How long till my website sees results

In July 2026 I pulled ninety days of search data for a plumbing company. Google had shown the site 315,407 times. It had earned 310 clicks. If that were your site, one person in a thousand who saw you would have visited.

Every first call ends with the same question. It usually comes after the owner has already decided they like you.

So, how long before any of this works?

There is a pause. What fills it is almost always a number between three and six months. It sounds careful. It is the number I heard when I was starting. It is the kind of number on page one of Google today. And it is not really a lie. It is an answer to a question nobody asked.

The question sounds like it is about time. It is really about which job you just paid for.

Not exactly. It is what the industry truly believes, and it is easy to check.

Search for how long SEO takes. You get about the same range from seven different publishers: four months to a year. Two to three months for words nobody is fighting over. Small moves in three or four months, steady spots in six to twelve. That is one trade magazine, two software companies, and four agencies. They mostly agree, and they are mostly reporting what they have seen.

That range has two problems. The second one is the serious one.

The first problem is that it gives one number for a pile of different jobs. Take two of them. One is fixing the setting that lets your email land in the inbox. The other is beating your competitors for a popular search. They are not the same kind of work, and nothing slows them down in the same way. My best guess is that the two are about three years apart.

The second problem is that the range is older than it looks. In a study published on 15 May 2025, Ahrefs tracked a million pages published in 2023 for a year and measured 1.3 million United States keywords. Patrick Stox wrote it up. Only 1.74 percent of those pages reached the top ten for anything. In 2017, that figure was 5.7 percent. The same study looked at who sits at number one. The average number-one page is now five years old, up from two years.

This is Ahrefs' chart of average page age in the top ten, from an earlier run. The 15 May 2025 update puts the number-one page at five years old.

Read those two lines together. Getting onto page one got about three times harder. The advice didn't change. One of the pages still on page one for this question was written in 2015. That is not a scandal either. This happens when an answer is copied instead of being checked again.

Because your website does not run on one clock.

Think of a website as a row of separate machines sitting side by side. Your email setup is one machine. The code that counts your leads is another. The pages that show up in Google are a third. The sales you want from it all are a fourth. Each one runs on its own timer.

When you ask how long, you hold up the whole row and point at one dial. Whatever number comes back will be wrong for most of what you are holding. That is where "three to six months" comes from. It is the average of things that should never have been averaged.

Here is what each dial really read, on work we have done, with the dates.

The email settings, the lead-counting code, and the hidden code are fast for one reason. You already control all three. Nobody is competing with you to fix your own email or your own tracking. Showing up in Google and getting visitors are the opposite. Someone else already holds those spots. They got there first, and they are still working to keep them.

The email to the existing list is worth a closer look, because I nearly printed it as a website result. I went back to the original message thread before publishing. The site was still being built that week. The orders came from an email to people who had already bought from her, plus a few social posts. The fastest win I have ever seen for a client had nothing to do with the website at all.

The first honest thing anyone can tell you isn't how long. It is which row of that table you are in.

This is a real account for a local barbershop. The numbers come from Google Search Console, not from memory.

In August 2025, its average spot on Google was 26.3. That is page three. By November, it was 12.0. The climb took about three months. Then came the part nobody shows you: it held. For ten months in a row it stayed between 10.5 and 13.2. Last month it was 10.6. That is not a spike, and it is not luck.

Now look at what happened to the visits over those same months.

Across those twelve months, the number of times Google showed the site fell by 55 percent. The clicks fell 22 percent, from 709 a month to 550. It climbed from page three to the edge of page one, but visits dropped.

That is not a failure. And I couldn't tell you that if I only printed the good line. The share of people who clicked went from 1.02 percent to 1.76 percent. That is up 73 percent. It went up because the searches that dropped away were never going to end in a haircut. But anyone who sells you a ranking as the goal should be able to explain those numbers. Most of them have never had to.

One more honest note about this account, because this article is about giving credit to the wrong thing. The site was redesigned, and I can date it. Tracking began in March 2016. The move to a new web address was planned that same week. They handed me the new site in May 2017. A year later, I told the owner that visitors had more than doubled. Every one of those dates is eight years older than the twelve months I just showed you.

