One of the big problems of the internet is that is practically impossible to keep up-to-date on preferred topics. You can limit your sources, but this can mean missing a lot of valuable data.
Google and other search engines help, and tag clouds and other social tools can make it easier to follow other people's recommendations, but wouldn't life be easier if someone was out there on the web scouting for the stuff that you really wanted to see?
Topikality is trying to fill that role. The brainchild of entrepreneurs Phillip Scott and Richard Heycock, Topikality is designed to learn what a user is interested in, and then scour the internet for matching articles that are delivered to the user on a regular basis.
Scott says the key difference is that Topikality presents what you have told it you are interested in, rather than just showing what is popular.
"It looks at everything to do with the articles you are reading, not just specific words — so phrases, words and the context of those words and phrases," he says.
Rather than searching the entire internet, Topikality is restricted to specific sources such as media outlets, blogs, governments and well-regarded organisations. Users can also nominate additional sources.
The value is in the artificial intelligence techniques used — in this case machine learning — which means that the more the system is used, the better it becomes at delivering the right information.
Search tends to be a once-off activity and does not record whether you were satisfied with the result, and hence the search engine does not learn your preferences.
"Once you've done a little training on the system, it delivers results to you ever day," Scott says. "So the net result and return for you increases over time."
Hence if you want to keep up to date on topics such as treatments for specific diseases, it can over time narrow down your interests.
There is a lot to like about Topikality. The service, which has entered a public beta, has a clean interface, and the technology appears to do what it is supposed to in terms of dragging back information that you might be interested in.
The company was founded in December 2007 and has been in development ever since. While the technology is not patented, Scott says the sheer hard work and complexity behind Topikality is a series barrier for competitors.
Heycock has a background in mathematics, while Scott has been the co-founder and CEO of Australian customer management company, Prosper Business Solutions, which sold to a US company in 2006.
But it looks like a good idea in search of a business model, and faces two significant hurdles.
Firstly, the founders are uncertain as to how they will make money out of it.
The company is not currently charging for the public beta service, but may look at subscription models for corporate users. The so-called free-mium model is a very difficult one to translate into dollars however, and will require Topikality to find a lot of extra functionality to make it worth companies paying money for it. Just ask the folk at Twitter about that one.
Scott also wants Topikality wants to sell the software as an appliance to companies with large information repositories, and is currently in an unpaid trial with a large media services company that processes hundreds of thousands of documents each day, and has another lined up. The other opportunity is in enterprise search, for information retrieving from the web or intranets.
But herein lies Topikality's other major hurdle. It is playing in a competitive market, against competitors including Google, and there is a good chance that its service can become quickly commoditised (another Aussie company, Isys, has had stored search for years). And with limited sales and marketing resources it will be difficult for the company to make its voice heard amongst the noise.
Scott says he isn't worried about the competition.
"If everyone in the IT industry worried about those things nothing would ever get developed," Scott says. "This is something worthwhile that's worth having a crack at. Who knows that the potential might be and where it will end up."
So while the opportunity is potentially big, until Topikality can batten down a reliable business model, its chances of being a boom company are slim. Thankfully the company is funded by its founders and has a very low burn rate, but at this point the question mark over its potential to generate revenue is a big issue.
- Does it address customer pain? TICK
- Does the technology work? TICK
- Is the business model robust? CROSS
- Does the company have sufficient resources to compete? CROSS
- Is it under threat from competitive pressure? CROSS
bootstrappr opinion: BUST