A SERP monitor is usually bought on the price per keyword. Ten dollars a month for 250 keywords, forty for 1,000, and the comparison shopping happens in that column.
That column is not where the cost is. We priced a monitor from both ends this month: what the raw data costs per query per check, and how often a daily check produces a change you would actually act on. The data came out cheap enough to be almost free. The reviewing did not.
Here is the arithmetic, so you can size your own monitor to the time you have rather than to the seat count on a pricing page.
What we measured
Two separate measurements, because a monitor has two costs.
The data cost. We priced live Google SERP requests at three depths, from the API we use for rank data, on September 12, 2026, US location, English. Costs are what the requests were billed at, per query, per check.
The signal rate. We calculated how often a tracked query produces a change worth alerting on, using our own Search Console property: 90 days of query-and-date data, filtered to the 71 queries with at least 20 days of history, giving 2,188 query-days. For each query we counted the days where the average position moved more than three positions from the previous day.
Neither number is a vendor benchmark. Both are measurable on your own account, which is the point of doing it this way.
The data cost, measured
Per query, per check:
Depth requested | Cost per check | Organic results returned |
|---|---|---|
10 | $0.002 | 12 |
20 | $0.0035 | 23 |
100 | $0.014 | 97 |
Depth is the only lever that moved the price by a meaningful factor. Going from 10 results to 100 results costs seven times as much, and on this query returned eight times as many organic results. Pricing scales roughly with what you ask for, which is fair, and it means the cheapest configuration is the honest one for most programs.
Monthly data cost for a monitor, checking daily, 30 day month:
Keywords | Depth 10 | Depth 20 | Depth 100 |
|---|---|---|---|
100 | $6.00 | $10.50 | $42.00 |
250 | $15.00 | $26.25 | $105.00 |
500 | $30.00 | $52.50 | $210.00 |
1,000 | $60.00 | $105.00 | $420.00 |
A thousand keywords checked daily at depth 10 costs $60 a month. That is a number most teams can approve without a meeting.
Data cost scales linearly with keywords and depth. The 100 result column is seven times the 10 result column.
The signal rate, measured
Cheap data is only useful if it produces something to look at. This is the part that pricing pages do not cover.
Across 71 queries and 2,188 query-days, 770 query-days showed a move of more than three positions from the previous day. That is 35.2 percent of all days.
Measure | Value |
|---|---|
Queries with at least 20 days of history | 71 |
Query-days in the sample | 2,188 |
Days with a move over 3 positions | 770 (35.2%) |
Median query's rate of such days | 35.0% |
Quietest quarter of queries | 21.1% of days or fewer |
Busiest quarter of queries | 50.0% of days or more |
Queries moving on fewer than 10% of days | 5 of 71 |
Queries moving on more than 50% of days | 16 of 71 |
The distribution is the useful part. Five queries out of 71 are quiet enough that a daily check is close to wasted on them. Sixteen move on more than half of all days, so a monitor watching only them will alert constantly.
Note the instrument. These are Search Console average positions, which smooth across the day and across devices. A live SERP scrape will show more movement than this, not less. Treat the 35.2 percent as a floor.
The full cost model
Put the two measurements together for a 100 keyword monitor at depth 10, checked daily.
Line item | Monthly |
|---|---|
Data cost | $6.00 |
Signals generated (35.2% of 3,000 query-days) | 1,056 |
Cost per signal in data | $0.0057 |
Review at 60 seconds per signal | 17.6 hours |
Review at 30 seconds per signal | 8.8 hours |
The data is 0.7 percent of the bill if you price an hour of review at $50. Everything else is attention.
This is the argument for treating a SERP monitor as a review-capacity problem rather than a data problem. Two consequences follow, and both point the same direction.
Buying more keywords is nearly free. Reviewing them is not. Going from 100 keywords to 1,000 adds $54 a month in data and roughly ten times the signals. If nobody has ten spare hours a month, the extra 900 keywords will be paid for and unread.
Reducing noise is worth more than reducing price. Halving the number of alerts saves more than halving the data bill, because the data bill was never the constraint.
At every size, review hours outscale the data bill by more than an order of magnitude.
Three ways to cut the review bill
Each option changes what the monitor can tell you. None is free.
