Landing Page Optimisation Strategy: What to Fix First and When to Rebuild

A landing page optimisation strategy answers three questions before anyone touches a page: which page gets attention first, whether that page needs small changes or a full rebuild, and how much effort the likely return justifies. The quickest way to answer all three is to score every page on traffic, conversion value, and fixability, then work down the list. A page with heavy traffic, high-value conversions, and an obvious defect always comes first.

Notice what is missing from that answer: nothing about headlines, button colours, or form fields. Those are process decisions, and they only matter once you know which page deserves the work. This post stays on the decision layer, the same layer we start from in our own conversion optimisation projects. For the audit, hypothesis and measurement cycle that follows once a page is chosen, our step-by-step landing page optimisation guide covers the full process.

Score Every Page on Traffic, Value and Fixability

You cannot optimise everything at once, and you should not try. Give each landing page a score from 1 to 5 on three factors, multiply the three numbers together, and rank the results. The arithmetic takes twenty minutes and settles arguments that meetings never do.

• Traffic: how many people actually see the page each month. Pull the real number from analytics rather than guessing, because teams routinely overestimate traffic to pages they care about.
• Value: what one conversion on this page is worth to the business, in pounds where possible.
• Fixability: how confident you are that you can name the page's problem and change it. A broken mobile form is a 5. A vague sense that the page "feels weak" is a 1.

A worked example makes the point. Imagine a plumbing firm with two pages. Its quote request page gets 1,200 visits a month, an average job is worth £1,400, and the mobile form cuts off below the fold: call that 4 x 5 x 5, a score of 100. Its newsletter page gets 3,000 visits, but a subscriber is worth very little for months: 5 x 1 x 3, a score of 15. The busier page loses. Traffic alone is a poor compass, and following it is the most common strategic mistake we see.

Where the Numbers Come From

None of this requires special tooling. Traffic comes straight from your analytics, filtered to the last 90 days so a one-off campaign spike does not distort the ranking. Value comes from your own records: average order total for e-commerce, or average job value multiplied by the rough share of enquiries that become customers for a service business. An estimate within 30 per cent is fine, because the scores only need to separate pages, not price them.

Fixability is the one you assess by hand. Open each page on your own phone, on mobile data, and try to complete the action a visitor would. Time the load, submit the form, read the offer cold. Ten minutes per page tells you whether the problem is nameable or vague, and nameable problems are where optimisation budgets earn their keep.

Value Is the Factor Teams Get Wrong

For e-commerce, value is easy: it is the order total. For service businesses the conversion is usually a lead, and a lead is only worth something if it turns into work. A page generating fifty enquiries a month from people outside your service area scores lower on value than a page generating eight enquiries from qualified buyers. Judge lead pages by what a qualified lead is worth, not by raw form completions. Offer, form design and follow-up all shape that quality, and our lead generation landing page guide treats them as one connected system.

Iterate or Rebuild: Four Signs a Page Is Past Patching

Once a page reaches the top of your list, the next strategic call is the size of the intervention. Default to iteration. Small changes are cheaper, they compound, and each one teaches you something about your audience that a wholesale replacement would erase.

Rebuild instead when you see one or more of these signs:

1. The offer changed but the page did not. If you now sell fixed-price packages and the page still pitches hourly consultancy, no headline tweak closes that gap.
2. Conversion sits near zero despite steady traffic. A page converting at 0.2 per cent rarely climbs to 3 per cent through element tweaks. Something foundational is wrong, usually the match between what visitors wanted and what the page offers.
3. The structure fights the goal. A brochure page repurposed as a landing page carries full site navigation, competing messages and no single action. You would have to remove more than you keep.
4. The technical base resists change. If every edit needs a developer, breaks the mobile layout, or takes a week to ship, the cost of iteration quietly exceeds the cost of starting fresh.

The deciding test is accumulated cost. Add up the hours you have spent patching the page over the last quarter and the hours the next round will take. Our landing page design service builds a single focused page for £595, and it is common for that figure to undercut what a business has already spent nursing a page that was never designed to convert.

A rebuild should not throw away what the old page taught you. Carry over any element with evidence behind it, a headline that lifted enquiries or a testimonial visitors mentioned on the phone, and treat the old page's conversion rate as the baseline the new one must beat in its first full month. If the replacement cannot beat a page you judged beyond saving, the diagnosis was wrong and the numbers will say so quickly.

When Optimisation Is Premature

The least discussed strategic decision is choosing to wait. Optimisation is a learning exercise, and learning needs data. A page receiving 80 visits a month produces a conversion signal so faint that almost any conclusion you draw from it is noise dressed up as insight.

Our working rule: below a few hundred visits a month per page, skip formal experiments entirely. Split testing in particular demands far more traffic than most small business pages receive, and our landing page A/B testing guide is honest about the sample sizes involved before a result means anything.

Low traffic does not mean doing nothing. It means limiting yourself to changes that need no measurement to justify: repairing broken links, cutting load time, shortening a ten-field form, correcting an out-of-date price. We keep a collection of these in our landing page conversion tips, and they pay off at any traffic level. Beyond that, a low-traffic page's real problem is traffic, so the budget belongs in search visibility or advertising until there are enough visitors to learn from.

Budget Effort Like a Portfolio, Not a Rota

Spreading attention evenly across ten landing pages guarantees that none of them improves meaningfully. Weight the effort instead. The one or two pages that top your scoring get the large share: research, redesign work if the signs above apply, and structured changes with proper follow-through. The middle of the table gets light maintenance, an hour or two a month to keep content accurate and forms working. The tail gets a deliberate decision, not neglect by accident.

Make the split concrete against your real capacity. A business with eight hours a month for this work might spend six on the top-scoring page, ninety minutes checking the middle tier, and thirty minutes reviewing whether any tail page should be retired. That feels unbalanced, and it should. Balanced effort across unequal pages is how a whole quarter passes with no page measurably better than before.

For tail pages, choose one of three options. Retire the page and redirect its URL to the closest relevant survivor. Merge it into a stronger sibling if it holds content worth keeping. Or leave it alone, consciously: a page that costs nothing to keep and cannibalises nothing is not hurting you, and pouring hours into it would be.

Whatever the allocation, document what each change taught you. A short log of what was changed, why, and what happened afterwards builds knowledge about your specific audience that no general advice can replace, and it stops new team members repeating experiments that already failed. Rescore the full set quarterly, because campaigns, seasons and new competitors shift traffic and value faster than most teams expect.