I tested three different landing pages for an avocado slicer. One approach pushed the sale, another appealed to people interested in cooking, and another emphasized how the page looked. We allocated budget across the pages and kept testing. The ugliest one won.
It had a simple package offer. I thought that simplicity helped, but I still can't tell you exactly why it beat the page I expected to do best. The campaign generated roughly $600K in sales in under 60 days. That result came from the offer, advertising and repeated testing together. It doesn't tell you what an ugly page will do for your business.
A landing page is a presentation made vertically. The visitor can leave at any screen, skip to the price, or scan the whole thing before deciding whether to read. Each section has to give them a reason to continue, and the next action has to make sense for the person who arrived.
A feature grid might explain something a buyer needs to know. It might also occupy a screen without answering the question that stopped the buyer from purchasing. The page's job gives you a way to tell those apart.
Start with the click that brought someone there and the commitment you're asking them to make.
Choose an offer that fits the traffic
The preceding message sets an expectation
An advertisement for a useful guide and an email about booking a call promise different next steps. Sending both to the same general page can break the connection between what prompted the click and what the page asks for.
I start with People: who the page is for, what problem brought them here, and what better outcome they're looking for. Product and Process follow. The offer might be a free resource, a paid audit, a strategy call or a direct purchase. Its scope has to fit the relationship and the problem.
A traffic-source label gives you somewhere to start. It doesn't prove how much the visitor knows. A search can come from someone comparing suppliers or someone trying to name a problem. Retargeting can reach someone who barely glanced at a product. Read the preceding message and the behavior behind the visit before choosing the ask.
These are possible pairings to investigate, rather than a requirement to build four pages for every product.
| Arrival | Possible next step | What the page needs to establish |
|---|---|---|
| Cold paid social | A useful free resource | What the resource contains and why it's relevant |
| A problem-led search | An audit or assessment | What will be examined and what the buyer receives |
| An email about a strategy call | Book that call | What the conversation covers and who it suits |
| A return visit to a known product | Purchase or trial | Terms, remaining objections and what happens next |
A useful audit can precede a larger engagement
The accounting work shows what a smaller commitment can accomplish. Prospects were wary of a large retainer before they'd seen useful work. We offered a $49 diagnostic audit: a structured review of their books and systems, followed by written findings and recommended fixes. The landing page explained an offer the firm could deliver before asking for a larger engagement.
The funnel included Facebook Ads and email follow-up as well as the page. It produced roughly $80K in collected revenue in about 60 days, much of it from audit buyers who went on to larger service packages. Those are the funnel's results. The record doesn't isolate the landing page's contribution.
An audit purchase was a meaningful step in that relationship. A form submission alone would have told us less about whether the firm was winning suitable clients. Decide what happens after the page action while you're deciding what the action should be.
Make the first screen useful on a phone
The scan should explain the offer
The opening screen should explain what you're offering, who it's for and what someone can do next. The headline can name a situation or an outcome, but the supporting copy still has to say what the buyer is being offered. A visitor shouldn't have to translate a slogan into a service.
Check the promise against the advertisement or email that led to it. A page can be clear in isolation and still disappoint someone who clicked for something else. Keep the language and offer recognizable across that transition.
My working approach is mobile-first and skimmable. I think about each screen as a part of the presentation. Someone scanning on a phone may look for evidence that it's worth a closer read; someone at a desktop may open supporting material and inspect the detail. Those are useful behaviors to design for, not rules about what every phone or desktop visitor does.
Give the scan something to communicate. Headings should carry the argument, an image should explain the offer or its result, and proof should sit near the claim it supports. A deeper read can then answer the questions the scan raised. Making the page skimmable doesn't require deleting the explanation.
The form and navigation need a purpose
Open the page on a phone and try the actual action. Can you read the terms, complete the form and understand an error? Does important evidence appear without waiting for decorative elements? Check desktop too. A form that looks short in a wireframe may still be awkward to complete.
Navigation deserves the same purpose test as everything else. A full-site menu can introduce unrelated destinations before the visitor understands the offer. Remove those distractions where they don't serve the page's job. Keep information needed to judge the purchase, including relevant evidence and policies. Counting links alone can't distinguish a distraction from an answer.
