Flash is bad for SEO because it locks text, links and navigation inside a single SWF file that Google always struggled to read and, since its announcement of 28 October 2019, ignores altogether: to Google, a Flash website is an empty page with a single URL. And since 12 January 2021 Adobe has blocked Flash content from running, so the people visiting that site can't see it either.
So the problem's solved, then?
Not quite. According to W3Techs, in October 2026 2.8% of websites still contain Flash as a client-side language. And in the UK the search "adobe flash player" gets an average of 880 searches a month on Google (Google Ads data via DataForSEO, peaking at 1,300 in September and October 2025): somebody, somewhere, is still trying to open an old site.
But that's not why we're writing about it in 2026.
The mistakes that made Flash invisible to Google didn't die with Flash. They moved to other technologies. In our audits we find them in three forms:
- links that are JavaScript buttons with no
href, like the buttons in an old Flash movie; - whole sites that change content without changing URL, with addresses built on the hash (#);
- important text written inside an image or loaded only on click.
The result is the same as with a 2009 Flash site: Google arrives, downloads the page and finds little or nothing to index.
And why listen to us?
Because we build the opposite of a Flash website every day. On the site of Macropix, a Milan-based LED wall manufacturer, every major installation has its own page with its own URL: the Italian query "ledwall" (LED wall) alone brought in 67,560 impressions in twelve months, on pages Google reads one by one. For Sunscape, which develops solar farms in Italy, we built a page for each region: from 329 to 5,809 organic impressions a month between March and August 2026.
A Flash site, given how it was built, couldn't have done either.
In this guide you'll find:
- what Google actually read inside an SWF file, in the words of the engineers who indexed it from 2008;
- why a single URL was enough to lose, even when the text was being read;
- our test of 10 October 2026: how many SWF files Google UK shows today compared with PDFs;
- the full timeline, from 2008 to the 2021 block;
- what to do if you still have a Flash website, in five steps;
- where the same mistake hides today, line by line against Google's documentation.
Let's start.
What Google read inside a Flash file (and what it never read)
Google started indexing the text in SWF files in 2008, and only the text. The Google Search Central post of 1 July 2008, written with indexing engineers Ron Adler and Janis Stipins, explains that the algorithm explored the file the way a person would, clicking buttons and filling in fields, and stored all the text it came across.
Everything else stayed out. These are the limits Google listed in the same post:
- images: text drawn inside an image wasn't recognised;
- buttons without text: no anchor text (the clickable text of a link) for graphic buttons pointing to a URL;
- files loaded via JavaScript: if the page called the SWF with a script, Google might not notice it;
- external resources: text in XML files or in other SWFs loaded by the movie was indexed separately, not as content of the page;
- Hebrew and Arabic: Flash content in right-to-left languages wasn't indexed.
So all you had to do was write the text in Flash and Google would read it?
In theory, yes. In practice Flash sites used exactly what Google couldn't read: headings as images, menus made of icons, text loaded from external files. The same post even suggested that anyone who wanted to hide a "loading" message from Google should turn it into an image. The technique for making text invisible to the search engine was the same one designers used to make it look good.
It's the same reason the alt attribute matters today: Google understands an image from the text you put next to it, not from the text drawn inside it.
One URL for the whole site (and so only one page to rank)
The worst flaw of a Flash website wasn't the unreadable text: it was that the whole site lived at a single address. Clicking "Products" or "Contact" didn't change page, it moved the movie to another frame.
For SEO there were four consequences, and each one on its own was enough to hold rankings back:
- One title and one meta description. The page on service A and the page on service B didn't exist: there was the homepage, and a single page ranks for a narrow group of searches.
- No deep links. Anyone who wanted to cite your price list could only link to the homepage, so all your backlinks ended up there.
- No structured headings. No H1, no H2: none of the heading tags Google uses to understand what each section is about.
- Blind analytics. In Google Analytics all the traffic showed up on one page: no way of knowing which section brought in enquiries.
Now compare that with Macropix. Of the nine keywords that bring the company the most qualified enquiries, none contains the name Macropix: they're Italian product searches such as "ledwall" (LED wall), "monitor pubblicitario" (advertising screen), "totem led" (LED totem) and "ledwall outdoor" (outdoor LED wall). Each one has a page that answers that search. On "monitor pubblicitario" the site went from position 88 to position 2 in five years.
With a single URL that climb wouldn't even have started. There was no page to push for "monitor pubblicitario": there was a movie with a button.
How to design the right structure is covered in our guide to website architecture.