Raise the trip threshold per query. Instead of three positions for everything, set it above each query's own normal movement. Our quieter quarter of queries sit inside 21 percent of days, and the busiest quarter move on half of all days; one threshold cannot serve both. This cuts alerts on stable queries without muting the noisy ones. It does mean you will miss small moves on quiet queries, which are sometimes the interesting ones.
Check weekly instead of daily. A weekly check on 100 keywords at depth 10 costs about $0.86 a month and generates roughly 151 signals, or about 2.5 hours of review at a minute each. You trade latency for a monitor you will actually read. For most programs this is the right trade, and the case for daily only exists when you can respond the same day.
Watch a shorter list more closely. Fifty keywords checked daily with a per-query threshold and a monthly review is a monitor somebody owns. Five hundred keywords checked daily is a data feed somebody subscribed to. The second one looks better in a proposal and produces less.
Sizing rule
Set the keyword count from the review budget, not the other way around.
- Decide how many minutes a week the review actually gets. Be pessimistic. This is usually the hardest number to be honest about.
- Divide by the average triage time per alert, which for a position change is somewhere between 30 and 60 seconds.
- Divide again by your own signal rate over the last 90 days. If you do not have one, 35 percent is a reasonable starting estimate and you should replace it within a month.
- The result is your keyword count at daily cadence. Anything above it should be checked weekly.
Run against our numbers: 60 minutes a week, 45 seconds per signal, 35 percent signal rate gives about 228 keyword-days per week, or roughly 33 keywords at daily cadence. A 100 keyword monitor needs about three hours a week to be reviewed properly.
That is the number a pricing page will never tell you.
What this does not cover
Three limits worth stating plainly.
The data costs are what we paid for live requests in September 2026. Vendors change list prices, and a subscription monitor bundles crawling, storage, and interface into one number that is not directly comparable to a per-request price. Use the per-request figures to understand what you are buying, not as a vendor quote.
The signal rate is from one property with its own ranking profile. A larger site with more head terms will have different movement. The method transfers; the 35 percent does not.
The cost model prices review time but not decision time. Some alerts lead to a page change, a content brief, or an escalation, and that work dwarfs the triage. If your monitor is producing alerts that lead to real projects, the review number is the small part of the true cost.
Where this connects to the rest of the setup
A monitor is three decisions stacked: what to watch, how often, and what to do when it fires. This article prices the second one. For the first, the useful question is how far down the page you actually need to see, which we measured in what rank depth actually returns. For the third, separating your own changes from SERP-wide movement matters more than the alert itself, because a third of the alerts you will generate are not about you.
If you build the loop yourself rather than buying a seat, the sequencing and the failure modes are in our two source rank tracking build.
FAQ
How much should a SERP monitor cost? Data for 100 keywords checked daily runs from about $6 to $42 a month depending on how many results you request. The larger number in any monitor budget is the review time, which lands between 9 and 18 hours a month at the signal rates we measured.
Is daily monitoring worth it over weekly? Only if you can act the same day. Weekly monitoring at 100 keywords costs about $0.86 a month in data and produces about 151 alerts, which is a monitor someone will read. Daily produces seven times that.
Why do I get so many alerts if my rankings are stable? Because the threshold is usually site-wide. Our quietest quarter of queries move more than three positions on 21 percent of days, and the busiest quarter do it on half of all days. One threshold cannot serve both groups.
Should I monitor at depth 100? Rarely. Depth 100 costs seven times depth 10 and returned 97 organic results on a query where nobody was reading past the first page. If your queries genuinely rank past 50, the depth question is worth answering first.
Can I calculate my own signal rate? Yes, from Search Console. Pull query and date dimensions for 90 days, filter to queries with at least 20 days of history, and count the days where position moved more than three from the day before. That is about twenty minutes of work and it replaces the estimate with your number.
Auspia view: price a SERP monitor on data plus review, and the buying decision gets easier. The data is close to free at any depth worth using, so the configuration that matters is the one your team will read every week. Set the keyword count from the minutes you have, not from the seat tier that looked affordable.
Author: Tessa Clarke, Content Governance Lead for 120+ Editorial Systems at Auspia. Tessa writes about review workflows, monitoring cadence, and the operating rules that keep a search program readable.