So the redesign didn't move this site from page three to the edge of page one last year. It couldn't have. Search Console only keeps sixteen months of data. There isn't a clear before-and-after for a redesign that old, so I won't claim there is. What does show up inside those sixteen months is new search words. The week the climb started, the site began showing up for 154 new searches. Then 33 the next week, then 30, then just a few a week by this spring. That is a different story from a redesign, and it is the one the data supports.

Read that account again, because it contains what no agency prints.

The ranking more than doubled, and visits went down. Both of those are true for the same twelve months, from the same tool, for the same business. If I showed you only the ranking line, it would look like a win. It would be true. And it still wouldn't be what the owner cares about.

What improved wasn't the number of visits. It was what each visit was worth. More people who saw the site clicked because the searches that dropped away never led to a haircut.

A good ranking is a first step, not the result. It opens the door. Whether anybody walks through is a second number. That number needs its own measurement tool and its own date. The answer to "how long before I see results" needs to specify which of the two you're promising. They do not show up together, and sometimes the second one never shows up.

Some searches are for people who are still learning, not buying. You can rank first for one of those, and your phone will stay quiet. That is not a timing problem. You selected the wrong search to pursue, and waiting won't fix that.

They buy the leads. Almost everybody does. Almost nobody says so in an article like this one.

Go back to the plumbing company from the start. I measured it in July 2026. In ninety days, Google showed its pages 315,407 times, and those showings brought 310 clicks. That is one click for every thousand showings. Its average spot across the whole site was 27.4, which is page three. The searches that bring in money sat even lower.

Page three is not a ranking. It is a lot of people not seeing you.

So the phone does not ring on its own. The company takes the same steps that any healthy business does when free visits can't sustain it. It pays. It buys roughly 175 leads a month at about $200 each. They come from search ads, from lead websites, and from companies that charge per phone call. One of those call companies charges a flat $155 every time the phone actually rings. In one recent month, one lead website alone sent 53 leads for $7,935. That is just under $150 per lead.

The company's own tracking shows where the work comes from. It recorded 240 leads in thirty days. 149 came from ads on one search engine and 49 from ads on another. About four out of five leads are paid for. Google search on its own, with no ads, brings in about two leads a month.

The owner books about eighty percent of the leads that come in. Their own notes say it plainly: the problem is getting enough leads, not closing them. They are not bad at selling. They are hard to find, so they rent their spot on the page every month while the slow work runs.

Here's the part that should prevent anyone from reading this as the story of one struggling company.

Go back to the bathroom remodeler from the twelve-month row in the table. It ranks far better than the plumber. Its average spot is 16.6 over the last four hundred days and 11.9 over the last month. Its showings in Google went up fifty-six times in a year. It runs no search ads. I checked that instead of assuming it.

Its site had 2,493 visits in twelve months. Almost every one came from free search, social media, other websites linking in, or things the business owns. One exception: eight visits from the one service it pays for. That service is a lead website subscription at $455 a month. It is the only thing the remodeler pays for to get leads.

That subscription brings in about 11 or 12 leads a month. That works out to about $40 a lead. The plumber pays about $200. That is five times as much for the same thing.

And still, about four out of five of the remodeler's leads come from that lead website. Its own website form brings in the remaining fifth.

So here are two businesses. One ranks much better than the other and pays a fifth as much per lead. Both still buy about four-fifths of their work.

Neither company's ranking is feeding it. One of them has a much bigger bill.

Now look at what those eight visits mean.

The lead website sends the business's own website eight visits a year. The leads show up inside the lead website's own system. They never touch the business's site.

So the thing that brings in most of the work is not connected to the thing everybody asks me to speed up. They are not connected at all. I mean that plainly. You can move one a long way while the other sits perfectly still. That is exactly what a year of rising showings next to leads that did not rise with them looks like.

This is why "how long till my website sees results" needs a second question answered first: results at what? The site can get much better while the phone rings at the same rate. And nobody involved is lying.

Go back to the plumber. There is one more layer under the $200.

Many of the people clicking those ads and being counted as leads were already customers. They had used the company before. When the drain backed up again, they searched, saw the ad at the top, and clicked it. And the company paid for another $200 lead.

This was not an accident, and I am not going to pretend it was a mystery. The ad account was set up to do it. Someone pulled the customer list out of the scheduling software. They loaded it into the ad platform so past customers would be shown ads for a full year. It had its own budget line. The written ad plan says ads shown to past customers work three to five times better than ads shown to strangers.

