Website marketing

Website marketing is two jobs. Most businesses do one.

Website marketing means both getting a website found and getting the visit to count — marketing OF the site through search, content, email and paid media, and marketing THROUGH the site so the person who arrives actually does something. They need different work and they fail differently, and almost every business we take on is competent at one and neglecting the other.

2

Jobs the phrase covers, which is one more than most plans budget for

20 +

Years building the sites as well as marketing them

300 +

Projects delivered, from local firms to Fortune 500

20,700 +

Client keywords in Google's top three, measured in Ahrefs

// The Two Halves

Getting found, and getting the visit to count

The same phrase covers both, which is why briefs written around it so often fund one and assume the other. They are not interchangeable: doubling traffic to a page that does not convert doubles nothing, and a page that converts beautifully with nobody on it is a very expensive piece of design.

Getting found: search

Organic positions for the questions your buyers already type, which is the cheapest traffic there is once it works and the slowest to arrive. This is also the half that now feeds AI answers, because the engines read organic results before deciding whose name to put in the summary.

Getting found: earned coverage

Being written about somewhere your buyers and the search engines both already trust. It is the part that cannot be bought cleanly or rushed, and it is what makes the difference between a site that ranks for its own name and one that ranks for its category.

Getting found: paid

The top of the page today, at a price that rises with the competition. Its real value on a site that is not yet ranking has little to do with the traffic. It is finding out which messages and which queries actually convert, before you spend a year of content on the wrong ones.

Getting the visit to count: speed and clarity

How fast the page renders and whether a stranger can tell within a few seconds what you do and what happens next. This is the least glamorous work on the page and it moves the number more reliably than anything above it, because it applies to every visitor from every channel at once.

Getting the visit to count: the ask

One job per page, a form no longer than it has to be, and a next step that matches how ready the reader actually is. A demo request placed in front of somebody still working out whether they have the problem will produce a bounce with extra steps rather than an inquiry.

Getting the visit to count: trust

Named people, real proof, checkable numbers, and a phone number that belongs to somebody. On a site selling a considered purchase this does more work than any layout change, and it is usually sitting three scrolls below where a visitor decides.

The half you are neglecting

There is a quick test. If traffic is flat and inquiries are flat, the problem is the first half. If traffic is fine and inquiries are flat, it is the second, and no amount of additional traffic will fix it — you will simply pay more per inquiry than you did last year.

Why we do both

Because we build the sites as well as market them, which means the second half does not need a third party, a change request queue or a fortnight. Most agencies can only recommend the fix. Twenty-odd years of production code is why we can also ship it.

// Process

What we actually do to a website

In this order, because the order is most of the value. Every stage produces something you can check rather than a status update.

Measure what the site does now

Where it ranks, what it converts at, which pages produce the inquiries you already get, and how much of the traffic is arriving on pages that were never meant to receive it. This stage frequently finds that one neglected page is doing most of the work, which changes the plan considerably.

Data in →

Fix what every visitor hits

Speed, clarity, mobile behavior, the path from landing to inquiry, and the tracking that will tell you whether anything after this worked. It comes first because it multiplies everything that follows, and because it is the only work that improves the visits you are already paying for.

Strategy →

Point the channels at it

Paid to learn quickly, organic and earned coverage to compound. The evidence from the first weeks of paid decides which pages get built and which keywords get a year of work, rather than the other way round.

Execute →

Keep testing the ask

Forms, offers, page order and the next step, changed one at a time and measured. Conversion work is where the compounding actually happens: a rate lifted this quarter keeps paying on every visitor from every channel for as long as the page is up.

∞ Loop
// From the archive

One of these we got right. One of these we got wrong.

Eight passages from our own archive, verbatim and dated. All of it was standard website marketing advice in 2010. Some of it still is, and one piece is now a way to damage a site, which is a more useful thing to show you than another paragraph about our expertise.

While article directories don't get as much love as they used to from search engines, they still get points for original, keyword relevant content, and the best article directories are still good sources of back links to your site. It is still not unusual for articles from the top article directories to rank on the first page of the search engine results for some topics.

This one we got wrong, and it is here rather than deleted because that is worth more to you than the pages where we were right. Panda landed the following year and took the category apart. Read the first line again, though, because the tell is in it: "don't get as much love as they used to". We could see the decline and recommended the tactic anyway, which is the failure mode to watch for in any advice, ours included.

While you can access services that submit articles to many directories, this isn't the best idea. For one thing, duplicate articles all over the place aren't going to get much love from the search engines (since they like original and not duplicated content). And for another, five high quality articles submitted to five high quality article directories will benefit you more than two dozen crappy articles submitted everywhere in a scatter-shot approach.

The venue died and the reasoning did not. Five good placements beating two dozen bad ones is how our digital PR work is run now, and the automated submission services are still being sold, under new names, to people who have not read anything this old.

First of all, know what kind of registration form works best with your particular landing page clientele. The conventional wisdom is that you lose up to one-third of your respondents for each registration field. Whether this is true or not, it is certain that seeing a lot of registration fields is a turn-off to most people, so be sure you're actually using whatever information you plan to squeeze out of your respondents.

