Traffic

How do I get more website traffic?

Almost every traffic problem is one of four things, and Search Console will usually tell you which one within twenty minutes. The expensive mistake is buying the wrong lever because it was the one somebody was selling.

Hub and spoke internal linkingA three-level link map. The home page feeds three hubs, one for services, one for industries and one for locations, and each hub feeds two child pages. Dashed links run sideways between siblings, so no page is left without a route in or out.HOW THE PAGES HOLD EACH OTHER UPHOMESERVICE HUBWHAT YOU DOINDUSTRY HUBWHO YOU DO IT FORLOCATION HUBWHERE YOU DO ITGUIDEANSWERGUIDEANSWERCITY PAGECITY PAGEDOWN TO CHILDREN, AND BACK UPACROSS TO SIBLINGSNO PAGE IS LEFT AN ORPHAN.
Hub and spoke internal linking
In short

How do I get more website traffic?

Website traffic grows through four levers: publishing pages that deserve to rank for questions people actually search, improving the click-through rate of results you already appear in, removing the technical problems that stop pages being found or loaded, and paying for visits through ads.

Which lever to pull is a diagnosis rather than a preference, and Search Console usually settles it: many impressions with a low click-through rate is a titles and snippets problem, few impressions is a coverage or authority problem, and plenty of clicks with no inquiries is a page problem rather than a traffic problem.

More traffic is also not always the right goal, because sessions that never become inquiries still cost money to acquire and make every other number harder to read.

Framing

Traffic is a symptom, not a goal

Nobody wants traffic. They want more inquiries, more booked appointments, more orders. Traffic is one input to that, and it is the input most likely to be increased in a way that does nothing for the business.

It is possible to double a site's sessions without gaining a single customer. Publish enough loosely related content, or buy broad enough ads, and the number goes up while the phone stays quiet. This happens constantly, and it usually survives several reporting cycles because the headline metric is moving in the right direction.

So the honest version of this question is narrower: how do I get more of the right people, from the searches that indicate they need what I sell, onto pages that are capable of converting them. That reframing changes which lever you pull, and it is the reason the first section below is a diagnosis rather than a tactic.

It also changes what counts as success. A site that loses a third of its informational sessions and gains a fifth of its inquiries has had a good quarter, and a traffic-first report will describe it as a disaster.

The context

Why impressions no longer convert into visits the way they did

SparkToro analysed a Similarweb clickstream panel of US Google searches to measure how many end with nobody clicking through to any website. This is the backdrop to every traffic conversation now.

68.01%of US Google searches ended without a click, January to April 2026
60.45%the same measure in 2024, shown for trend

SourceSparkToro, zero-click search analysis 2026 (Similarweb clickstream panel)

This does not mean search stopped working. It means ranking is now worth a mix of clicks and visibility rather than clicks alone, so a flat traffic line can sit on top of rising exposure. Check inquiries before concluding anything from sessions.

Anatomy

Where your impressions are actually going

Before deciding you have a ranking problem, look at what the results page for your terms actually contains.

The anatomy of a results pageA stylised search results page with each block called out in turn: the search box, an AI overview, paid results, the local map pack, the organic results and the people also ask questions.ONE QUERYSUMMARYADSEARCH BOXONE QUERY, SEVERAL RACESAI OVERVIEWSUMMARISES, THEN CITESPAID RESULTSBOUGHT, PRICED PER CLICKMAP PACKTHREE LOCAL LISTINGSORGANIC RESULTSEARNED, NOT BOUGHTPEOPLE ALSO ASKTHE QUESTION AFTER THIS ONEONE PAGE, SEVERAL SEPARATE RACES. EACH BLOCK IS WON A DIFFERENT WAY.
Ads, a map pack, an AI answer and People Also Ask can all sit above the first organic result.

For a commercial query there are usually paid results at the top. For anything with local intent there is a map pack, which for a business with a physical location is often worth more than the organic listing beneath it. For informational queries there is increasingly an AI Overview, and People Also Ask panels can push organic results a long way down.

The practical consequence is that position three on a page with four ads, a map pack and an overview is a very different piece of real estate to position three on a plain results page. Two terms with identical rankings can deliver wildly different traffic for reasons that have nothing to do with your site.

This is worth checking manually before spending anything. Search your five most important terms, from a logged-out browser, and look at what is above you. Sometimes the answer is not to rank higher organically but to appear in the map pack, or to accept that the informational term will not send clicks and go after the commercial one instead.

