Today marks an important day in our mission to leverage data & AI to improve Brandable domaining capabilities. Before I share more details, let’s start with why we are doing this.

The Discovery Challenge

It’s no secret that a large majority of domains never sell. Even when investing in high quality, brandable domains, the annualized sell through rate is rarely above 5% – an incredible loss of opportunity.

The main issue with brandable domains is discovery. Many companies are looking for unique brand names that evoke positive emotions, metaphors or imagery to reflect their brand promise. Such names are usually distinctive and don’t include generic industry keywords. Think of popular clothing brands such as Zara, Shein, beauty brands like Clinique or Lush. It’s highly unlikely buyers would stumble across these names without getting very specific in their search. Most marketplaces utilize a text based search technology to present names to buyers. Since the buyers are often not able to find relevant results for their specific use cases, they resort to searching for broad and generic keywords such as “Clothing”, “Fashion” or “Tech”.

We’ve been tackling this challenge by building a complex genome of root words, emotions, visual imagery and metaphors to connect the right names to the right buyer intent.

The new model, coupled with deep AI based integration is now ready and we are calling it Discovery 2.0.

What’s included in Discovery 2.0?

There are several aspects of this framework that serve as core building blocks. Some of those aspects are already live, while others will be enabled in the near future.

1. Rich & Persuasive AI Descriptions

An important aspect that helps in selling Brandable domains is persuasive descriptions. A good description can often provide additional context that can help buyers move forward with the purchase.

Consider the domain “Lazuni.com” as an example. To a casual buyer, the name looks like a decent 6L brandable domain; however if you read the description (which was auto generated by our new AI integration), you can see how it explains the Latin origins and ties it to Luxury and timelessness.

A strong description can also improve the SEO aspects of the landing page, making the domain more likely to be discovered in Google search.

You can try generating a description for any of your domains by visiting our recently revamped Domain Insights Page.

Needless to say, a good description alone is not enough to improve the discovery of domains. Our new discovery framework goes much beyond AI generated descriptions.

2. Deep Domain Classification For Better Discovery

Let’s say you are a buyer looking to launch a natural skincare brand or a sustainable clothing brand. You will likely search for names that match those specific themes. The challenge is that domains are typically classified at a broad level (e.g. Beauty or Clothing). For example, we currently have over 20,000 “clothing” domains in our Premium Marketplace. As a result, it makes it difficult for the most relevant names to be recommended even if the search is powered by an Advanced AI based solution.

Ultimately, the strength of any AI-related discovery tool is the underlying data that feeds that tool. Until now, we relied heavily on root words and manual classification to improve relevance of our results. With Discovery 2.0., we have made significant improvements in the classification capabilities that can help us identify specific use cases related to the domain.

Here’s an example of a domain PurityCompass and the specific use cases detected by the new algorithm. If a buyer searches for use cases such as Sustainable clothing brand or Natural Skincare brand, it will now allow us to prioritize the discovery of names like these which are much more relevant.

In addition to identifying specific use cases, we have also improved the capabilities to detect the emotions or feelings associated with names. This way, if a buyer’s brand mission is all about “Purity”, they can easily find those names that evoke the feeling of Purity. Previously, most of this classification needed to be done manually, which makes it extremely limiting. Now, we will automatically enhance the classification for domains in the Marketplace.

The new classifications will also allow us to recommend the most relevant names to contest holders with much more precision. Since we already know the specific use case of a contest holder’s business, our AI can now recommend the most relevant domains from our premium marketplace based on the styles of names that received high ratings by the contest holder.

You can try generating the potential use cases and emotions for any of your domains via our Domain Insights Page.

3. Better connection with buyer journey

While Deep classification is one of the necessary ingredients for better discovery, it is not enough. We need to pair that data with how buyers are browsing our marketplace to create meaningful associations based on real end user signals.

For example, if a buyer searches for a certain keyword and shortlists 5 different names, we can start building dynamic associations based on search intent and those domain attributes. With over 3M unique monthly visitors, we are now able to create aggregated insights that can improve the discovery experience. The new buyer journey is not yet live, but we have begun optimizing our model with over 200K daily recommendations.

What’s Next?

While our foundational capabilities are now in place, we still need to do lot of work to continue to improve the discovery experience. We will be announcing several updates to our platform in the near future using this new framework. AI will never replace human learnings and insights but it has the potential to make significant improvements in the discovery of brandable domains, where context, emotions and use cases play an extremely important role.