Five Signs SEO Is Working Before Your First Click
The signs SEO is working show up weeks before the first visitor does, and not one of them is traffic. They are crawl speed, indexing, impressions, position and the queries you never targeted, and all five are visible for free in Google Search Console.
Most people publish, wait ten days, see a flat traffic graph and conclude the whole thing failed. They are reading the last metric in the chain first. A page has to be found, then visited, then stored, then shown, and only then clicked. Judging a two week old article on clicks is like judging bread on the crust before it has gone in the oven.
TL;DR
- Clicks are the last signal to arrive, never the first. On a young site they take months, and a flat traffic graph tells you almost nothing.
- Signal one is crawl speed, signal two is indexing, signal three is your first impression, signal four is position direction, signal five is the queries you did not plan for.
- Your first impression is a bigger event than your first click: it proves Google judged you a plausible answer to a real query.
- A first position around ninety is a starting block, not a sentence. It drifts both ways for weeks before settling.
- Google's own documentation tells site owners to expect four months to a year before a change shows an effect.
Table of contents
- Signs SEO is working when your traffic is still zero
- The ladder every new page climbs
- Signal one: your pages get crawled faster
- Signal two: the index fills up
- Signal three: the first impression arrives
- Signal four: your position starts moving
- Signal five: queries you never targeted appear
- What is not a signal, and what it costs you
- How long to wait, and when to stop waiting
- Your ten minute check, once a month
- FAQ
- Conclusion
Signs SEO is working when your traffic is still zero
A new page is not competing for a click yet. It is competing for attention from a crawler, then for a place in an index holding billions of documents, then for a slot on a results page. Each of those is a separate decision made by a separate system, and each one leaves a trace you can read.
That is why "no traffic" is not a result. It is the absence of the final step, which tells you nothing about the four steps before it. A page can be doing everything right and still be invisible, simply because it is at stage two of five.
The useful question is not "am I getting visitors". It is "how far along the chain has this page travelled, and did it move since last month". That question always has an answer, even at zero traffic.
The ladder every new page climbs
Search Console names these stages almost literally, in the Page Indexing report and the URL Inspection tool. Here is the whole ladder, with the label you will actually see.
| Stage | What Search Console says | What it means | What to do about it |
|---|---|---|---|
| Not found | URL is unknown to Google | Nothing links to the page and it has not been read from your sitemap | Link to it from a page that is already indexed |
| Discovered | Discovered, currently not indexed | Google has the address but has not visited yet | Add internal links, be patient, do not republish it |
| Crawled | Crawled, currently not indexed | Visited and judged not worth storing for now | Make the page deeper or more distinct, link it better |
| Indexed | Submitted and indexed | The page is stored and can be shown | Stop watching indexing, start watching impressions |
| Shown | Impressions appear in Performance | Google served it to a human on a real query | Watch which queries, and at what position |
Two things are worth noticing. Four of the five stages happen before any visitor exists, so most of the early work is invisible on a traffic graph. And the fix is different at every rung: rewriting an article stuck at "Discovered" is wasted work, because nobody has read the old version yet.
Signal one: your pages get crawled faster
The first thing to check on a new page is not its ranking. It is whether it has been crawled at all, and how long that took.
Crawl delay is the earliest health signal a young site produces. If your first articles took eight days to be visited and the ones you publish now are visited within a day, something real has improved: Google decided your site is worth coming back to more often. That happens before rankings move, and it is the most encouraging thing a new site can see.
The opposite case is just as informative. A page sitting at "Discovered, currently not indexed" for weeks with no crawl date is not being judged on quality, because nothing has been read. That is a discovery problem, and the fix is structural: point internal links at it from pages that are already indexed, and check that it appears in your sitemap. This is one of the quiet reasons a real content distribution strategy beats publishing into a void. Pages that receive links get found, orphan pages wait.
Be honest about the limits, though. Internal links reliably move a page from unknown to discovered. Getting it from discovered to crawled is much slower, and on a young domain that can take several more weeks. Adding one link is not a switch you flip.
Signal two: the index fills up
Indexing is the moment your page stops being a candidate and becomes a document the engine can serve. Until then, nothing else on this list can happen.
