Adapte Dijital
Anasayfa
AINEO
Dijital Danışmanlık Dijital Denetim
Web & AI
Kurumsal
Paketler Blog

How to Rewrite Titles and Descriptions

Yayın Tarihi: 18 Ağustos 2026 Yazar: Adapte Dijital Kategori: Tips
How to Rewrite Titles and Descriptions — Adapte Dijital cover image
💡 Kısaca: On the results page, the title and description are not the site’s advertisement; they are the business’s only sales representative present at the scene.

On the results page, the title and description are not the site’s advertisement; they are the business’s only sales representative present at the scene. Everything invested in the page — content, speed, design — takes the stage only if those two lines win the click. And as the click pool shrinks, the auction for those lines hardens: five rivals on one screen, competing for the remaining clicks in the same three seconds. The oddity of this race is that most runners do not know they are on the track — they never wrote their shopfront and are racing with whatever formed on its own.

This piece binds the rewriting of those two lines to a one-page working routine: the formula, the constraints, the test window and the measurement.

WHAT

What Is the Problem?

BU BÖLÜMÜN ÖZETİ

  • The title reads like a filename
  • The description is a machine’s random pick
  • The two lines and the page tell different stories

Most sites’ shopfronts were never written; they accreted — and an accreted shopfront carries three diseases.

The title reads like a filename

“Our Services | Company Name” is a label, not a title. It states no difference and answers no question; it leaves “why should I pick you” for the user to answer. In the three seconds on screen, users do not reason — they recognise; a line offering nothing to recognise donates its three seconds to the neighbour below.

Most sites’ shopfronts were never written; they accreted — and an accreted shopfront carries three diseases.

The description is a machine’s random pick

Left empty, the description field gets filled by the engine with a fragment from the page — varying by query and usually a context-free middle sentence. It is a random interior page taped to the shop window: it sometimes fits, mostly does not. Filling the field does not fully reclaim the choice, but it earns the right to make an offer — and a page without an offer enters a raffle.

The two lines and the page tell different stories

The costliest shopfront is the lying one: the title promises, the page does not deliver. The click arrives, the trust departs — and the third layer of visibility failing to become sales opens exactly here. Winning the click rate while losing the conversion costs more than never being clicked, because the first spends only opportunity and the second spends reputation with it.

WHY

Why Does It Happen?

BU BÖLÜMÜN ÖZETİ

  • Nobody owns it
  • Unmeasured, it passes for cosmetic
  • The summary era enlarged the job; the habit stayed small

Shopfront neglect has three structural causes.

Nobody owns it

The title comes from a developer’s template, the description from a plugin default; no human wrote either with a customer in mind. The content has an owner, the shopfront does not — yet the customer reads the shopfront first. The fix is as small as it sounds: one name owning two lines — writing them, measuring them, holding the sole right to touch them.

Unmeasured, it passes for cosmetic

With the label’s effect on clicks unmeasured, the work gets filed under decoration. Install the measurement and the table flips: at an unchanged position, the rate can multiply — the deep-rank champions in the inventory tip are the standing proof.

The summary era enlarged the job; the habit stayed small

That AI summaries split the click is now a wide literature; the behavioural data relayed by Search Engine Land’s coverage is among the known examples. In the era of shared-out clicks, shopfront writing graduated from cosmetics to a revenue line — while habits remain calibrated to the age of plenty, when neglect carried a small fine; in scarcity, the same neglect gift-wraps the remaining clicks for the rival.

HOW

How Is It Done?

BU BÖLÜMÜN ÖZETİ

  • Panel 1: write the query and its intent
  • Panel 2: fill the title formula
  • Panel 3: write the description, open the window

One worksheet per page; three panels.

Panel 1: write the query and its intent

The one main query the page wants to win, and that query’s intent (information or decision), go at the top. The shopfront is written to the query, not to the page — on the screen, the user hunts for the echo of their own question. If the intent is unclear, the match test comes first; an intentless shopfront is a letter without an address.

Panel 2: fill the title formula

The title’s building blocks, in reading order: the query’s own language (the user’s word) + a concrete difference (number, scope, timeframe, region) + a reason to click — and the brand last, if room remains. The distance between “Plywood Prices 2026: Current List by 18 mm Sheet” and “Our Products | Company” is the whole formula. The order is deliberate: the user scans for their word first, the difference second, the brand last — a title assembled against the reading direction shelves the right parts in the wrong place. Length stays inside the band the screen will not crop; a cropped promise is half a promise.

Panel 3: write the description, open the window

The description has one job: enlarge the title’s promise with one proof and close with one action — what they will find, why it is credible, what they can do. Nothing absent from the page may be written; shopfront-shelf agreement is conversion’s pre-contract. Then the change date is noted, a four-to-six-week window opens, and no second change touches the page inside it; the comparison runs on the same query set against the prior period. With position steady, the rate’s movement is the shopfront’s report card — and an unnoted change is an unregistered exam: even a good result cannot be credited.

HOW

How Long, at What Cost?

BU BÖLÜMÜN ÖZETİ

  • Half an hour per page
  • The real cost is sentence discipline
  • The return arrives at zero extra traffic cost

The work scales with the inventory; the unit cost stays low.

Half an hour per page

Fed by one good example, the worksheet takes under thirty minutes. A ten-page repair round is a day’s work, run in the inventory’s priority order — the lost-click product.

The work scales with the inventory; the unit cost stays low.

The real cost is sentence discipline

The difficulty is not writing but shortening: if the difference will not fit one sentence, the difference is usually not yet clear. Shopfront work therefore yields a by-product — it confronts the business, page by page and without escape routes, with “why should they pick us”.

The return arrives at zero extra traffic cost

Raising the rate at an unchanged position wins clicks without entering the ranking war; every click gained is the collection of visibility already paid for. In budget terms this queues ahead of producing new content.

THE

The Common Mistake

BU BÖLÜMÜN ÖZETİ

  • Sliding into click-baiting
  • Changing everything on the same day
  • Writing without measuring

The shopfront gets written; three reflexes spoil it.

Sliding into click-baiting

An inflated promise lifts the rate briefly and sinks return rate and trust durably. The shopfront is the page’s best true sentence — not its brightest imagined one. A title the page cannot honour is advertising that works for the competitor.

The shopfront gets written; three reflexes spoil it.

Changing everything on the same day

Relabelling the whole site in one pass makes it forever unknowable which change worked. Repair moves in batches with windows; the formula each batch teaches travels to the next.

Writing without measuring

A repair never compared at window’s end sits in the same ledger as a repair never made. A flat rate is also a finding: the fault was not the shopfront, and the next suspect is intent — work without a finding is only fatigue.

FREQUENTLY

Frequently Asked Questions

Sık Sorulan Sorular

What if the engine rewrites my title?

The rewrite is a signal: the engine judged the title insufficiently tied to the query. Carry the query’s language more directly into the title and check its harmony with the page’s main heading; a harmonious pair is retained in most cases.

Is one template across hundreds of pages acceptable?

As a skeleton yes, as a sentence no. A template whose variables — product, region, number — fill from real data works; a hundred copies of the same sentence turn the shopfront back into a label. Rule: one template, page-specific proof.

Test the title first or the description?

Title first; it carries most of the rate. The description converts the glance the title won into the click — the order is the glance-to-click order. Changing both in one window makes their contributions inseparable.

Bu Konuyla İlgili Diğer İçerikler

TREN