Twenty minutes

Diagnose before you spend anything

All of this uses tools you already have, for free. Do it before agreeing to any proposal, including ours.

  1. Check that your pages are actually indexed

    In Search Console, open the Pages report under Indexing and read the reasons pages are excluded. Noindex tags left over from a staging site, canonical tags pointing at the wrong URL, and pages Google has crawled but chosen not to index are all common and all invisible from the front end. A page that is not indexed cannot receive any traffic at all, so this check comes first every time.

    • Compare the number of indexed pages against the number you expect to exist
    • Look specifically for "Excluded by noindex tag" and "Alternate page with proper canonical tag"
    • Check that your sitemap is being read and is not listing dead URLs
  2. Read the queries you already appear for

    Still in Search Console, open Performance and sort by impressions. This is the single most useful report most businesses never look at. It shows the questions people are asking that you are already relevant to, in their exact words, along with your position and click-through rate for each.

  3. Split the problem into impressions or clicks

    If impressions are high and clicks are low, people are seeing you and choosing someone else, which is a titles, snippets and relevance problem you can often fix in an afternoon. If impressions themselves are low, you are not competing for enough searches, which is a coverage and authority problem measured in months. These two situations need completely different work and money.

    • High impressions, low click-through: rewrite titles, tighten the page to match the query
    • Low impressions, decent position: you are ranking for too few things, so coverage is the gap
    • Low impressions, poor position: it is an authority and relevance problem, not a formatting one
See the remaining steps: Diagnose before you spend anything2 more stepsHide the remaining steps: Diagnose before you spend anything
  1. Check whether clicks are already being wasted

    In analytics, look at which pages receive traffic and what happens next. If pages are getting visits and producing nothing, more traffic will produce more nothing. Conversion problems are cheaper to fix than traffic problems and they improve every future visit, so they usually come first.

  2. Find out where your existing inquiries come from

    If you cannot say which channel produced last month's calls and forms, no traffic decision can be judged. Call tracking, form source fields and correctly configured conversion events are the minimum. This is boring and it is the thing that makes every subsequent decision arguable or not.

At the end of these five steps you will know which of the four levers applies to you. Most sites we look at need indexing fixes and better matching of existing pages to existing queries, not new content and not more ad spend.

Core Web Vitals

The speed thresholds worth measuring against

Google publishes the Core Web Vitals thresholds it considers good. They matter for traffic partly as a ranking input and mostly because a slow page loses the visitor before the content loads.

2.5sor less for Largest Contentful Paint, the main content appearing
200msor less for Interaction to Next Paint, responsiveness to a tap or click
0.1or less for Cumulative Layout Shift, unexpected movement of the page
75thpercentile of page loads, segmented separately for mobile and desktop

SourceGoogle web.dev, Web Vitals thresholds

Interaction to Next Paint replaced First Input Delay as a Core Web Vital, so an older audit report may be measuring something that no longer counts. Measure at the 75th percentile on real visits, not on a single lab test from a fast connection.

Start here

Quick wins that are usually still available

In rough order of how often we find them unclaimed, and how quickly they pay back.

  • Pages excluded from the index by a leftover noindex tag or a mistaken canonical
  • Two or three near-duplicate pages competing for the same query, which should be merged and redirected
  • Titles that describe the business rather than answering the query the page ranks for
  • Queries with high impressions and a weak position, where the existing page nearly matches and needs sharpening rather than replacing
  • A Google Business Profile that is incomplete, miscategorised, or missing services and hours
See the full checklist: Quick wins that are usually still available3 more itemsHide the full checklist: Quick wins that are usually still available
  • Images several megabytes larger than they need to be on the pages people land on most
  • A phone number that is not a tappable link on mobile, and a form that asks for more than it needs
  • Internal links missing between related pages, leaving good content with nothing pointing at it

None of these require new content or new budget. Most of them are an afternoon of work on a site that already exists, which is why they are worth clearing before anyone proposes a programme.

Quality over volume

Traffic you should not want

It is easy to buy or attract visits that cannot become customers. They still cost money, and they poison the reporting.

From a search to a booked jobA path running left to right: a search, then your page, then a branch into either a phone call or a form and chat, then a booked job. A faint branch drops away from the page to show the people who leave instead.FROM A SEARCH TO A BOOKED JOBSEARCHA QUERY WITH INTENTYOUR PAGEPROOF AND A NEXT STEPCALLFORM OR CHATBOOKEDTRACKED TO ITS SOURCELEAVESNOT EVERY CLICK CONVERTS.THE PAGE’S JOB IS TO LOSE FEWER OF THEM.EVERY STEP IS A PLACE TO LOSE SOMEONE, OR A PLACE TO MAKE IT EASIER.
The only part of the path that pays is the part that ends in an inquiry you can serve.

