Hey Indie Hackers,
I've run Meta ads myself, and one thing that always bugged me: building a report that actually shows what's improving (and what's not) takes way more effort than it should.
Yeah, tools like AgencyAnalytics, DashThis, Whatagraph, etc. already do this — but they're full-stack platforms. You pay for 50+ integrations when all you actually need is clean Meta Ads reporting for your clients. That pushes small agencies into $100-800/month plans for one feature they'll actually use.
So I'm building something small and focused: a tool that auto-generates a clean, white-labeled, shareable PDF report from your Meta Ads data — no dashboard to configure, no 15 metrics to wade through, just the numbers that matter for your client's goal (leads, sales, awareness, etc.), ready in minutes.
I want to talk to people who manage Meta ads for clients (agency owners, freelancers, media buyers) — what your reporting process looks like today, what's annoying about it, and what a report actually needs to have to make a client happy.
If that's you, drop a comment or DM me — happy to hop on a quick call.
The focused approach makes sense. One question I would test early is whether agencies really want a better PDF or an easier client conversation. A report can be clean and still create work if the client immediately asks what changed, why it changed and what happens next. I would want a short plain language summary beside the numbers, plus a clear comparison with the previous period. Have the agencies you spoke with complained more about building the report or explaining it afterward?
I like the question you raised. Although I didn't talk any agency yet. And this is one of the reason why i am here. I need to understand more deeply the problem. Do you run ads?
The focused approach makes sense compared to paying for broad agency reporting platforms.
What would convince you that Meta Ads reporting alone is painful enough to become a standalone workflow, rather than something agencies tolerate as part of a larger reporting stack?