recommends whether each negative should be Exact, Phrase, or Broad match, with a plain-English reason
Checks your positive keywords for conflicts before you block, so you never silently kill your own campaign
It also estimates how much waste is hiding in the search term data Google won't show you (they hide 20-80% of it).
I just launched on Product Hunt last week and also shipped a free instant analyzer: paste up to 10 search terms and get AI-powered verdicts with match type recommendations. No signup, no email, no CSV upload.
Try it: https://www.search-savior.com/audit
Stack: React/TypeScript, Node.js, Supabase, Google Ads API, Gemini AI. Bootstrapped, built with Claude.
Would love feedback from anyone who manages Google Ads or builds in the PPC space. What would make this more useful?
the 'mistake nobody warns you about' framing is doing a lot of work but keyword self-competition is genuinely underrated. hard part for a free tool is the people making this mistake don't know enough to go search for a fix. might land better as a 'here's $X you're burning' teardown on someone's real account than a tool they have to find first, idk, depends if you can get read access
Hey ouuki, yes! Sorry four months late and that's entirely my fault, IH only just emailed me the notification today! Go figure. Apologies.
You are right, and I've spent the last month arriving at the same conclusion the slow way. Nobody searches for this. I checked the volume: the terms that describe what my tool does get almost nothing, because you can't look for a fix to a problem you don't know you have.
The teardown idea is where I've landed too. I'm about to offer a handful of practitioners a free run on a real account, in exchange for telling me what it found and whether it was something they didn't already know. That last part is the bit I actually need to know. If people run it and shrug because they knew everything on the list, that's a real answer.
Read access is the sticking point you'd expect, particularly for anyone managing client accounts. The way round it so far is a browser-based version that reads two CSV exports and never sends anything anywhere. It's a worse tool and it asks for nothing.
Thanks for taking the time back then. Wish I'd seen it in April.
the conflict-detection angle is smart because it's invisible until it's already cost you money, worst kind of bug. i've had overlapping campaigns quietly bidding against myself for weeks with no clue. tricky bit is distribution though, a free tool for a pain people can't feel yet is hard to get in front of them, the detection's kinda the easy half
Hey Lainey, thanks for the reply. Sorry for the four-month silence, IH only sent me the notification today (!!).
"A free tool for a pain people can't feel yet" is the best description of my problem I've read, and I've been circling it for months without putting it that plainly.
The bidding-against-yourself thing you describe is the same shape: invisible until somebody goes looking, and nothing in the interface prompts you to look. That's what makes it survive.
You're right that detection is the easy half. I've since found that Google itself flags some of these, and it's worth knowing how much. I built an account with four planted conflicts across all four levels and ran Google's own check against it. It found two, and the only fix it offered was deleting the negative, which throws away whatever it was protecting.
So the detection is partly solved and the distribution isn't solved at all. Thanks for saying so in April. ;-P
This is a really sharp use case — especially the conflict detection part, that’s something most people don’t even realize is happening.
One thought — with tools like this, I’ve seen that how clearly the value is communicated upfront (including the name) can make a big difference in early adoption.
Curious, how did you land on “SearchSavior”? Was it something you tested or just went with initially?
Hey aryan_sinh, thanks for the reply,five months late, apologies, IH only notified me today.
On the name: no testing at all, I just went with it. Which is roughly the honest answer to most naming questions from solo founders.
You're right that it doesn't communicate the value. It says something about search and something vaguely heroic, and nothing about the specific thing the tool does. Someone hearing it has no idea it's about negative keywords, let alone about the mistake they're making with them.
I haven't changed it, mostly because the domain, the API integration and everything else is wired to it, and because I have bigger problems than the name. Worth knowing it's a known weakness rather than a decision I'd defend.
Really appreciate you raising it, and sorry it took this long to answer.
That’s a useful answer, especially the fact that you already see the name as a known weakness. I’d be interested in continuing the conversation beyond the thread — would you be open to sharing the best email to reach you on?
Happy to keep talking here, easier for anyone else reading. What did you have in mind?
Mainly continuing the product/positioning conversation. Your point about the name being a known weakness caught my attention, especially since the core value seems much clearer than “SearchSavior” communicates.