I built my own SERP tracker
Most SEO platforms are priced like a toolbox when the job needs one screwdriver. What it cost to build a focused rank tracker instead, and who shouldn’t.
If the need is narrow enough, owning the data beats renting a platform — but only if someone will maintain it.
I built a small SERP tracker because I kept running into the same problem. I didn’t need another SEO platform. I needed to know whether a defined set of keywords moved up or down, on desktop and mobile, over time. That’s it.
I’m not a developer by trade and I don’t pretend to be one. But modern SEO work is getting harder to separate from data pipelines, APIs and reporting — and if I understand how data is pulled, stored and reported, I’m better at the job I actually sell.
The problem isn’t the platforms. It’s the bundle.
The big SEO platforms aren’t bad. I use them, and there are cases where they’re clearly the right choice. But a lot of SEO software is priced like a toolbox when you only need a screwdriver. The cheapest plan often includes more than a small business will ever use, and the price reflects the product category rather than your use case.
That makes sense from the vendor’s side — it’s easier to sell a platform than one workflow. From the client’s side it can be a lot, especially when the real need is: check these 300 keywords once a week and show me whether we’re going in the right direction.
For that I don’t want another login and another dashboard. I want the data in BigQuery, connected to Looker Studio, comparable with Search Console, GA4, and later leads or revenue.
Is custom cheaper? Not automatically.
A custom tool isn’t free just because AI helped write the code. It needs an API provider, a cloud service, storage, scheduled jobs and maintenance. Build it badly — too many checks, everything stored without thinking, careless queries — and the cost climbs.
The argument isn’t “DIY is cheaper”. It’s narrower: if the need is focused enough, the custom route can be better value, because you’re not buying a whole suite to use one part of it. That’s especially true if the company already has a reporting stack. If BigQuery and Looker Studio are already in the picture, SERP data is just another source.
What the numbers looked like
Prices checked 2026-05; API and platform pricing changes, so treat these as the shape of the comparison rather than current figures.
| What you’re paying for | Rank tracking angle | |
|---|---|---|
| Ahrefs | A full SEO platform | Lite plan: 750 tracked keywords, weekly updates |
| Semrush | A full SEO and marketing platform | Entry SEO plan: 500 tracked keywords, daily |
| DataForSEO | SERP data through an API | Request-based; cost depends on frequency and depth |
| Custom observer | Your own tracking system with cloud storage | Keyword count, device, location and schedule under your control |
For 750 keywords weekly on DataForSEO’s Standard Queue, at $0.00465 per SERP with 100 results: $3.49 per run, roughly $15 a month. I calculate with the top 100 rather than the top 10 — checking only the first ten results is usually too shallow.
Cloud costs sit on top: Cloud Run, Cloud Scheduler and BigQuery. At this size that stayed under $1 a month, thanks to the free tiers — unless you store or query carelessly.
Why not just Search Console?
Because it answers a different question. Search Console is still one of the first places I look, but average position isn’t a clean SERP check: it mixes impressions, devices, locations and query variants, and whatever Google decided to show that day.
A SERP tracker is more boring, and that’s the point. For this keyword, in this location, on this device — where did the domain appear? That kind of boring data is useful for monitoring specific pages and keyword groups. A hundred well-chosen keywords often tell you more than five thousand random ones.
Search Console shows what happened in the market. SERP tracking shows what happened in a controlled check. I’d rather use both than pretend one replaces the other.
Can it be built with AI-assisted coding?
Yes, if you don’t try to build the next Ahrefs in a weekend.
That’s where these projects usually go wrong: the scope inflates. First it’s a SERP checker, then it needs accounts, a dashboard, alerts, competitor tracking, billing, and an AI feature because everything apparently needs one. At that point you’re not solving your SEO problem — you’re accidentally building SaaS.
The first useful version only had to post SERP tasks, fetch results, parse the target domain’s position, store it in BigQuery and make it reportable. Two or three days with AI-assisted coding. The hard part isn’t writing the code. It’s knowing what the tool should not do.
That’s also why I like this work: it forces the measurement to be defined properly. Devices, locations, keyword sets, schedules, data structure, reporting. Those aren’t developer problems. They’re SEO problems that happen to need code.
What AI search changes here
A rank tracker thinks in keywords: one query, one SERP, one position. AI search doesn’t behave that neatly — a single question fans out into several hidden ones. When I looked at the fan-out around SERP checkers, the topic moved into free tools, paid tools, local tracking, mobile versus desktop, APIs, SERP features, accuracy problems and the future of rank tracking in generated results.
So a SERP tracker isn’t “the answer” to AI visibility. It’s one layer, and a stable measurement point in a search environment that’s becoming less stable. Because the data already sits in BigQuery, it’s easier to combine with everything else — which is where the custom setup gets more interesting than a closed dashboard.
There’s a related direction worth noting: DataForSEO runs an MCP server, so some SEO data can be pulled straight into AI workflows. Useful for exploration. For monitoring I still want a database — a chat window isn’t a reporting system.
Who shouldn’t build this
A company shouldn’t build this if nobody will maintain it. That’s the boring answer and it’s the true one. No technical confidence, no reporting setup, no interest in owning the data — then a commercial platform is the safer choice. Pay the subscription, use the UI, move on.
But if the need is a focused rank-tracking workflow and the company already cares about data ownership or custom reporting, building a small tool isn’t a developer fantasy. It’s a normal business decision.
The question usually isn’t whether Ahrefs or Semrush is worth it in general — they often are, for the right user. The question is whether this company needs the full package every month.
Where mine is
I published it: SERP Observer on GitHub. It isn’t a product and it isn’t polished. It’s a working example of one specific approach: track selected keywords, store the results, connect them to reporting, and leave room for weekly AI-assisted summaries later.
Every time I build something like this I understand the measurement side of SEO a bit better — and that’s what makes the work I actually sell better.