Why this matters
When a page gets visits but few enquiries, the instinct is to redesign it, which replaces one set of guesses with another. Annotating comes first: read the page as a sequence of sections, each with one job, and mark where the sequence breaks. "The page doesn't work" becomes specific, checkable problems.
This exercise applies the guide to the anatomy of a landing page and the discipline of the conversion rate optimisation hub: an annotation produces problems to investigate, not changes to ship. The page's CVRShare of visits or clicks that convert., as defined in the glossary entry on conversion rate, is the number every problem on your list should explain.
The landing page
Sunbeam Solar is a fictional installer of solar panels and home batteries. Its search ads send steady traffic to a quote request page that produces very few requests. Here is the page as it appears on a phone, top to bottom, with last month's figures.
Sunbeam Solar: the ad, the page and last month's figures
The most-served ad reads: "Solar Panels Installed Locally · Free Quote Within 24 Hours · Battery Options".
| Order | What the visitor sees on a phone |
|---|---|
| 1 | Logo and the full site menu: Home, About, Products, Blog, Careers, Contact |
| 2 | Headline over a stock photo of a sunset: "Powering a brighter tomorrow" |
| 3 | Line under the headline: "A family business since 2009, proud of our values" |
| 4 | Button: "Learn more", which scrolls to the company history |
| 5 | Company history: three paragraphs about the founders and the first van |
| 6 | Eight technical bullets: panel efficiency, inverter type, monitoring app, mounting system and so on |
| 7 | FAQ accordion, collapsed: "What does it cost?" (answer: "Every home is different, contact us"), "Do I need planning permission?", "How long does fitting take?" |
| 8 | Quote form with ten fields, including electricity meter number, annual usage in kWh and roof direction; button reads "Submit" |
| 9 | Carousel of three reviews: "Great job!" signed "J.", "Very happy" signed "Mrs P.", "Recommended" with no name |
| 10 | Footer with the full menu again, a newsletter sign-up and links to the blog |
| Measure | Last month |
|---|---|
| Visits from search ads | 2,400 |
| Visits on a phone | 1,680 (70%) |
| Visitors who scrolled as far as the form | 480 (20%) |
| Visitors who typed in the first form field | 120 |
| Quote requests submitted | 31 |
| Conversion rate (requests ÷ visits) | 31 ÷ 2,400 ≈ 1.3% |
Labelling each section
For each row, ask which part of a landing page it is meant to be (some rows are none), what its job is, written as the visitor's question it should answer, and whether it does that job: working, weak or missing, with a reason. Then list the parts of a good page that do not appear at all; a missing guarantee is a finding just as a weak headline is. Finally, read the figures as a small FunnelModel of the customer journey stages.: visits, reaching the form, starting it, submitting it. The biggest loss shows where the annotation matters most.
Task
- Copy the page table and add three columns: the anatomy label, the job, and your verdict with a reason.
- List the parts of a good landing page that the page does not have at all.
- Compare the headline and first screen with the ad and say whether they continue its promise.
- Work out where the funnel loses the most visitors and say which sections that points to.
- Write the problems you would investigate first, in order, with the evidence you would gather to confirm each before changing anything.
Worked answer
Sunbeam Solar: annotated sections
| Order | Anatomy label | Job for the visitor | Verdict |
|---|---|---|---|
| 1 | None; site navigation | None; every link is an exit | Weak: six exits before the offer |
| 2 | Headline | "Is this what I clicked for?" | Missing: a slogan; no panels, fitting or quote |
| 3 | Subhead | "Is this for someone like me?" | Weak: about the company; true of any firm |
| 4 | Primary call to action | "What do I do now?" | Missing: "Learn more" leads away from the quote |
| 5 | None; brand story | Belongs after the offer, if at all | Weak: takes the most valuable screens |
| 6 | Benefits | "Why does this matter to me?" | Weak: features, not outcomes such as lower bills |
| 7 | Objections | "What is the catch?" | Weak: right questions, collapsed; price unanswered |
| 8 | Form | "Is this worth the effort?" | Weak: details people lack to hand; "Submit" |
| 9 | Proof | "Why should I believe you?" | Weak: anonymous, generic and below the form |
| 10 | Footer | "Is this a real company?" | Weak: more exits; no address shown |
Missing altogether: a statement of the offer (what the free quote involves and when it arrives), a "how it works" strip, a guarantee or risk reducer, and any credential or accreditation.
Funnel: 2,400 visits → 480 reach the form (a loss of 1,920) → 120 start → 31 submit. The largest loss is before the form, so the top of the page matters most.
Message match. The ad promises local installation, a fast free quote and battery options; the first screen mentions none of them. The page fails the recognition test before anything else can work.
Ranked problems to investigate.
- First screen does not continue the ad. Confirm with phone scroll data and the share who leave without scrolling.
- No call to action near the top; the only button leads away. Compare clicks on "Learn more" and the menu with clicks towards the form.
- The offer and price are never stated. Run an on-page poll and read the sales team's call notes.
- The form asks for details people do not have to hand. Add field-level tracking to find where typing stops.
- Proof is anonymous and out of sight. Check for attributed reviews on an independent profile.
Each item is a problem with a location and a way to confirm it, not a fix.
Common mistakes
- Annotating on a desktop. Most visitors use phones, where the history pushes the form far down.
- Labelling without judging. A label says what a section is; the verdict says whether it works.
- Missing what is absent. The missing offer and guarantee matter as much as weak sections.
- Jumping to fixes. A new headline written now skips the evidence.
- Starting with the form. Most visitors never reach it.
Return to the beginner conversion optimisation level page to continue, or test yourself in the checkpoint on conversion rates and landing page anatomy.