They do. Of course they do. These are people who have already paid you once.

So the ad platform reports a success rate several times higher than anything else. And that higher number is what gets used to justify the spending. The ad did not create the demand. The customer's last job did, and that job was already paid for. The account is charging a toll on a road the business already built.

Now think back to the email sent to the existing list. The fastest result in this whole article was an email to people who had already bought, within an hour. It cost nothing. It is the same group of people, reached in two ways. One of those ways bills you every time somebody remembers your name.

Nobody means that when they say three to six months. And that is why the question has to be, "results at what?" Some of what a marketing budget buys is new customers. Some of it is a bill for customers you already had. Those are not the same purchase. They don't take the same time. And only one of them is worth waiting for.

The other choice was never "wait longer for the rankings." It was written into the same plan as the ad budget.

That plan included three other ideas. The first was a membership offer to make at the end of every finished job. The second was an email to the existing customer list. The third was a program to win back past customers, aiming for fifteen to twenty repeat jobs a month. All three reach the same people the ads were paying to reach. None of them charges a fee per click. None of them stops working when the card is declined.

The ad budget got funded. The other three stayed on paper.

Now compare that to the remodeler. The difference isn't talent, or luck. That business spends $455 a month. It owns everything else: the Google listing, its own web pages, its social media accounts, its email list. When its showings on Google grew fifty-six times in a year, it kept that growth. It did not rent it for a month.

Here's the honest answer to the timeline question, and it's difficult for my industry. Paid leads are fast, and they drop to zero the day you stop paying. Free leads you own are slow, and they keep adding up. Three to six months is about right for the slow kind. It is the wrong way to think about the fast kind. And many people who quote that number are selling you paid leads while describing the timeline for free ones.

There is a free option in between, and its problem is not what you would guess.

The remodeler's Google Business Profile, the free listing on Google Maps, costs nothing to run. It is already the business's third biggest source of visits. Only people typing in the web address and Google search itself send more. Turning viewers into customers is not its weak spot. But in ninety days, Google said the listing was found through just two searches, and one of them was the company's own name.

So the problem isn't the listing, and it isn't the ranking. The listing is fine. Google just isn't showing it to anybody.

Ranking and being found are not the same thing. A page can hold a top spot for a search nobody types. A listing can be built perfectly and still be shown twice. A report that only shows ranking makes both look the same. And neither one is fixed by waiting longer.

The same account has an even clearer example. I went into its data, hoping to find a win. I found the opposite.

Over ninety days, that site ranked between first and third for 74 different searches. Those searches showed the site to people 388 times. They brought one click.

Now compare a single blog post on the same site, from the same data pull. It sat down at ninth place. That one page was shown seven and a half thousand times and earned twenty four clicks.

Work out the rates, and they come out the same: about 3 clicks in 1,000, whether the page ranked ninth or first. Moving up eight places bought nothing anybody could measure.

Two warnings before you lean on this. First, it rests on a single click. One more visitor would change the rate a lot. Second, it is not a fair side-by-side. The ninth-place number comes from one blog post. The first-place number comes from seventy-four different searches, and people click on each one for different reasons. So treat this as a warning sign, not a rule.

But the direction is uncomfortable. It is the opposite of what a ranking report suggests when it shows you an arrow going up.

That list of searches has a second lesson, and it is the more useful one. The searches the site ranks first for include a misspelling of its own trade, the word "bleach", and the phrase "48 hours". Most of them were searched once or twice in three months. It ranks first for those because almost nobody is asking. That is not winning. It is an empty room with your name on the door.

This is not only true of one remodeler. In a study of over 80 million keywords and billions of search results, Sistrix measured how often people actually click each position on page one. Johannes Beus published it on 14 July 2020, and last updated it on 7 July 2025. The first result gets 28.5 percent of the clicks. The second gets 15.7 percent, the third 11 percent, and the tenth gets 2.5 percent.

Flip that around. Even the spot everyone fights for is skipped by seven out of ten searchers. By third place, nine out of ten skip it. Page one is not one place. A ranking gets you a share of attention, and that share is usually smaller than the bill for it suggests.

