Why long pages sometimes convert better than short ones

The advice you get is contradictory, which is a sign that both camps are describing their own customers. Keep it short, people do not read. Go long, more information means more trust. Both are true for different buyers.

The useful question is not how long. It is what does this specific buyer need to know before they will pick up the phone.

Does page length affect conversion?

Not directly, and there are two controlled experiments large enough to settle it.

Bing runs a dynamically sized results page, and Ron Kohavi, Alex Deng, Roger Longbotham and Ya Xu report both directions of that test in their 2014 KDD paper, written out of Microsoft and LinkedIn.

In the first, they truncated the page to four results instead of ten for queries where a single block already took over three quarters of the clicks. More than 8 million triggered users. The abandonment rate did not change to any statistically significant degree, with a p value of 0.92. They shipped it, because the shorter page was faster.

In the second, they extended the page to 20 results and removed related searches. Over 3 million triggered users. Revenue fell 1.8%, page load slowed by 30 milliseconds, and pagination dropped 18%. Abandonment again did not move, p value 0.93. They did not ship it.

Read those two together. Cutting a page in half and more than doubling it both left abandonment untouched. Across roughly 11 million users, length itself did nothing. What mattered was what the length cost: speed in one case, revenue and a slowdown in the other.

That is the honest version of the debate. Buyers do not read pages the way you write them. They scan for the thing they are worried about, and if they find it they keep going. A long page full of answers reads fast to someone scanning, because they skip everything that is not their question.

Nobody ever left a page because it was too long. They left because they stopped finding answers.

Is a shorter page automatically faster?

Which businesses need long pages?

It tracks with how much risk the buyer is carrying, and how unfamiliar the purchase is.

High price, infrequent purchase, and irreversible outcomes push toward longer pages. A roof replacement, a legal matter, a medical procedure, a wedding venue. The buyer does this once, they cannot undo it, and they will read everything.

Low price, familiar purchase, and urgency push toward shorter. Someone whose drain is blocked right now needs your phone number and your response time. Nothing else on the page will be read.

Competitive markets push longer too. If four businesses look interchangeable in the map pack, the one that explains its process, its pricing, and its guarantees gives the buyer something to decide on. The others are relying on being first.

And technical or unfamiliar services push longer, because part of your job is teaching the buyer enough to evaluate you at all.

What belongs in the extra length?

Answers, in the order the buyer thinks of them. Not sections in the order your business is organized.

What exactly is included, and what is not. This is the most common gap on service pages and it is where most quotes stall.

What it costs, at least as a range. This is the question people scroll hunting for, and giving it early makes the rest of the page readable.

How the process works, step by step, with timing. Who arrives, what they do, how long it takes, what you need from them.

Proof attached to the claim, not pooled in a testimonials section. A review about punctuality belongs beside the paragraph where you claim to be punctual.

Objections, handled directly. Why you cost more than the cheapest quote. What happens if something goes wrong. What if they need to reschedule.

And the action, repeated after each major section, because a visitor who is convinced at the halfway point should not have to scroll to the bottom.

What does not belong: your founding story at the top, awards nobody recognizes, and paragraphs written to hit a word count. That last one shows, and it is one of the signs a website was built for the builder rather than the business.

There is one structural finding worth importing here, with its limits stated. The Baymard Institute, which has tracked checkout flows since 2012, concluded after more than a decade that the number of steps matters far less than the number of fields a user has to consider. That is ecommerce checkout research from a company that sells UX research, and it is not about service pages. The transferable idea is that splitting content across more screens does not reduce the work; it just hides it. A long page that answers everything beats four short ones that each answer a quarter.

How do I know if my page is too long or too short?

Watch what people actually do rather than guessing.

Scroll depth tells you where attention stops. If most visitors stop at the same point, the section above it is where you lost them. That is not always a length problem. Often it is a boring section that could be cut, which makes the page shorter and better at once.

Recordings tell you more. The first thirty session recordings will show you people scanning back and forth hunting for something, which means the answer is missing or badly placed.

Your inbox tells you most of all. Every question a lead asks on the phone is a question your page did not answer. Write those down for a month and you have your section list.

What will not tell you is a split test. At the traffic a local business has, the arithmetic does not allow it. The standard two proportion sample size formula, at 5% significance and 80% power, is 16 times your conversion rate times one minus your rate, divided by the square of the difference you want to see. At a 3% rate, catching a fifth more inquiries needs roughly 12,900 visitors per version. Two versions of a service page is nearly 26,000 visits, which is years at local volume. Decide page length on the question list, not on an experiment you cannot finish.

Then check whether the page is doing its job at all, because healthy traffic sitting next to flat inquiries is the situation these edits are meant to fix, and you need the before number to know.

One caution. When you rebuild a page, the tracking often breaks with it, so make sure your conversion tracking survives the change before you start moving sections around.

What to do this week

Pull your last twenty inquiries and list every question a lead asked you before booking. Group the duplicates and rank them.

Open your main service page and mark which of those questions it answers. The unanswered ones are your new sections, and they go in the order the list ranks them.

Run the six step weight audit above before anyone cuts a word for speed. If the words are a fraction of a percent of the page, the argument is over.

Then cut one thing anyway. Almost every service page has a section about the company's passion for quality that could be deleted with no loss. Removing it makes room and costs nothing.

Add the phone number after each major section while you are in there.

Do this before you increase spend. If you are still deciding whether to run Google Ads at all, the page that answers buyer questions is the thing that determines whether the answer is yes.

Questions about your service page?

Email me at eric@seod.com with a link to your main service page and the three questions your leads ask most often on the phone. I will tell you which ones the page already answers, which it does not, and where the missing answers should go in the order of the page.

It takes me a few minutes, there is no call attached, and if the page is already complete I will say so rather than inventing work.

Otherwise, keep reading the conversion library.