One primary call to action gives the page a clear destination. You can repeat that action at useful decision points so a prepared visitor can take it without searching. The explanation still needs to work for someone who has more questions.
Answer the buying questions before asking for a commitment
A page review can follow five questions: is this relevant to me, does the seller understand the problem, does the proposed approach make sense, what risk am I taking, and what happens when I say yes? They give each section a purpose. They also help distinguish an explanation the buyer needs from material the business happens to have available.
The sequence is a starting point. A returning buyer may already have answers to the first questions. Someone comparing suppliers may go directly to the terms. The page should let those visitors act while giving an unfamiliar buyer enough information to judge the same offer.
Name who the offer can help
Relevance begins with a specific situation. Name the customer, the problem or the outcome clearly enough that a visitor can tell whether the offer applies. A broad promise to grow a business leaves the reader to work out what kind of work is being sold and whether it addresses their problem.
Qualification belongs here too. Describe the work's scope and who will get value from it. A page that attracts inquiries for work the business can't deliver creates a problem after the click, even if the form-completion rate looks good.
Describe the problem in the buyer's language
I collect the phrases buyers use in places such as competitor reviews, community discussions and support tickets. I call that collection a Lexicon of Pain. It gives the copy a basis in what people have described, rather than what the business assumes they ought to care about.
Keep the context with the phrase. A complaint about a hidden charge and a complaint about slow delivery need different answers. Copying both into a paragraph about general frustration would throw away the distinction that makes the research useful. A quotation must remain someone's actual words; a summary needs to preserve what the person meant.
Explain what has to change
I connect a buyer's problem to a desired future, then ask what has to change to reach it. The page needs to make that change concrete. Explain what the buyer receives, what the business does and what the buyer must contribute.
A capability can help explain that path. It earns its place when the reader can connect it to the result. A list of available services without that connection makes the buyer assemble the approach themselves. Scope and deliverables give them something more useful to evaluate, including the limits of what they're buying.
The terms explain the buyer's commitment
Risk reduction needs facts. State the price, scope and relevant conditions. Explain a refund policy or guarantee only when the business actually offers it, with the conditions that determine whether it applies. Put proof from a comparable situation beside the claim it supports, keeping the differences visible.
The cost of doing nothing can be relevant, but it has to come from the customer's circumstances. Invented losses and exaggerated consequences can't make an offer safer. A clear boundary may help more than another reassurance: what happens if the diagnosis finds something outside the service, or the buyer turns out to be a poor fit?
The button names the next action
Make the action and its consequences clear. A purchase, application and booked call are different commitments. The page and the following screen should agree about which one the visitor is making. Form fields should collect what the business needs for that step; qualification may justify more than a minimal email field.
Read each section against those questions. A team introduction, feature grid or animation can stay when it supplies a useful answer. Cut or change a section whose removal costs the buyer nothing. An unanswered question may also expose a weakness in the offer that more page copy won't fix.
Check the offer before adding another section
A keychain became the entry product
The premium menswear project had a page problem and an offer problem together. The brand sold denim in the $400 to $500 range. Its design-heavy site was converting cold traffic at under 1%. I handled conversion rate optimization and site work on that project, not the advertising.
We changed what a cold visitor was asked to buy. A small keychain became the entry product on a dedicated landing page, with the jeans introduced alongside it and through a bundle. Moving to Shopify Plus also allowed us to experiment with the checkout process. We simplified options and clarified the offer while keeping the brand's identity.
The conversion gain followed an offer and checkout change
Conversion climbed above 4% on cold traffic for the bundle flow. That result covers changes to the offer and checkout together. It doesn't establish that removing an animation or rewriting a headline caused the gain. The business was already successful, and I can't take credit for all its revenue.
The useful distinction is what the visitor had to decide. An unfamiliar buyer was being asked to commit to expensive jeans. The keychain offered a smaller entry into the same brand, and the bundle gave the jeans a different context. Another section praising the denim would have left the original buying decision intact.