Visitors from outside your service area, researchers with no intent to buy, clicks on broad ad keywords that match the wrong meaning of a phrase, and traffic from content written about subjects you do not actually sell. All of these inflate sessions and none of them produce revenue.

They also distort every average on the dashboard. Add enough unqualified traffic and your conversion rate falls, your engagement rate falls, and the reports start recommending fixes to problems you do not have.

The cure is to judge channels on inquiries rather than sessions, and to be willing to see a traffic number fall. Cutting a broad ad keyword or consolidating a set of thin pages will often reduce visits and increase calls, which is the correct trade.

Reporting

Measuring it without fooling yourself

Use both Search Console and your analytics, because they answer different questions. Search Console tells you what people searched, how often you were shown, and whether they clicked. Analytics tells you what happened after they arrived. Neither one alone can tell you whether the work is paying.

Know what your metrics mean before you act on them. Bounce rate in GA4 is defined as the inverse of engagement rate, not as the share of single-page visits it meant in the older version, so a bounce rate imported from an old mental model will mislead you. Sessions and users are different things, and comparing one period's sessions against another period's users is a common and confusing error.

Read the full breakdown: Measuring it without fooling yourself2 more paragraphsHide the full breakdown: Measuring it without fooling yourself

Set the comparison honestly. Compare like periods, allow for seasonality, and remember that a change in Google's results layout can move your clicks without anything on your site changing. Google publishes a decision tree for debugging search traffic drops that is worth working through before assuming a penalty.

Then hold the work to the right standard. Technical and conversion fixes can show up within a month. Coverage and authority compound over weeks and months, not days. Anyone promising a fixed traffic increase by a fixed date is describing something they cannot control.

Want someone to run this diagnosis with you?

We will look at the same reports, tell you which of the four levers is actually your constraint, and be straight with you if the answer is that your traffic is fine and the problem is elsewhere.

Questions

Straight answers.

Why did my website traffic suddenly drop?

Work through the causes in order of likelihood: a technical change such as a noindex tag or a broken migration, a change in how the results page looks for your terms, seasonality, a broad algorithm update, or a genuine loss of rankings to competitors.

Google publishes a decision tree for debugging search traffic drops, and Search Console will usually show whether impressions fell, clicks fell, or both. Impressions holding steady while clicks fall points at the results page rather than at your site.

How much traffic does a small business website need?

There is no useful universal number, because the figure that matters is inquiries rather than visits. A local practice can run comfortably on a few hundred well-targeted sessions a month if the pages convert.

Work backwards instead: how many new customers do you want, what share of inquiries convert, and what share of visitors inquire. That produces a traffic target specific to your business rather than a benchmark from someone else's.

Does a high bounce rate mean the page is failing?

Check which definition you are reading first. In GA4, bounce rate is the inverse of engagement rate, so it reflects sessions that were not engaged rather than visits to a single page. The old single-page definition no longer applies.

Because of that, there is no clean universal target. Compare a page against similar pages on your own site and against its own history, and treat a sharp change as a signal to investigate rather than an absolute score.

Does social media bring website traffic?

Some, and it varies hugely by platform and audience. For most local businesses social traffic is modest compared with search, and its bigger contribution is awareness that later shows up as branded searches.

It is a reasonable supporting channel and a poor substitute for being found at the moment someone is looking for what you sell.

Is buying traffic ever a good idea?

Paid search is legitimate and often sensible, particularly for testing whether demand exists, for terms you cannot realistically rank for soon, and for filling gaps while slower work compounds. It stops the moment you stop paying, which is the trade.

Buying traffic from traffic-selling services, pop-under networks or bot providers is not the same thing and has no value. The visits do not belong to anyone who could become a customer.

How long does it take to increase traffic?

Indexing fixes and title improvements can show within days or weeks because the rankings already exist. New content and authority building compound over months, and competitive markets take longer than quiet ones.

A fair review point is about a quarter. Anyone promising a specific percentage increase by a specific date is overselling, because nobody controls the results page.

Sources

Where this comes from.

Primary documentation and published research behind the guidance on this page.

Next step

Talk to the team

A short call, a look at how the business currently shows up, and a straight answer on what we would do first.