Searcheye Launches AI Recommendation Engine

[tldr_widget]

 

When our team set out to build a world class link-building product, we wanted to answer the most common question users ask us: “well, which links would you recommend?” At a small scale (i.e. your local personal injury lawyer), the answers are easier to determine by answering some basic questions:

  • Is the publisher’s site relevant to your industry?
  • Does the site you’re getting the link from get traffic or is it a ghost town?
  • Is the site good for your SEO (does it have a high domain rating, etc?)
  • Is the value worth the price the publisher is asking?

As we sign up more publishers (and our user base continues to grow), these questions become hard to answer in faster, and more automated ways. At some point, humans are just too slow to digest 1,000s of possible links and identify the best links for. This is hard for one user, let alone 1,000s of users.

The solution is obvious: algorithms and computing are much better at this particular task.

Building v1 of the Recommendation Engine

One of the useful things about working with search engines and the open web is that we have access and visibility to the data we need. By definition, sites that want search engine traffic must be easy and available for anyone to view.

At the same time, we also have user analytic data, and signals around what kind of links they prefer. For each user domain on-boarded, we track:

  • Keyword rankings & traffic
  • Each page on their site
  • The site’s domain rating and monthly traffic 
  • The industry categories the user is in
  • The above metrics for key competitors

We take a similar approach for each publisher (e.g. vimeo.com, godaddy.com) that is on our platform. 

It’s interesting how often the best answer tends to be the most elegant one. We create a knowledge graph of every client domain. We also create a knowledge graph of every publisher’s site.  We then compare these two graphs and surface the content opportunities that lie at the intersection of the user’s content wheelhouse, and publishers who are writing about those subjects. These opportunities are shown as a relevance score. The higher it is, the more relevant the content opportunity is to the user. Beginning a month ago, we show the relevance score on every product card:

Above: An example of a product card, including the relevance score on the right.

This is an important step towards our overall vision of making it fast and easy for users to secure predictable marketing traffic that makes their business bigger, lowers their customer acquisition cost, and creates long term defensibility in ways other traffic channels can not.

While the tool is already accelerating results for users, we have a journey ahead of us to improve the quality of our recommendations.

Searcheye vs the Competition

There are many ways people build links.  Techniques include guest blogging, broken link building, resource link building, along with many other approaches, and variations within these approaches. What’s common to how most people execute these approaches is that they lack operational consistency, analysis, and numbers driven accountability . An absence of industry-wide standardization and transparency has created a haven for incompetent and dishonest operators who take advantage of users.

Search engines are also deliberately vague (understandably) about how their products work, which only exacerbates the issue. Ironically, by not providing clarity, the large search engines make it easier, not harder, for a spammy ecosystem of low quality links, and content, to proliferate. This is not good for anyone involved (even the bad actors, who discourage users from participating). 

This lack of industry standards has given rise to dishonest practices by some vendors who sell ineffective link-building services, making it difficult for businesses to identify legitimate providers who can deliver high-quality links.

This current state of affairs ultimately damages the industry’s reputation and long-term growth potential.

Sunlight is the best disinfectant. That’s what our recommendation engine (along with our future product releases) do – they demystify the search engine blackbox to return control and power back to the user. 

It is necessary to establish industry standards and best practices for link-building so that users are not taken advantage of.

Here’s how most link-building is done today:

The Searcheye recommendation engine means algorithms do the heavy lifting of identifying good link opportunities so users no longer have to sift through hundreds of spam emails and thousands of sites on multiple lists to find valuable links. Our platform provides you with confidence that every link recommendation will have a significant positive impact on your website’s overall SEO and search rankings.

>> Learn more on how we source Guest Post Opportunities (for ⏰3pm Links)

Start Building Links at Scale

Discovering the most valuable backlinks for your business has never been simpler. By signing up for 3pm Links, you can receive the internet’s top link-building opportunities delivered straight to your inbox every day at 3pm.

[cta_banner_link]

Frequently
Asked Questions

Book a Demo Sign Up