Google has ignored Flash since 2019: our test on SWF and PDF (0 files out of 8 versus 9 out of 10)
On 28 October 2019 Google closed the matter. In the "Goodbye, Flash" post Dong-Hwi Lee, an engineering manager at Google, wrote that on pages with Flash content Google Search would ignore the Flash part and would stop indexing standalone SWF files. In the same post he recalled how widespread it had been: the Flash runtime was installed 500 million times in the second half of 2013 alone.
But did Google really do it?
We checked. On 10 October 2026 we queried Google UK with the filetype: operator, which filters results by file type, and downloaded the first page with DataForSEO. Same word, two formats: one Google indexes (PDF) and one it has said it no longer indexes (SWF).
| Search | Market | Organic results on page one | Files of the requested type |
|---|---|---|---|
filetype:pdf game | Google UK, desktop | 10 | 9 out of 10 |
filetype:swf game | Google UK, desktop | 8 | 0 out of 8 |
site:.uk filetype:swf | Google UK, desktop | 0 | no results |
The result is clear. With PDFs the operator works and returns nine files out of ten. With SWFs Google can't find a single file to show and falls back on generic pages (shops, game portals, a news story); on .uk domains it returns nothing at all. As far as Google is concerned, an SWF file today doesn't exist.
The comparison with PDFs tells you something else too: the problem isn't having a file instead of a page. A PDF ranks because Google extracts its text, and we explain how to handle that in our guide to PDF SEO.
Does your site show Google less than you see?
You've just seen how a whole site can drop out of the index while working perfectly on screen. Today it happens with JavaScript menus, hash URLs and text loaded on click. We look for these by comparing the page the browser sees with the one Googlebot downloads, and we tell you what to fix before you think about rebuilding everything.
Book a strategy consultationPhones, security and file size: why Flash died outside Google too (timeline 2008-2021)
Google stopped reading Flash when hardly anyone was using it any more. The decline had started with browsers and smartphones: according to the Chromium blog, in 2014 over 80% of daily Chrome desktop users visited sites with Flash; by July 2017 the figure had fallen to 17%.
| Date | What happened | Effect for anyone with a Flash website |
|---|---|---|
| 1 July 2008 | Google improves indexing of text in SWF files | Text visible in the movie can appear in snippets; images and external resources stay out |
| 25 July 2017 | Adobe announces the end of Flash Player for the end of 2020, together with Apple, Facebook, Google, Microsoft and Mozilla | Three and a half years to migrate |
| 2019 | Flash disabled by default in Chrome 76 and Firefox 69 | Visitors have to allow the plugin to see the site |
| 28 October 2019 | Google announces it will ignore Flash content and stop indexing SWF files | The text in the movie drops out of the index |
| 31 December 2020 | Adobe ends support for Flash Player | No more security updates |
| 12 January 2021 | Adobe blocks content from running in Flash Player | The site can't even be seen on screen |
Three factors did the rest:
- Smartphones. iPhone and iPad never supported Flash Player. And since Google started indexing with its smartphone crawler (mobile-first indexing, completed between October 2023 and July 2024), what can't be seen on a phone doesn't count.
- Security. Kaspersky counts more than 1,100 Flash Player vulnerabilities in MITRE's CVE database, some with the maximum severity score, CVSS 10.0. A plugin that had to be updated by hand on every computer was a risk browsers stopped taking.
- File size. Before showing anything the browser had to load the plugin and then the whole movie, often behind a "loading" screen. Today the same wait is measured with Core Web Vitals.
And how did Adobe explain the decision?
With a line that still holds: in its announcement of 25 July 2017 it wrote that open standards such as HTML5, WebGL and WebAssembly now provided many of the capabilities that plugins had introduced. The web didn't lose its animations: it moved them into code Google can read.
Still have a Flash website? What to do in five steps (and why old # URLs can't be redirected)
If a site or one of its sections still uses Flash, the job is a proper migration, not a design touch-up. In order:
1. Find where Flash is still used
Open the page source (Ctrl+U) and search for ".swf", "<object" and "<embed". Often the site is already in HTML and Flash survives in a banner, an old video or a virtual tour. The 2.8% measured by W3Techs is partly made up of leftovers like these.
2. Recover the content
Text, images and price lists inside the SWF need extracting before you throw it away. To view the movie in a modern browser there's Ruffle, an open source Flash emulator that works without a plugin.
3. Rebuild with one URL per topic
One page per service, per product, per area: this is the step that creates the pages that didn't exist before. Each with its own title, its own H1 and a readable slug.
4. Map the redirects, with one exception
If the old site had real URLs, each one needs a 301 redirect to the equivalent new page. Many Flash sites, though, used addresses like example.co.uk/#/products: the part after the # never reaches the server, so a server-side redirect can't tell them apart. In that case the only URL to protect is the homepage.