Read it as a ratio rather than a status: pages indexed against pages published, checked once a month. A site where that ratio climbs is a site being taken more seriously. A site where a third of the pages sit at "Crawled, currently not indexed" is being told something specific, which is that the pages were read and judged not worth keeping.
The two cases call for opposite responses. Discovered but never crawled is a linking problem, so add internal links and wait. Crawled but not indexed is a content problem, so the page needs to be more distinct, more specific or more useful than whatever already covers that query.
One warning that saves weeks: being indexed quickly is not a quality verdict. Indexing is a filing decision. Plenty of pages get indexed within a day and are never served to anybody.
Signal three: the first impression arrives
An impression means your page was displayed to a real person who typed a real query. To get there it had to be indexed, judged relevant to that query, and ranked above every page that finished below it. That is a genuine editorial verdict from the search engine.
A click, at the position where a young page lives, is mostly a lottery. Almost nobody scrolls to the eighth page of results. So on a new site, impressions carry the meaning and clicks are the noise, which is the exact opposite of what most dashboards imply.
Read impressions three ways:
- Do they exist at all. Zero impressions after a month on an indexed page means the page is a candidate for nothing. That is a targeting problem, not a patience problem.
- Are they growing. Four impressions this week against one last week is a fourfold move on a tiny base. Weak evidence, but it points up.
- Which queries produced them. That is signal five, and it is where the real information lives.
The same logic applies when you start watching where your brand appears in AI answers rather than in blue links: presence comes long before traffic. If that is on your list, measuring brand share of voice in AI answers sets out what to count, and AI visibility tracking turns it into a monthly routine instead of a weekly worry.
Join the waitlist if you would rather have this loop run itself. The reading stays yours either way: a machine can collect these numbers, but deciding a topic is not worth another month is a judgment call.
Signal four: your position starts moving
New pages enter the results somewhere between position sixty and one hundred. That number feels like a rejection. It is closer to a provisional placement while the engine works out what your page is for.
What matters is not the number but the direction over several weeks, and it moves both ways. A page can enter at ninety five, climb to eighty two as it gets matched to better queries, then slip back when a more established page is re-evaluated on the same query. No single move means much. A month of consistent drift in one direction does.
| What you see over a month | Likely meaning | Reasonable response |
|---|---|---|
| Position improving, impressions rising | Being matched to better queries | Leave it alone, send it more links |
| Position stable, impressions stable | Settled at its true level for now | Nothing to fix on the page, work on authority |
| Position drifting down, impressions flat | Competitors re-evaluated above you | Normal on a young domain, do not rewrite |
| No impressions at all | Candidate for no query | Reconsider the target query, not the prose |
The only line worth internalising is the last one. The rest is weather.
Signal five: queries you never targeted appear
Search Console shows you the queries that produced your impressions. Some will be the phrase you wrote the page for. Many will not.
Those unplanned queries are the most valuable data a young site produces, for two reasons. They tell you what Google actually thinks your page is about, which is often narrower or wider than you intended. And they hand you the phrasings real people use, in their words rather than yours.
Three patterns deserve a reaction:
- A query with clear intent, ranking badly, on a page not built for it. That is a new article waiting to be written, and the most reliable content idea you will ever get, because the demand is already proven on your own domain.
- Several near identical queries spread across two of your pages. Those pages are competing with each other. One should own the intent, the other should link to it.
- Queries far from what you sell. Attention you cannot convert. Interesting, not actionable.
This is the same instinct that makes keyword research for an online store work when you start from buying intent instead of search volume: the phrase people type tells you what they want. The difference is that here you are not guessing, because the queries came from your own pages.
What is not a signal, and what it costs you
Some numbers feel like progress and are not. Each one has a real cost, because chasing it burns a week you could have spent publishing.
- Your position when you search for yourself. Personalised, localised, biased by your own history. Use the Performance report, never your own browser.
- SEO scores from plugins and audit tools. They measure compliance with a checklist written by the tool vendor. Nobody at Google sees that number.
- Word count. A longer page is not a better page. It is a page that took longer to write.
