Last week we introduced the Matrix API get call. We didn't expect the overwhelming response for more information, so in the absence of finalised documentation we'll publish a few articles detailing the most common requests. The get request is made when a page is required to be indexed for the first time (or subsequently with the appropriate index or crawl instruction), and this functionality wasn't previously described in the required manner because of the low public utilisation of the service. It's important to note that subsequent API requests used to retrieve crawled pages are made to the page endpoint, and this article details the (truncated) response. We've redacted certain fields we're not ready to be made public, and we've removed certain data that identifies the overall rank of websites.
The purpose of the API is introduced in an article titled "The Matrix API, but in short, the API graphs what we call The Financial Web - a full and complete index of all broker websites and advertising. Armed with definitive and virtually complete understanding of the broker website landscape, we're better positioned to support our SEO clients, and we're more informed with everyday decision-making. In terms of our own broker website framework, it arms us with the necessary unbiased confirmation to make the objective and definitive claim that our broker website performs better than others. Given that we don't disclose client details, we generally lean on our far more advanced technology to support our claims... and to question those of others.
Do We Provide an SEO Service: Do we provide an SEO Service? The short answer is yes and the long answer is no - not really. SEO and AI-indexing has become a function of simply creating more video and article content, and demonstrating the Lantern attributes of Expertise, Education, Entertainment, and Experience (E+AT) via your own content and social outreach. Unique, compelling, engaging, and relevant content is the key to organic success, and this is supported by Yabber's SEO features. While we do provide an SEO service, it's well outside the budget of most businesses. Instead, our excellent website framework provide a platform upon which to thrive.
An article titled "Using the Matrix API to Check for Duplication, AI, or Plagiarism" details how we use the API to assess duplicate content - a significant factor in the effectiveness of your blog - and in a scheduled article we'll review the popular endpoint that'll return competitor page or site keywords. There's now over 70 endpoints, so we'll probably wait on the documentation before we share anything else.
API Key: Until recently, the API didn't require authentication, and we published the data in full on our website. Reasoning for publishing data in full dates back to comments I made on a specific aggregator website resulting in frivolous legal action (basically, an aggregator didn't want to expose their incompetence). I published the evidence, obtained supporting documentation from Google, and sourced dozens of supporting statements. The claim was later dropped. Opening up data clearly proved our point when it came to exposing the performance of certain aggregator websites, and generally speaking, they all rank lower than even poor broker websites. While the API now requires authentication, API Keys are generally provided to those that need one, and we expect to publish the data at its permanent home in full (once again) sometime soon.
PHP Function: If you want to have a play with the API yourself (and you have the necessary access), we've published a crude little PHP function in an FAQ titled "How to Use a PHP Function to Access the Matrix API". The simple function makes it easy to access the 70+ primary endpoints. API Keys associated with BM Clients will return all data without exception. Non-client API Keys include some redacted data, and only clients are permitted the 'crawl' or 'index' instruction (amongst others). No rate limit of any type applies to our clients.
Website SEO Update: After a complete review of Xena (statistics, heat maps, scroll records etc) and the Matrix data, we're pushing a completely revised SEO module to all our client websites this month. The module is exceptional and provides holistic SEO management (although the previous module was just as efficient). Significant changes were also made to the Partner Plugin SEO features and the free Funnels product.
The Matrix page Endpoint
The page endpoint returns information relating to a specific page. The domain_id, url_id, record_id, hash, and a number of other keys are used in subsequent API requests that'll return more information or comparative data. All the values returned in the specific response below should be considered invalid as we've altered some of the identifiers.
An extremely large amount of information was redacted in the response because it exposes proprietary methods, such as inclusion of the advertising-focused Insights API and others.
We'll query a page on Otto's Home Loan Experts 
(HLE) website. I'm using HLE's website only because they typically do quite well, although as you'll see, their SEO presence is predicated on basic good practice and not the typical keyword stuffing inclusions employed by most businesses. Because HLE's page doesn't include all response data, we'll follow by including a few examples sourcing data from our own site.