5. Take a snapshot of the data before launch
Save the old site's impressions and queries from Search Console. Without that starting snapshot you won't know what you've gained, or what you've lost.
When rebuilding everything is the right call, the numbers show it. Scovaventi, an Italian food producer, had an online shop that was slow and awkward to use on a phone, with traffic and zero sales: rebuilt on Shopify, it closed its first year with €77,000 in sales and an average basket of €190. It wasn't Flash, but the problem was the same: a technology working against people arriving from their phones.
The full procedure, with how long the dip lasts and the checks to run on launch day, is in our guide to redesigning a website without losing rankings.
Flash's heirs: where the same mistake survives in 2026 (with the line from Google that says so)
Each row in the table is a flaw we'd have found in a Flash site in 2009 and that we now find in sites built with page builders, JavaScript frameworks or heavy themes.
| Flaw of the Flash site | Today's heir | What Google says | How to check |
|---|---|---|---|
| Graphic buttons with no text | Links created with onclick or with a span instead of <a href> | Google can generally crawl a link only if it's an <a> element with an href attribute (guide to links) | Inspect the menu: every item should be an <a href="..."> |
| The whole site on one URL | Single-page apps with hash addresses (#/products) | Use the History API rather than fragments to load different content: Googlebot can't reliably resolve those URLs (JavaScript SEO) | Click through the menu: the address should change, with no # |
| Text inside the movie | Headings and prices inside images, sliders or canvas | As early as 2008 Google excluded text inside images; today it understands them from the alt text and nearby text | Try selecting the text with your mouse: if it won't select, it isn't text |
| Content reached by navigating the movie | Text that appears only on click or swipe | Don't load primary content in response to user interaction: Google won't load it (mobile-first indexing) | URL Inspection in Search Console, crawled HTML |
| Movie loaded via JavaScript | "Shell" page that shows the text only after running JavaScript | Google renders the page later, when its resources allow (JavaScript SEO) | Ctrl+U: the main text should already be in the source |
The last row needs a warning. Googlebot does run JavaScript, even if it queues it. Most AI crawlers don't. According to an analysis by Vercel of traffic on its own network, none of the main AI crawlers executes JavaScript: not OpenAI's (GPTBot, OAI-SearchBot, ChatGPT-User), not Anthropic's ClaudeBot, not PerplexityBot. The exception is Gemini, which uses Googlebot's infrastructure.
In plain terms: a page that shows its text only after JavaScript runs is, for ChatGPT and Perplexity, what a Flash site was for Google before 2008. Which bots actually visit a company website is covered in our guide to AI crawlers.
If you recognise your site in more than one row, technical SEO comes before any work on content. It's the check we start with on every SEO consultancy project.
Frequently asked questions about Flash websites
No. Since its announcement of 28 October 2019 Google has ignored Flash content on pages and no longer indexes standalone SWF files. In our test on Google UK on 10 October 2026, the search "filetype:swf game" returned no SWF files at all, while "filetype:pdf game" returned 9 PDFs out of 10.
No. Google ignores the Flash part and indexes the rest of the page. The damage is indirect: as far as Google is concerned, the text and links inside the movie don't exist, and visitors see an empty box.
It depends: not in modern browsers, because since 12 January 2021 Adobe has blocked content from running in Flash Player. You can view it with emulators such as Ruffle, which are useful for recovering text and images before rebuilding the site.
It depends: HTML5 makes text readable, but an HTML5 site with menus without href, hash URLs or text only in images repeats the same mistakes. What matters is that every piece of content has its own URL and sits in the HTML.
It depends: content that's still valid should be extracted and reorganised, but the technical side has to be rebuilt anyway, because an SWF file doesn't turn into a set of indexable HTML pages. The real work is deciding which pages to create, one for each search you want to capture.
Summary in five lines (and what Flash teaches anyone building websites today)
Flash locked text in a file Google could only half read.
It put the whole site on a single URL.
Google has ignored it since 2019, and since 2021 it won't open.
2.8% of websites still carry traces of it.
And its mistakes live on in sites built purely in JavaScript.
There's one thing people rarely say, though. Flash didn't die because of SEO: it died because of security and smartphones, and Google acknowledged it years later. For eleven years, from 2008 to 2019, Google tried to read it, and anyone with a Flash site spent those years waiting for it to get better at it. With AI crawlers that don't run JavaScript today, the same scene is playing out: if you wait for them to learn, you stay out of their answers in the meantime; if you write your text in the HTML, you're in straight away.
If you want to know what Google, ChatGPT and Perplexity see of your site, book a strategy consultation: we start by comparing the page you see with the one the crawlers download.
See you next time!