- Daily rank checks. Positions fluctuate for reasons unrelated to you. Checking daily turns noise into anxiety, and anxiety into pointless rewrites.
- Social shares of the article. Pleasant, and unrelated to whether the page ranks.
One more trap deserves its own line, because it costs more than the others: rewriting a page that has never been crawled. You are editing a document nobody has read. And if you want the full version of what search engines do and do not penalise, AI content and Google penalties separates the documented rules from the folklore.
How long to wait, and when to stop waiting
Google's own guidance for site owners, in the Search Central documentation on whether you need an SEO, says changes usually take four months to a year to have an effect, and that this belongs in your expectations from the start. That is the publisher of the search engine setting the timescale, not an agency managing a client.
Read against that, most self assessments happen far too early. Two weeks is not a result. One month is barely a first data point. A quarter is the first window where a trend means anything. It is also why publishing rhythm beats publishing intensity: ten articles in one weekend then silence gives a crawler one burst and eleven empty months. How often to publish blog posts without a team covers what a sustainable cadence looks like when nobody is paid to write.
Three months of indexing with zero impressions, though, is a real verdict. It usually means one of four things. The query has no demand, and nothing about the page will fix that. The query belongs to marketplaces and ten year old brands, so you should go longer tail instead. The page answers no clear intent, covering a topic without addressing a question anybody asks. Or the page is fine and the domain is simply young, which is the most common case and the least satisfying: the answer there is to publish more, link better and wait.
Search behaviour is shifting underneath all of this too. A growing share of the questions your buyers used to type now goes to an assistant, which moves where visibility lives without removing the need for it. ChatGPT versus Google search looks at where buyers actually go now, getting mentioned in ChatGPT covers what makes a brand quotable, and the generative engine optimization guide lays out the foundations. The measurement discipline is the same one described here.
Your ten minute check, once a month
Monthly, not weekly. Weekly measurement on a young site produces noise and bad decisions.
- Indexing. Open the Page Indexing report. Count pages indexed against pages published, and note anything stuck at Discovered or Crawled.
- Crawl delay. Inspect your two most recent pages. Days between publication and first crawl, compared with last month.
- Impressions. Total impressions this month against last month, site wide. Ignore clicks until you have some.
- Best pages. The three pages with the most impressions. Are they the ones you expected.
- Queries. The ten queries producing the most impressions. Does any of them deserve its own page.
- Orphans. Any indexed page at zero impressions two months running. Note it, do not act yet.
That is the whole routine. Write the six numbers in a file so that next month you are comparing rather than remembering. Several of these steps can be automated once the habit exists, and SEO automation is fairly clear about which ones genuinely can and which ones only look like they can.
FAQ
How long before I see any sign at all that SEO is working? Crawling and indexing within days to a few weeks. First impressions typically within one to two months on a new site. First clicks are a matter of months, and Google's documentation points at four months to a year for a change to show an effect.
Is zero traffic after one month normal? On a new domain, yes, and close to expected. Check that pages are indexed and starting to gather impressions. Those are the two things that should be moving. Clicks come later.
What if my pages are indexed but get no impressions? That is a targeting problem, not a patience problem. The pages are stored but never judged relevant to a query. Look at what people actually search around your subject and write for that phrasing instead.
Should I check Search Console every day? No. Data lags by a couple of days and positions fluctuate constantly. Monthly is enough to see trends, and it stops you rewriting pages because of a bad Tuesday.
Do impressions without clicks have any value? As a signal, yes: they prove you are a candidate answer on a real query. As traffic they are worth nothing, which is why they should be read as a leading indicator and never reported as a result.
Conclusion
The five signs SEO is working all arrive before the first visitor: pages crawled faster than the last batch, an index that fills up, a first impression on a real query, a position that drifts up over a month, and queries you never planned for. None of it looks like success on a traffic graph, and all of it comes before success.
The work in between is repetitive rather than difficult: publish, link, distribute, measure once a month, repeat. Distrify is built to run that loop across search, social and AI answers at once, on the same keywords, and the engine behind it already publishes daily for live sites. The product is not open yet, so Join the waitlist is currently the only way to hear about it first.