Tracking your rankings without obsessing

The rank tracker somebody sold you works by sending automated queries to Google. Google's spam policies name that practice, describe it as scraping results for rank-checking purposes, and state that such activities violate both those policies and its terms of service.

Which is an awkward foundation for a monthly report. It is also, on its own, not the strongest argument against the habit. The stronger one is that Google publishes what to do when your position moves, and the instructions are almost entirely about waiting: a full week before you even look, a small drop treated as no reason to act, and months rather than days before a genuine improvement registers. What follows is that guidance, and what to watch instead. Everything adjacent is filed on the getting-booked hub.

What does automated rank checking cost?

More than it returns, and the policy is explicit.

Under the heading of machine-generated traffic, the spam policies describe the practice of sending automated queries to Google.

They state that this includes scraping results for rank-checking purposes, or other types of automated access to Google Search conducted without express permission.

The stated reason is that such traffic consumes resources and interferes with the ability to serve users, and the stated consequence is that these activities violate the spam policies and the terms of service.

That is not a warning about your rankings. It is a statement about the tool's method.

Which is worth knowing before paying for something monthly, and worth asking any agency about directly.

What the answer tells you about the rest of their work is usually more informative than the report itself.

The policies are published as Google's spam policies for web search.

Lost Coast, FL

What does a rate-limited request look like?

A specific status code, defined by an internet standard.

Any server dealing with more requests than it wants from one source has a defined way of saying so.

The public registry of status codes lists 429 as Too Many Requests, referencing the specification that introduced it.

That specification states that the code indicates the user has sent too many requests in a given amount of time, describing the condition as rate limiting.

It adds that the response should include details explaining the condition and may include a header indicating how long to wait before making a new request.

None of which is exotic. It is simply what the internet does when something asks too often.

A tool built on repeated automated queries is designed around getting that answer and working past it, which tells you what kind of tool it is.

The registry is published by IANA, and the code is defined in RFC 6585.

What is a core update?

A broad change, several times a year, announced.

The documentation says that several times a year Google makes significant, broad changes to its search algorithms and systems, that these are called core updates, and that notice is given on a published list of ranking updates.

It then says something most people never hear: in general, most sites do not need to worry about core updates and may not even realise one has happened.

It describes the changes as broad in nature and as not targeting specific sites or individual pages.

Its own analogy is a list of favourite restaurants written in 2019, revisited later: some new places are now candidates, some move up, and the ones that move down are not necessarily bad, there are simply others that now make the list.

That framing is worth adopting, because it removes the assumption that a drop is a punishment.

Most of the anxiety in this subject comes from reading a change as a verdict.

The published account is Google's page on core updates and your website.

What are you told to do about a drop?

Wait, twice, before looking at anything.

The first instruction is to confirm the update has finished rolling out, checking the status dashboard and noting the start and end dates.

The second is the one nobody follows: wait at least a full week after a core update completes before analysing the site.

Only then does it suggest comparing that week against a week before the update began rolling out, so the change can be pinpointed.

It also advises reviewing your top pages and queries, and analysing different search types separately, since a drop in one surface is a different problem from a drop across all of them.

Add the rollout period to the mandated week and the honest interval before you may form a view is often a month.

A daily rank report is therefore producing thirty readings for a question that cannot be answered until the thirtieth.

The measurement that is permitted and useful is described in the local search piece.

How big does a drop have to be?

Bigger than most of the ones people panic about.

The guidance divides the response by size, and its examples are specific.

A small drop, illustrated as moving from position two to position four, warrants no drastic action, and it goes further: it recommends avoiding changes to content that is already performing well.

A large drop, illustrated as moving from position four to position twenty-nine, is what triggers a deeper assessment.

Between those two examples sits almost every position change a small operation ever notices.

So the documented answer to the most common observation in this subject is to do nothing at all.

Which is a difficult instruction to sell as a monthly service, and a very easy one to follow.

The self-assessment a genuine drop should trigger is discussed in the blogging piece.

Port Aransas, TX

What are you told not to do?

Three things, and the first is the entire industry of quick fixes.

Avoid quick fix changes, with the guidance's own example being removing a page element because you heard it was bad for search.

Focus instead on changes that make sense for users and are sustainable in the long term, and consider improving content in meaningful ways, such as rewriting or restructuring so a page is easier to read and navigate.

And treat deleting content as a last resort, only where it cannot be salvaged.

Then it adds a line worth sitting with: if you are considering deleting entire sections of your site, that is likely a sign those sections were created for search engines first rather than for people.

Which is a diagnosis rather than a rule, and an uncomfortable one for anybody who built nine town pages from a template.

Where that applies, it says removing the unhelpful content can help the good content perform better.

What those pages should have contained instead is set out in the piece on place pages.

How long does a genuine improvement take?

Days at best, months realistically, and possibly the next update.

The guidance says some changes can take effect in a few days, but that it could take several months for the systems to learn and confirm that a site as a whole is producing helpful, reliable, people-first content in the long term.

If months pass with no effect, it says that could mean waiting until the next core update.

It also notes that smaller core updates happen continually and are not announced, because they are not widely noticeable, and that these are another route by which improved content can rise.

Then the sentence that should end most arguments: there is no promise that changes will produce a noticeable impact, and positions are not static or fixed in place.

It describes results as dynamic by nature, because user expectations evolve and the web itself constantly changes, and says this can cause both gains and drops.