This one we got right and would not change a word of. The one-third figure is still roughly what we measure, and the sentence after it is the part people skip. The commonest offender in 2026 is the same as it was in 2010: a phone number field on a form whose next step is an email.

Put yourself in your landing page user's shoes. What would he or she think would be sufficient information for you to have from them? Maybe their job title, company, phone number, and potential time frame of purchase would be appropriate, but go much beyond this, and people won't bother at all.

Still the whole test, and still the one nobody applies to their own form. The useful version asks what the visitor gets in return for each field, and whether they would think that trade fair if they were not already inside your business.

All it takes is a statement at the top welcoming new subscribers. Users take a little bit of a leap of faith when they click on an ad. It's nice to know you've landed on an actual landing page rather than gone down some virus-laden internet rabbit hole.

Message match, described in 2010 without the jargon that later grew around it. The leap of faith is larger now, not smaller, because a good share of visitors arrive from a summary they did not click through and have no idea what they are looking at.

As you may have guessed, it is common to design a Web site with a wonderful appearance that is easy to use and has useful, interesting content but still fail to capture a solid percentage of visitors. In many cases this happens because the site doesn't include a call to action in the proper place and at sufficient frequency. A call to action should be a highlight of your Web site.

This is the second half of this page in one paragraph, and the diagnosis has not moved. A beautiful site that nobody asks anything of is the most expensive thing in this business, because everything upstream of it looks like it is working.

Your call to action should also be presented in the right form. Does your visitor have the opportunity to request a sample or demonstration? Can he or she ask for a free trial or a no-cost consultation? This type of call to action should garner you more leads than simply asking the visitor to provide contact information. Give them something in exchange. After all, they are giving you themselves.

"After all, they are giving you themselves" is the line, and it is the one nobody writes any more. Match the ask to how ready the reader actually is, and give them something for it. Sixteen years of testing has added detail to that and has not replaced it.

Google wants you to speed up your site. They did some experiments where they deliberately slowed search results page loads to see how users would respond. They found out that slowing down a page by 100 to 400 milliseconds produces 0.2 to 0.6% fewer searches. Not only that, searches dropped even more over a period of weeks, and, even if the page loads returned to normal, it took users a couple of weeks to return to their normal search habits!

Written the week speed became a ranking factor, and the part worth keeping is the tail: the damage outlasted the slowness by weeks. That is still true of every kind of bad experience a site can give somebody, and it is why the speed work comes before the traffic work rather than after it.

Around 700 posts, published since 2003

Read the archive
// Straight Answers

The questions people arrive here with

Usually asked at the point where the site has been live a while and the numbers are not moving.

What is website marketing?

Two jobs under one phrase: marketing the website so people find it — search, content, earned coverage, email, paid media — and marketing through the website so the people who arrive act. Both are needed, and a plan that names only one of them has a hole in it whether or not anyone has noticed yet.

Is website marketing the same as web marketing?

Close enough that nobody will correct you. Web marketing is the wider program across every browser-based channel; website marketing is normally used for the part centred on the site itself. Our main page on the wider version sets out all eight channels and how to sequence them.

We get traffic and nothing happens. What is wrong?

Usually one of three things, in this order of likelihood: the traffic is arriving on pages that were never built to receive it, the ask is wrong for how ready the visitor is, or the traffic is the wrong traffic — real people, genuinely searching, who were never going to buy. The third is the expensive one, because it looks like success on every report that counts sessions.

How much traffic do we actually need?

Fewer visits than most people assume, and the sum works backwards rather than forwards. Take the inquiries you need, divide by a conversion rate you can evidence rather than hope for, and that is the number. Businesses that do this arithmetic often discover the traffic problem they thought they had is a conversion problem they can fix for a fraction of the money.

Should we redesign the site?

Usually not, and this is the recommendation we make least often and get thanked for most. A redesign resets the rankings, the tracking history and the things that were quietly working, in exchange for a look. If the diagnosis is speed, clarity or the ask, those are fixes rather than a rebuild. If the platform itself is the constraint, that is a different conversation and we build sites too.

How long does it take?

Conversion work shows up fastest: changes to speed, clarity and the ask can move the rate within weeks because they apply to traffic you already have. Getting found is slower — a quarter for movement in the position bands, a second or third quarter for the clicks. Which is another argument for doing the second half first.

// Results

Numbers don't lie.