And that share is still shrinking. In a study published on 4 February 2026, Ahrefs compared 300,000 keywords, half with an AI Overview on the page and half without. Ryan Law wrote it up. An AI Overview is the summary Google now writes at the top of some search results. When one was there, the top page got 58 percent fewer clicks than it would without one. A year earlier, that gap was 34.5 percent. The ranking did not change and the page did not change, but the gap grew. No timeline mentions this because a timeline is written once, and this changed after it was written.

Not everyone agrees on that number, and the disagreement tells you more than either figure alone. In a study of 5.47 million queries across 53 brands, covering 2.43 billion impressions from January 2025 to February 2026, Seer Interactive found that clicks were starting to come back on searches with an AI Overview. Tracy McDonald, Hannah Cooley and Marketa Williams published it on 24 April 2026.

By early 2026, 2.4 out of every 100 people clicked on those searches. On searches with no AI Overview, the number rose faster, from 2.8 to 3.8 out of 100. So the gap did not close. Whichever study you trust, a page that used to be the answer is now one of several choices on the screen.

Seer also found the one thing anyone has shown helps here, and it is not ranking. Pages named in the AI Overview got 120 percent more clicks per showing than pages that were not named. Being on the page is not what matters. Being named in the summary is.

So there are three steps, not one. Ranked, then clicked, then a lead. Most of the industry sells the first step and hints that the other two come with it. A ranking is not the goal. From what we have seen here, it is not always even halfway there.

That is what the three-to-six-month answer leaves out. It does not lie about search. It just says nothing about the bill you pay the whole time you wait, or about whether that bill ever stops.

Here are four questions. If you know what you're selling, you can answer them on the first call.

• Which part of the work is this, and how long does that part take? If one number covers everything you bought, it averages things that do not belong together.
• What is the first number you will show me, and on what date? Not a report. One number, from a named tool, on a date you can put in your calendar.
• What result would tell us this is not working? Agree on that number before anybody starts. A plan that cannot fail is why vague timelines last so long.
• Once the ranking comes, what should happen next, and how will we see it? People ask this last. It should be asked first. Agree on the second number before the first one starts to move.

A vague timeline is not a gentle answer. It is a useful one for the person giving it. Three to six months is two quarters of bills where nothing can be judged. And in the end, nobody can say it missed, because the number was never clear enough to miss.

I would rather be held to something. So here is mine. The technical fixes take hours or days, and you will see the numbers change while we are on the call. Showing up more often in Google takes months, and the first honest check is at ninety days. Beating strong competitors for a popular search takes years. I will tell you before we start whether the search you want is one of those.

Answer out loud.

The last time somebody gave you a timeline, did they say which part of the work it was for?

If it covered all of it, that was not a real estimate. It was a number no one could check. By the time anyone could check, everyone would have moved on.

Below are three questions that tell a real estimate from a guess. Tick each one you can answer about the last timeline you were given. The box at the bottom tells you what your answers mean. It changes with each box you tick, because each missing answer causes a different problem.

How long does SEO really take? For a search nobody else is going after, it can take a few weeks to a few months. For a search your competitors are working hard on, plan for years. The usual answer of four months to a year is the middle of that range. It hides both ends.

Why do agencies all say three to six months? Because it is about right for the middle case. It is also long enough that you can't judge anything before two quarters of bills are paid. Most people who quote it are repeating what they were told, not what they measured.

What should I see in the first month? Fixes you can check for yourself, and a starting number to measure from. If month one brings a report and nothing you can check on your own, ask what was measured and how.

Is it worth doing if it takes that long? The fast jobs are worth doing no matter what. Several of them take an afternoon and keep paying off. The slow jobs are worth it when the search is one your buyers really use before they buy. Those are two different choices, so make them one at a time.

Does a ranking guarantee more customers? No. One of the businesses in this article more than doubled the quality of its ranking over twelve months. Its visits fell 22 percent in the same year. A ranking is a first step. Customers are a different number, and they need their own measurement.

What if it has been six months and nothing has happened? Then nobody agreed up front on what would count as working. That is the real problem. Ask what the first number was, what it read at the start, and what it reads now. If nobody can show you those three numbers, time was not the problem.

The next article in this series is for anyone who is not sure which job their timeline was for. It covers what happens when someone gives up too early: We tried it for a month and it did not work (coming soon). If you are not sure your numbers can be trusted, start with running ads with broken conversion tracking. A timeline is only as good as the tool that measures it. And if you would like someone to look at your own account and tell you which jobs are fast and which are slow, request a free site analysis.