HLE: Otto is known and recognised in the industry for his exceptional knowledge of policy and process, and his website content is a testament to his real-world knowledge base. I've often wondered if Otto developed his knowledge by writing for his website - a true method of knowledge reinforcement and professional development - or if he came across the expertise via other means. Either way, he does quite well and Matrix consistently ranks his website in the top-30 nationwide,
Endpoint: matrix/page/matrix.json?id=[URL_ID]&apikey=[APIKEY]. The 'URL_ID' may be the full 'base64 URL-encoded' URL, 32-character Record ID, or the numeric url_id returned in most responses. The response below includes all fields returned in the get request in addition to others. The result:
Social: The 'social' data found on the general page isn't entirely accurate, but with the sameAs key in the Organisation JSON-LD markup we're delivered far more reliable and accurate information. If social data is provided in the JSON Linked Data we'll always use this for page analysis ahead of resolved information. This same practice applies in all cases where Linked Data provides applicable information.
Site Crawling & Indexing: If you have requested a crawl of a website, each link will include a crawled_at timestamp in the returned internal link array. To get an update of how many pages are indexed and how many are crawled, send a request to the matrix/review/matrix.json?id={SITE_ID}&apikey={apikey}. A limit of 10'000 crawled pages applies to any single site. Client website are crawled very differently since each site includes over 75-million pages.
The information returned via page is far more comprehensive than get because it's retrieved after the page was fully indexed and assessed against all other pages. Basic information such as JSON Keywords, licence information, and general longtail keywords are probably the most noticeable inclusions, although duplicate and similar articles are probably the most significant additions... although in this case no similar articles within defined similarity parameters were found.
What Can Be Resolved at a Glance? By simply looking at the returned data we're able to quickly resolve that HLE put moderate effort into their JSON-LD (Linked Data) for easy ingestion by search engines, although they don't seem to apply a great deal of effort into their keyword considerations and consistency. The article JSON-LD markup shows the current keywords as shown below (a strong indication of page purpose):
While the 'rates', 'property' and similar general keywords were weighted more heavily than others, we would have expected longtail phraseology to match against those declared via the structured markup. The article only included a two-word longtail presence via the phrases 'growth rate', 'cash rates, 'previous months', 'mortgage brokers', and 'housing prices', with none matching those that are used to define the page objective. No three, four, or five-word phrases were found (used subtly and without stuffing to indicate the page objective). HLE have a good Search presence with established experience, expertise, and authoritativeness, so they don't have to work as hard as others in order to gain good SERP... although in this case the article doesn't rank particularly well against competing pages.
HLE include seven JSON-LD blocks which is admirable, and their consistent usage of structured data is a defining feature of their page architecture. That said, there were elements that I expected to see but didn't, such as (but far from limited to) the geo coordinates - a good feature for local search authority.
At a glace, we can see that the article includes three article links (which are measured independently to page links): an assessment form, a phone number link, and a calculator link. None of the conversion assets were on the target page, and the funnel practice isn't something that resembles what we'd describe as best practice. One of our core organic linking considerations is orientated around education, and since precisely zero internal links (or graphs, images etc) were provided we can only assume that this particular article is a reproduction of an email campaign (not uncommon for RBA news), or it was created as a destination that email recipients were sent - it's clearly not part of the internal education funnel architecture.
I'm not going to pick this page apart since HLE effectively own Search, and they don't have to work nearly as hard as others in order to return results. What the result does do is give us an indication of the basic yet reasonably comprehensive data extracted from the page which can be measured against the Finance Matrix - or The Financial Web - in order to position page presence against others in the industry.
We have an article scheduled that talks a little more about keywords and longtail keywords, but for the sake of completion we'll look at only the keywords extracted from one of our articles tha describes features of our Conditional Website. Note that the below response includes longtail keywords. If we had more than one occurrence of any five-character word, that phrase would be returned as well. The article on keywords will detail how to specifically query the counts of any similar string in article text.
Spot the Possible Error?: In creating this article we've spotted an error with how special characters are removed from text before evaluation. We'll have to go back to the drawing board and debate the use of hyphens and other similar characters.
The standalone keyword and keywords endpoint returns more detailed keyword information which includes aggregate keyword counts.
In an article titled "Duplicate Content, Attribution, and Citing Sources in Article Content" we noted that citations, footnotes, and quotes are all returned in the page response, and this information contributes towards our resolved SEO effectiveness (in company with crawling the externally linked pages). An example of returned citations and attribution is as follows:
Once the cited page (if a cite URL is present) is crawled, a cite_id will be returned with each record, and this ID may be used on the cite endpoint to query the cited page. This is one of only a few cases where we follow links outside of the crawled website domain.
