Summarize with AI
If you sell one product, you do not need both. For a single-product SaaS startup, the homepage is the product page, and a separate /product page that repeats it makes both pages weaker. Add pages for the product only when a feature is big enough to deserve its own page, and give each one a name a buyer would search for.
This matters because a stranger gets one pass at your site. You know which page holds the real explanation, since you wrote both. A visitor who clicks Product in your header expects something new. If they get the homepage again in a different order, they have spent a click and learned nothing, and the page they needed does not exist yet.
A /product page that reshuffles the homepage's three screenshots
Picture an example: an AI tool that reads support tickets and turns them into bug reports. It connects to Zendesk and Intercom, groups the tickets that describe the same bug, and files one Jira issue with the steps to reproduce already written.
The header reads Product, Pricing, Customers and Blog, then Sign in and a filled button that says Connect Zendesk. The homepage hero reads Turn support tickets into bugs your engineers can fix. Below it sit three screenshots. The first is a card titled CSV export drops the last column, with 37 tickets this week and a button labelled Create Jira issue. The second is the Jira issue it files, with Steps to reproduce filled in. The third is a Monday Slack message headed Top 5 bugs by ticket count.
Now click Product. The headline reads The AI feedback platform for product teams. The same three screenshots appear again, this time with the Slack message first and the CSV card last. Under them is a grid of six cards: Smart clustering, Duplicate detection, Jira sync, Slack digests, Priority scores and SSO. Each card has one line of text and no link. At the bottom is the same Connect Zendesk button.
In the analytics, /product sits near the bottom of the page list, below the blog post about the seed round. Nearly all its visits come from the header link, and most of those visitors go to Pricing next or leave. In Search Console, the two pages trade places for the company's own name from week to week.
Why the founder built it, and what he does instead
The site started from a template, and the template had a Product slot. Every SaaS site he admires has one too, so leaving it out looked like a gap. He has rewritten the /product headline twice this quarter. Each version is broader than the homepage and says less about tickets.
On sales calls he never opens it. He shares the homepage, scrolls to the Jira screenshot, and then answers questions out loud. Which Jira fields does it fill? Does it work with GitHub Issues? What happens to a ticket in Spanish? He answers each one in a sentence. None of those answers is on any page.
That is the real gap. The site has two pages telling the same story, and zero pages for the questions buyers actually ask. He knows the product well enough to answer anything on a call. A stranger reading the site at night gets the same three screenshots twice.
What a copy of the homepage costs you
Most website advice treats a product page as standard kit, next to Pricing and About. For a company with one product, that advice is wrong. A product page only earns its place when it says something the homepage cannot hold.
The first cost is search. Google groups pages whose main content looks the same and picks one to show, a process it describes in its guide to canonicalization. The same guide notes that duplicate pages can confuse people, who wonder which one is the right page. So the two pages compete for one slot, and you do not get to choose which one wins.
The second cost is the visitor. Every visit was paid for, in ads, in content or in your own hours posting on LinkedIn. A click on Product that returns the homepage spends some of that visitor's patience on nothing. The ones who were close to signing up are the ones most likely to click it, because they wanted more detail.
The third cost lands six months later. The homepage gets the new screenshot after a redesign. /product keeps the old one, because nobody remembers it exists. Now the two pages disagree about what the product looks like, and a buyer who sees both trusts neither.
Three tests a feature has to pass before it gets a page
The fix is not to delete the idea of product pages. It is to build them one feature at a time, and only for features that pass at least two of these three tests.
- The search test. Someone types the feature into Google without your brand name. Check your Search Console queries and your call notes for the words buyers use. In the example, people search for a Zendesk to Jira integration. Nobody searches for smart clustering.
- The proof test. The feature has its own input, its own output and a screenshot that is not on the homepage. It also needs more explaining than one homepage section can hold, such as setup steps, limits and what it does not do. The Jira sync has a field mapping screen and a list of what lands in each field.
- The link test. You would send this page as a link to answer one question from a sales call. Which Jira fields does it fill? Here is the page.
In the example, Jira sync passes all three, so it gets the first page, at /integrations/jira, with a title that says Zendesk to Jira. Smart clustering fails the search test but is the core of the product, so it stays as the main homepage section. SSO passes none. It belongs in one line on the pricing page and one line in a security FAQ.
A single-product site often ends up with no feature pages for its first year, and that is fine. The homepage carries the whole story, and what makes a good SaaS website is mostly about making that one page clear. A second product is a different problem with different pages, covered in what to change on your website for a second product.
How Linear splits one product across pages
Linear sells one product and still has many product pages, and the way it does it is worth copying. Its homepage walks through the product in sections. Several sections end with a Learn more link to a page of their own: Intake, Plan, AI and Build. The footer lists more, such as Insights, Asks and Customer Requests. Each of those pages has its own title and its own description, and none of them is a second copy of the homepage.
The order matters. The homepage tells the whole story in short. Each feature page goes deeper on one part, for the visitor who already cares about that part. The homepage section is a promise, and the feature page is where it is kept.
The opposite case works too. In our website for SATHI, an affiliate platform for Shopify brands, the header shown in the case study reads Demo, Compare and Pricing, with no Product link. The features, from cookieless tracking to tiered commissions, sit on the homepage as numbered sections. The site launched seven weeks after kickoff.
The header when there is no product page
If /product goes, the Product link goes with it. Do not point it at a homepage anchor either. A link that scrolls the page you are already on feels broken to most visitors.
Keep the header to links that lead somewhere new: Pricing, Customers, Docs, Changelog. Once two features have pages, bring Product back as a menu that names them, such as Jira sync and Slack digests. Do not add an Overview item at the top. Overview is the homepage. The longer argument for a short header is in our post on SaaS website navigation.
This week: fold /product into the homepage
- Open the homepage and /product side by side. List every block on /product and mark the ones the homepage already has.
- Check what /product earns. Look at how many visits start on it and which search queries reach it. If both are close to zero, nothing depends on it.
- Move anything unique to the homepage, or into a note for a future feature page. Often that is one card worth keeping.
- Redirect /product to the homepage with a permanent redirect. Google says a permanent redirect shows the target page in search results, so the homepage keeps the slot.
- Remove Product from the header and run every link once to check nothing points at the old page.
- List your features and run the three tests. If one passes two of them, write it a page with its own screenshot, its own input and output, and a title in the searcher's words.
Where Studio Maydit fits
If you would rather hand the site to a studio, start by comparing website design agencies for B2B AI SaaS companies.
A header full of pages that say the same thing tells a stranger the company is smaller than it is. Studio Maydit designs websites and products for AI founders in the US, UK and Europe, and builds them in Framer, Webflow or custom code. A fixed-scope project takes three to four weeks and ends with a diagnosis of what is leaking. If your Product link leads back to your own homepage, book a 30 minute call.
