SEO for iGaming affiliate sites in 2026

Most SEO advice for affiliate sites starts with keywords. That ordering is wrong, and it is expensive. Before a page can rank for anything it has to be found, fetched and stored. On a new gambling-adjacent domain, almost all of the loss happens in those three steps — not in the competitive part everyone worries about.

When a site is finally audited URL by URL instead of trusting a dashboard summary, the finding is usually blunt: the majority of the URLs in the sitemap are not in Google's index, and a large share have never been fetched at all. Nothing about that was a ranking problem. It was a discovery problem wearing a ranking problem's clothes.

This guide is written in the order the bottleneck actually appears: discovery, indexing, then the parts that decide position. It pairs with our guide to affiliate content in the AI era, which covers what a page has to contain once it is indexed.

The order nobody teaches: discovery, then indexing, then ranking

Search works as a funnel with four gates. A page has to pass each one before the next matters.

Google documents the first three in its Search Console guidance and the page indexing report, which separates "Discovered — currently not indexed" from "Crawled — currently not indexed". Those two states have completely different fixes, and they are the two states a young affiliate site lives in. Treating either as a content-quality problem sends you off rewriting pages that were never fetched.

Illustrative arithmetic. Say you publish 60 pages in your first quarter and 40 of them are never crawled. You could double the quality of all 60 and your traffic ceiling would still be set by the 20 that exist as far as Search is concerned. Fixing discovery is a structural change with a multiplier; rewriting is linear.

Why most of a new site's pages are never fetched at all

Crawl attention is rationed. A domain with no history, no links and no traffic gets very little of it, and gambling-adjacent domains start from a worse position than most because the category is saturated with thin, near-identical pages. Google's own guidance on managing crawl budget is explicit that crawl demand follows popularity and change signals, not your publishing schedule.

The first-hand version: on two of our domains, an entire day of verified Googlebot activity — checked against the reverse-DNS method Google publishes for verifying Googlebot — consisted of nothing but requests to the robots file. Not one HTML page. Everything we had shipped that week was, to Search, invisible. No amount of on-page work moves that. Links and a reason to return do.

The three ways Google finds a page, and why a sitemap is only a hint

• Links from pages it already crawls — the strongest route, and the only one you fully control. This includes your own internal links.
• External links — someone else references your URL. Slow, and the only legitimate version is being worth referencing. Bought links, paid guest posts and link exchanges are named directly in Google's spam policies under link spam; we do not use them and do not recommend them.
• Sitemaps — a hint, nothing more. Google's sitemap documentation states plainly that submitting a sitemap does not guarantee crawling or indexing.

The practical consequence: a URL that appears only in your sitemap is a URL you have asked about politely, once. A URL linked from three pages you update regularly is a URL you keep asking about. Keep the sitemap honest too — indexable URLs only, real modification dates. Padding it with pages you have set to noindex, or bumping dates to look fresh, teaches the crawler that your signals are noise.

Internal links are the discovery mechanism, not a bonus

Internal linking is usually taught as a way to move authority around a site. On a new site that is the secondary benefit. The primary one is that an internal link is how a page becomes findable at all — which is why Google's guidance on making links crawlable insists that a link is an anchor element with an href attribute. A button that runs JavaScript, a card wired to a click handler, a filter that swaps rows on the client: none of those are links.

Three rules worth applying on every site:

1. Every published URL gets at least one link from a page that is itself crawled — normally a category hub that lists its children in full, server-rendered.
2. Anchors describe the destination. "Read more" tells a crawler and a reader nothing; "how to get approved by affiliate programs" tells both.
3. Hub pages list everything, not the first page of results. If a listing paginates or hides rows behind a control, the hidden rows need their own crawlable path.

The orphan-page audit you can run in an afternoon

An orphan is a URL in your sitemap with zero internal links pointing at it. They accumulate in the dozens on a mid-sized site, mostly articles whose category hub quietly stopped listing everything after a template change. The audit is mechanical:

1. Export the URL list from your sitemap.
2. Crawl your own site from the homepage, following only anchor elements with href attributes, and collect every internal URL you reach.
3. Subtract the second set from the first. What remains is orphaned.
4. For each orphan, decide: link it from a hub, or remove it from the sitemap. Both are valid. Leaving it in limbo is not.

Run the subtraction the other way as well, to catch the opposite fault — pages that are linked and crawlable but missing from the sitemap. Our content operations guide covers how to stop this from re-breaking every time you publish.

Rendering: if a crawler cannot see your list, the list does not exist

This is where affiliate sites lose the most silently, because the page looks perfect in a browser. Comparison tables, filter bars and "show more" reveals are all client-side patterns, and Google does not click anything. Its JavaScript SEO basics documentation describes rendering as a deferred, resource-limited stage — not a guarantee that whatever your interface can eventually produce will be seen.