16
Case studies with published numbers
19
Client brands
12
Countries the work has run in
22
Press features
TMS Brain Health
81.8%
Of all site traffic from organic search
Monthly organic users, against March 2020
2,000%
Phone-call conversions from organic search
52%
Timeline: Mar 2020 – Jan 2023 · #1 in LA & Santa Monica for TMS therapy
Utopia Management
247
Keywords in Google's top three, across city-level local search
Organic visibility, up 12.5 points
26.61%
Of the previous year's monthly organic users
280%
Timeline: 679 local terms tracked with the map pack — and still seven times the nearest competitor in 2026
Link Building & Digital PR
DR 82
Average authority across 81 earned editorial placements
DR 90+ placements
33 of 81
Featured in
CNN · Forbes · U.S. News
Timeline: Healthline, Nasdaq, Yahoo! Finance, Seattle Times & more
Budget Trucks (budgettrucks.com.au)
440%
Organic traffic against the penalty floor — an all-time high
Weekly organic sessions under the penalty
~1,300
Weekly organic sessions after the fix
~5,700
Timeline: Oct 2011 – Nov 2014 — twelve months suppressed, then the best weeks the site had ever recorded
SNS Nails
49.92%
Organic visibility — three times the nearest competitor
Monthly organic traffic, against the starting level
440%
Keywords in Google's top three
61
Timeline: Jun 2022 – Jun 2024 — up from 3% visibility, with nothing lost or declined in the top three
Budget Rent a Car (budget.com.au)
+20.6%
Organic users, year over year — 640,088 against 530,730
Australian-locale sessions
+696%
Bounce rate on organic traffic
27.6%
Timeline: Jun–Oct 2014 against the same four months in 2013 — the audience got bigger and it got local
Selby Acoustics (selby.com.au)
24.52%
Organic visibility — first in the category, ahead of JB Hi-Fi
Keywords in Google's top three
228
Estimated organic traffic
+174%
Timeline: Aug 2023 – Aug 2026 — up 11.5 visibility points while JB Hi-Fi, Officeworks and Harvey Norman all lost ground
On Campus Marketing (rhl.org)
+61.0%
Unique visitors from organic search, year over year
Visits from non-paid search
+49.0%
Selling-season window, measured a year apart
13 days
Timeline: May 2013 against May 2012 — the move-in selling season, non-paid search only
My Site Buddy (mysitebuddy.ai)
20
Active development days for a build costed at six months of senior work
Manual estimate
820–1,060 hrs
Shipped
120+ pages
Timeline: 15 Jul – 12 Aug 2026 · 361 commits, 9 concurrent workstreams, 11 releases — every one approved by a person before production
Ross Metals (rossmetals.com)
33
Keywords at number one — gold, chain, wire and permanent jewelry
Organic visibility, up from 0.40%
15.30%
Top-three rankings lost in twelve months
0
Timeline: Jul 2025 – Aug 2026 — about 3,900 searches a month held at position one, including "wholesale gold" at $3.72 a click
Isentia (isentia.com)
+61.4%
Organic sessions, year over year
Contact form submissions
+76%
Form completion rate, up from 8.1%
10.6%
Timeline: Aug–Sep 2015 against the same weeks a year earlier, across ten markets
George Clinical (georgeclinical.com)
+286%
Monthly sessions, January 2015 to August 2018
Organic share of all sessions
57%
Contact inquiries from organic search
55%
Timeline: January 2015 to August 2018 — organic finished as the largest channel, ahead of every other source combined
Maine Cottage (mainecottage.com)
80 points
Swing in year-over-year traffic growth, inside twelve months
Year-over-year traffic at the start
-40%
Year-over-year traffic twelve months on
+40%
Timeline: Jan–Oct 2013 finished +14.7% on visits and +18.3% on unique visitors
GoPromotional (gopromotional.co.uk)
1,682
Keywords at number one in Google UK — 240% of Vistaprint, 510% of 4imprint
Keywords in the top three
2,894
Keywords on page one
4,166
Timeline: Of 8,027 tracked — first on every band, and first in the category since 2010
Expedia Japan (expedia.co.jp)
20
Brand terms at number one on Yahoo Japan — every one tracked
Brand terms at number one on Google Japan
17 of 20
Generic travel terms on page one of Yahoo Japan
17
Timeline: May 2010 — including 海外ホテル, "overseas hotels", at number one on Yahoo and number two on Google
Private Jet Charter (privatejetscharter.net)
8
Country sites run at once — and every one of them grew
Average growth across the seven country sites
+77%
Germany, the fastest of them
+123%
Timeline: Eight markets in six languages — with the localized sites holding visitors two to three times longer than the global one
// Testimonials

What people say about working with us

We have been working with John and Web Moves for 14 years and the results have been astounding. John is a straight shooter — if he says he can do it, you can be confident he will deliver.
S

Shaun O'Brien

Owner, Selby Acoustics

Shaun was John's client

The great thing about John is his honesty — he tells it like it is. He is respected by major players including Microsoft and eBay, and media seek out his views: the front page of the Wall Street Journal among them.
P

Phil Leahy

Founder, MedCart Marketplace & Retail Global Events

Phil was John's client

John is the single most knowledgeable individual in the SEO industry I have ever worked with. Whether as a client, vendor or business partner, I cannot recommend him more highly.
M

Matthew Kligerman

Building Dirt — AI land development platform

John was Matthew's client

// Next Steps

Traffic fine, inquiries flat?

Send us the URL. We will come back with which of the two halves is actually costing you — where the traffic is landing, what it does when it gets there, and the two or three changes we would make first. If the answer is that the site is fine and the problem is elsewhere, we will tell you that instead.