That doesn't make a cheap entry product the right answer for every premium business. The smaller purchase has to lead to something the buyer values and the business can fulfill profitably. It can also attract customers who never want the main offer. Follow what those customers do after buying before calling the entry product a success.
The same caution applies to a page with a strong offer but a broken form or confusing checkout. Rewrite the part responsible for the obstacle. A diagnosis should identify what prevents a purchase, including conditions outside the copy, before turning into a request for a longer page.
The implementation choice follows from that diagnosis. You need a way to build the useful change without buying unrelated complexity.
Build custom when a template blocks a useful change
A template can carry a well-matched offer
A template supplies a layout and a way to publish it. The work of deciding what to say still belongs to the business. You can use a template to present a well-matched offer, and you can commission a custom page that explains the wrong one beautifully.
Start with what the page needs to do. Can the existing builder present the offer, supporting evidence and terms in a useful order? Can a visitor complete the intended action on their device? Can the business trace that action into the sale or service that follows? Reuse what can do those jobs.
Custom work becomes a candidate when a specific constraint prevents a useful change. The obstacle might be a checkout the team can't alter, a form that can't support necessary qualification, or a layout that makes essential evidence hard to understand. Name the constraint before commissioning the work. A general wish for a more distinctive page doesn't tell a developer what result would justify the cost.
The investment needs a business case
The investment depends on the economics of that change. Consider the traffic that could benefit, the value of a suitable customer, the margin left after fulfillment and the cost of building and maintaining the alternative. Keep the expected improvement separate from a measured result. A page with a high-value offer can still have too little traffic to establish whether a small apparent gain will last.
A template can be sufficient for learning whether anyone wants an offer. It may remain sufficient afterward. Evidence that the offer works doesn't automatically justify replacing the builder; it gives you a better basis for deciding which constraint is worth removing.
I use experiments to compare page alternatives. Research helps choose plausible options, but a preferred design is still a hypothesis. Spend the implementation effort on making a meaningful alternative possible, then compare what happens. Replacing a template without changing the proposition, the buying experience or your ability to measure either can leave the original question unanswered.
Review one arrival and follow it past the button
The review starts with an actual arrival
Choose an actual route into a landing page: an advertisement, a search result or an email that sends meaningful traffic. Keep that arrival in view while inspecting the page. A page for several genuinely different situations may need separate versions, but inspect the differences before multiplying pages.
The review should produce a specific change you can explain. Work through the arrival and purchase path in order.
- Write down what prompted the click, who the offer can help and what action the page asks for. Compare the offer on the page with the preceding promise. Resolve a mismatch before adjusting the layout around it.
- Open the page on a phone and a desktop. Try the form, calendar or checkout. Check instructions, required information and failure messages as well as the successful path. Separate a working-page defect from a question the copy needs to answer.
- Read only the headings and visible proof. Identify what a scanner can learn about relevance, approach, risk and the next action. Then read the detail to see whether the page supports that account or contradicts it.
- Find an unanswered buying question. Use customer-language research, sales objections or observed behavior to support the diagnosis. A low scroll depth can locate a point worth investigating; it can't tell you on its own why someone left.
- Describe a substantive alternative and the result that would make it useful. An offer change, a clearer explanation or a different checkout path tests something the business can learn from. Keep a record of what changed and compare the outcome beyond the button.
Measure what happens after the page action
That last step prevents a form-completion rate from becoming the whole definition of success. Evaluate a page asking for a call against suitable bookings and the sales process that follows. A purchase page needs to account for the resulting orders and their economics. Decide which outcome you're trying to improve before looking for a winning number.
Make comparisons under conditions you can interpret. Changing the traffic, offer and page together may produce a better campaign, but it won't isolate the page's effect. Limited data can leave the result uncertain. A test that doesn't support the proposed change is a reason to revisit the diagnosis; it doesn't guarantee the next version will win.
The upstream choice of offer and route belongs with Conversion Funnel Design. The ongoing experiment design belongs with Conversion Rate Optimization. The page connects those jobs by giving the visitor enough information to make the next decision.
The sales conversation continues after the click. Keep the promise, the confirmation and the actual delivery aligned. A page that wins a commitment the business can't fulfill has created another problem for the customer.