This fault is easy to ship without noticing. One of our comparison pages rendered its first twenty rows server-side and loaded the rest on interaction. Every operator past row twenty — the majority of the table — was present for readers and absent for crawlers, along with its links, its offer text and its outbound anchor. The fix was structural: render every row into the HTML, and use a native details element for the reveal so the content sits in the DOM whether or not anything is clicked.

Three checks worth running on any page you care about:

• Fetch the URL with JavaScript disabled, or from the command line, and search the source for the last item in the list. If it is not there, it is not indexed.
• Count the rendered rows against the number of records in your data file. A mismatch is a bug, not a design choice.
• Confirm the outbound affiliate links exist in the server HTML with their rel attributes intact — our guide to affiliate tracking and postbacks covers how those links should be built and logged.

What actually makes a page worth indexing on a gambling-adjacent site

Once a page is fetched, the question becomes whether storing it adds anything to the index. Two of Google's spam policies decide most affiliate cases, and both are worth reading in full rather than in summary. Thin affiliation is defined as publishing content with affiliate links "where the product descriptions and reviews are copied directly from the original merchant without any original content or added value". Scaled content abuse is "when many pages are generated for the primary purpose of manipulating search rankings", including "creating large amounts of unoriginal content that provides little to no value to users, no matter how it's created".

Read those two together and the affiliate trap is obvious. A page per operator built from the operator's own bonus copy fails the first. A page per operator multiplied by country, generated from a template, fails the second and the doorway policy at the same time. We do not build program-by-country matrices, and a doorway page is one of the few mistakes in this business that is genuinely hard to recover from.

What passes, in our experience, is a page containing at least one thing that could not have been assembled from the merchant's own site: a clause we read in the terms, a calculation we ran, a decision we made and what came of it. Google's helpful content guidance phrases the same test as a question — does the content "clearly demonstrate first-hand expertise and a depth of knowledge". On most affiliate pages the honest answer is no, and that gap is the opportunity. Our affiliate program directory is built to that standard: facts taken from the terms, our own reading of them, and an explicit note where we do not know something.

Titles and descriptions: the one on-page property tied to traffic

Titles and meta descriptions are the on-page elements Google ties most directly to how your result is presented, in its documentation on title links and snippets. They will not rescue an unindexed page, but they change the click-through rate of every page that is indexed, which makes them the cheapest work on this list.

The most common failure here is mechanical truncation: descriptions generated by slicing the first 155 characters of the article. Every one of them ended mid-clause. Hand-writing a hundred of them is a day of work, and it is worth the day.

Timelines: what months 1, 3, 6 and 12 realistically look like

These are the ranges we would give a new affiliate, based on launching new domains ourselves. They are estimates, not promises, and a competitive English-language market will sit at the slow end of every row.

Two honest notes. First, impressions arrive long before clicks, and early impressions are often junk queries with nothing to do with your topic — for a new site, most of what looks like early traffic is query spam and bot activity rather than readers. Read the query list before celebrating the chart. Second, "no traffic in month three" is not evidence of failure; "no indexed pages in month three" is. Those are different diagnoses with different fixes.

If you want to turn a traffic assumption into a revenue estimate before committing a year to it, our traffic to revenue estimator makes the funnel explicit: visits, click-through, signup rate, deposit rate, value per depositor. It usually shows that the traffic number people plan around is an order of magnitude too small.

What we stopped doing after measuring it

• Publishing before linking. An article not linked from a hub on the day it publishes is an orphan by default. The hub link is now part of publishing, not a follow-up task.
• Shipping client-side-only lists. Any list that matters goes into the server HTML. Interactivity is layered on top and never load-bearing.
• Chasing structured data for rich results. Shipping a great deal of markup on an affiliate site rarely produces a rich result, because most page types in this category have no eligible result type at all. We kept the markup that describes the page honestly and stopped treating it as a traffic lever.
• Counting pages as progress. Page count is an input. Indexed, linked, differentiated pages are the output. Track the second and ignore the first.
• Treating search as the only channel. Search compounds, but a site with one channel has one point of failure. See our comparison of traffic sources for iGaming affiliates for what the alternatives really cost.

Everything above sits on top of the licensing frame, not beside it: gambling content is for adults only, 18+ or the legal age in your market, every page naming an operator carries responsible-gambling messaging and a route to help, and no ranking gain is worth publishing for a market where the brands you list are not licensed.

What we would do this week

1. Run the orphan subtraction. Export your sitemap URLs, crawl your own site following only anchor links, subtract one set from the other, and either link every orphan from a hub or drop it from the sitemap — before publishing a single new page.
2. Fetch your three most important pages with JavaScript disabled and confirm that every row, every internal link and every outbound CTA appears in the raw HTML. Fix what is missing at the server-render level, not with a workaround.
3. Hand-write titles and descriptions for every indexable page — title under 65 characters, description 120–160, unique, no truncation artefacts — then open the Search Console page indexing report and write down how many URLs sit in each state, so that next month's number means something.

None of this is fast and none of it is optional. Discovery and indexing are the parts of SEO you can control on a new domain; ranking is what you earn afterwards. Terms used above are defined in our glossary of affiliate and search terms.