Read plainly, position is a moving quantity in a system nobody has promised you an outcome from.

Which is precisely why it is a poor thing to measure a business by.

So what should a guide watch?

Four things, none of them a position.

Enquiries, counted, by month, which is the first thing search work is supposed to produce.

Where each booking client says they found you, asked at the ramp and written down, which is the only measurement that describes your market.

Impressions and clicks from the search console, which is Google's own data about your own site and requires no scraping.

And whether the pages you built are actually reachable and indexed, which is a yes-or-no question rather than a number.

Those four take about twenty minutes a month between them.

None of them moves daily, which is the point, and all of them describe something you can act on.

The verification that unlocks the third is covered in the search fundamentals piece.

Why does position vary anyway?

Because it is not a property of your page.

A rank is the outcome of one query, from one place, on one device, at one moment.

Change any of those and the answer legitimately changes, which means a single number reported as your position has quietly discarded three of the four variables.

For a guiding operation the location variable is the one that dominates, since the searcher is often hundreds of miles away and the operator is checking from the launch town.

That single mismatch invalidates most kitchen-table rank checks before anything else is considered.

An averaged position across many actual searches, which is what the console reports, is a different and far more honest quantity.

It is also less satisfying, because it moves slowly and cannot be improved by refreshing.

How the distance factor works against a service-area business is examined in the map pack piece.

What about a genuine, sustained drop?

There is a documented response, and it is uncomfortable.

Where a site as a whole has fallen a long way and stayed there, the guidance sends you to a self-assessment about whether the site is delivering content that is helpful, reliable and people-first.

It asks you to look at the site as a whole rather than at individual pages, and to try to be objective about it.

Then it suggests something most operators will not do, which is asking somebody you trust and who is unaffiliated with the site to run the same assessment.

And it asks you to evaluate the pages most affected specifically, noting that there may simply be other pages on the web doing a better job for that searcher.

That last clause is the honest one. A drop can mean somebody else got better rather than that you got worse.

Neither reading is comfortable, and only one of them is fixable by editing your own site.

What a genuinely useful page for this trade contains is described in the schema markup piece and the pieces around it.

What does an agency report actually show?

Usually a measurement of the wrong thing, presented monthly.

The standard report leads with positions for a set of chosen terms, which is the quantity least connected to whether anybody booked.

It is also the quantity most sensitive to the variables a report cannot hold constant, meaning the searcher's location, device and history.

A more useful report would lead with impressions and clicks from the console, then enquiries, then bookings, in that order, because each is closer to the outcome than the one before it.

Ask for that order explicitly, and ask what data source each line comes from.

If the answer to the first line is a rank tracker, you have learned something about the method rather than about your site.

None of this means an agency is acting in bad faith. Position reporting is simply the convention, and conventions persist because clients expect them.

Changing what you ask for changes what you get.

What else is worth asking before hiring anybody is set out in the vetting questions piece.

Where do guides go wrong here?

Seven habits, and the first is checking at all.

Searching your own terms from your own device and treating the result as a measurement.

Paying for a tool whose method the spam policies describe as violating the terms of service.

Reacting to a two-place move, which the guidance explicitly says needs no drastic action.

Making a change, waiting four days, concluding it failed, and making another.

Assessing during a rollout rather than after it, against a direct instruction.

Deleting pages in a panic, which the guidance treats as a last resort.

And measuring position instead of enquiries, which is the substitution underneath all six.

The listings that genuinely change visibility are assessed in the directories piece.

What surprises operators here?

That the documentation is mostly telling them to relax.

Most people expect a page of tactics and find a page of intervals, thresholds and cautions against acting.

The second surprise is that most sites are said not to need to worry about core updates and may not notice one happened.

The third is the instruction to avoid changing content that is already performing well, which contradicts the instinct to optimise everything.

The fourth is that the tools sold for this measurement are described, in method, by a spam policy.

The fifth is that the honest waiting period before judging anything is measured in weeks and months rather than days.

Taken together they describe a subject with far less to do in it than the market around it implies.

The markup and structure work that is genuinely finite is covered in the on-page basics piece.

How would you set this up?

One page, one calendar reminder, twenty minutes a month.

Write down the four things worth watching and the month you will look at them.

Add a column for enquiries and one for where each booking client said they found you, both filled in as the season runs rather than reconstructed.

Open the search console once a month, look at impressions, clicks and average position across the site rather than for one term, and write the three numbers down.

Check the status dashboard only when something has visibly changed, and note the start and end dates before drawing any conclusion.

Then close it and do not look again until the following month.

A year of that produces twelve honest readings, which is enough to see a trend and too few to panic about.

The keyword work those readings should inform is described in the keyword piece.

The habit, in order

Measure enquiries, wait properly, change rarely.

Stop checking positions from your own device, since the result describes your kitchen rather than your market.

Stop paying for automated rank checking, whose method the spam policies address directly.

Count enquiries and ask every booking client where they found you, which is the measurement that is actually about you.

Open the search console monthly, not daily, and read impressions and clicks rather than a single position.

When something moves, confirm any update has finished, then wait a full week before analysing.

Treat a small drop as nothing, and leave content that is performing well alone.

When you do change something, expect days at best and months realistically before it registers.

And accept that positions are described as dynamic rather than fixed, which makes them a poor foundation for a decision.