What Do We Do With Page Data?
The Matrix page data in isolation is nice, but it doesn't really tell us anything. When data is measured against all other industry pages and then weighted against other pages with similar positioning, we're able to compare the results against those results returned via search engines and pinpoint the specific page attributes that likely contributed towards the higher score. For our own clients, we're then able to manufacture content with knowledge of what might be escalated consumer interest. Paired with trending data from our Athena News API, we're armed with more than the guesswork that tends to guide most businesses.
Athena News: Athena News is a module that ingests news and industry RSS feeds. We extract keywords and other data in order to understand trending news, consumer interest, and cyclic sentiment, and we build content with this understanding. The millions of news articles and specifically defined feeds may be queried against the news.json endpoint, or returned via a simple RSS feed. It's a big module that'll require its own discussion. Note that all news is queried against Matrix for general page and localised data, but links are not followed. Athena news is also queried for references to specific businesses or individuals to identify authoritative links.
Knowledge of a single page is rather pointless without peripheral context, and this is where Matrix shines. While Matrix isn't a search engine, it does perform the same function in order to garnish massive data sets that allow us to accurately crunch numbers that are algorithmically assessed for broad website performance... and it is extremely accurate.
There aren't that many mortgage broker websites in the market, so when you weed out those that may as well not exist, those with errors, aggregator websites (which may as well not exist), and some other groups, we're not left with a great deal of competition. If you apply consistent effort in your content creation, you're only ever competing against a very small market, so you're able to capitalise on organic traffic far easier than most. Those with a strong search presence will return triple-digit high-intent leads every single month without a great deal of effort (assuming, of course, your website is structured in the manner we constantly espouse publicly).
Malware, Dead Links, and Plugins
Outside the scope of this article, each page is assessed for potential malware, dead links, and website plugins. All this data is measured against our Malware API and heuristic engine to identify suspicious patters, vulnerabilities, and Indicators of Compromise (IOCs). This data is returned via the malware, plugins, and broken endpoints.
Landing Pages: We don't specifically identify landing pages. While every page is a type of landing page, we can generally resolve a page as a landing page but the system can only make assumptions. However, in company with the Insights API, we categorically resolve those pages linked to in advertising as purpose landing pages, but this doesn't account for those landing pages that are used internally but are not associated with paid promotion. It's worth noting that Insights indexes every mortgage broker ad on virtually every platform and evaluates it in a number of ways, and we use this to measure the performance of our own ads against those created by others.
Conclusion
What becomes clear through the Matrix API - and particularly through the page endpoint - is that real insight comes from context. A single data point is inert; dozens start to form a shape, and only when paired against hundreds of thousands of others does it become meaningfully actionable. This is the central premise of the Matrix project: to allow businesses, marketers, and analysts to remove the guesswork and operate with clarity.
In a space where every agency and aggregator is claiming SEO superiority, it’s no longer good enough to rely on anecdotal success or vague impressions of performance. The Financial Web as mapped by Matrix doesn’t just enable us to validate our claims - it compels better decision-making, finer targeting, and more efficient content creation. Data can - and should - be the standard against which all marketing efforts are measured.
Whether you're auditing your JSON-LD usage, evaluating keyword depth, or benchmarking your SEO structure against industry leaders, Matrix lets you step beyond superficial metrics and ask better questions. With the full context of your competitive landscape at your fingertips, decisions become strategic rather than speculative. If you understand the data, you can understand the behaviour. And if you understand the behaviour, you can engineer better outcomes.
In future articles, we’ll continue to unpack the Matrix API’s capabilities and the broader role of structured data, comparative keywording, and longtail strategies in finance SEO. Until then, remember: performance without proof is irrelevant. Data is the only language that scales.
Matrix is just one of many reasons we consistently get better results.
the depression, which saw real GDP fall 17 per cent over 1892 and 1893, and the accompanying financial crisis, which reached a peak in 1893, were the most severe in Australia's history. The overextension of the 1880s property boom and its unravelling led to an abrupt collapse of private investment in the pastoral industry and urban development and a sharp pullback in public infrastructure investment. A fall-off in capital inflow from Britain, adverse movements in the terms of trade and drought in 1895 accentuated and prolonged the depression.. [ View Image ]
