Nestoria is a property-focused vertical search engine based in the UK, founded by former engineers at Yahoo.
I talked to co-founder Ed Freyfogle to find out more about the site, the challenges of starting up in the UK and APIs…
What prompted you to launch Nestoria? Where did the idea come from?
I moved to the UK in the summer of 2005, and found searching for a place to live to be harder than it should have been. At the same time, there was a great deal of innovation happening around online mapping, led by the launch of Google Maps in the spring of 2005. It seemed natural to combine the improvements in mapping with search.
You were formerly at Yahoo – what did you do over there and how did
that experience prepare you for the launch of your own vertical search venture?
I joined Yahoo! in 1998 as the first developer based in Germany. For the first few years I focused very much on various content products (news/sports/finance/etc). Later I was in charge of search engineering for all European markets. That experience opened my eyes to the power of search and especially the need to create a compelling user experience.
Why do you think vertical search is becoming such a big deal?
I think consumers are becoming more and more savvy internet users. Sometimes people want to browse, to do research or to be entertained, but sometimes they just want a certain set of information quickly. When they’re in that mode a search engine is the optimal experience.
So what were the key challenges in starting up Nestoria? Why did you
start up in the UK?
I think the most difficult challenge for any entrepreneur is getting the right people to help you. We’ve been very fortunate to put together a first class team. In that respect, being in London was a definite advantage – there are great people here.
How do you prioritise tasks / feature development?
It’s a challenge. Because our site seems simple, everyone thinks they know how to do it better. We always try to base our decisions on numbers, not on opinions.
What’s the business model for Nestoria?
We add value by organising data in a way that is relevant and user-friendly.
Our fundamental business model is advertising on a per lead basis.
How do you collect the search listings? Crawling / scraping, RSS,
APIs, spreadsheets? Unless you have APIs in place it must be difficult to extract meaning from all of that unstructured data? How do you overcome this?
We have a combination of crawling and direct feeds. Data quality is a massive challenge. Agent’s have a way of presenting things that’s, well, let’s just say “unique”. We talked about it a bit on our blog.
We’ve built some sophisticated natural language processing tools to try to organise the data and extract a bit more meaning. It’s an ongoing challenge.
It’s good to see that Nestoria has an API… can you evangelise
about APIs for the benefit of our readers?
Someone selling a property wants to get that property listing in front of as many potential buyers as possible. We think we have some good ideas on helping people search for property, but we know we don’t have all the ideas. So we’ve opened our database to the community.
Should there be a standard API for the property sector as a whole?
Would you be happy to share an API with your competitors, in the same way that Google, Yahoo and MSN are sharing Google Sitemaps?
We’re happy for anyone to use our API.
A couple of months ago I read about ‘Nestoria Rank’. Can you explain
more about the rules that govern your proprietary algorithm?
Building a compelling search experience is about several factors; comprehensiveness, freshness, usability, and relevancy. Nestoria Rank is the name we give to our efforts on relevancy. It’s a series of algorithms we use to prepare and sort the data. It’s something we spend a lot of our time thinking about. If all we have to go on is “I want a cheap flat in North London” establishing relevance isn’t simple.
Anyone who is interested (or has suggestions) should subscribe to our blog where we go into more detail.
Do you think social factors can play a part in driving relevant /
accurate / up-to-date search results?
Yes, this is something we’ve been discussing quite a bit recently. There are a lot of interesting sites out there building communities around location and geodata (one I like a lot is Tagzania). There’s a lot to learn from sites like that.
What facets of the user interface and site functionality do you feel
are critical to the success of Nestoria?
People always tell us they like the local content (schools, transport, pubs, etc) that we’ve aggregated. Also, they seem to like the interaction with the maps. Beyond that we’re always experimenting with new usability features. We still have a lot to learn.
Why don’t you allow people to post classifieds directly? Is this to
preserve good relationships with the classifieds / property listings websites?
The reality is that very few people in the UK want to sell their own house. When push comes to shove, most people don’t trust themselves to get the best deal or they don’t have the time. I would guess that over time as people become more familiar with the internet this might change.
You’ve gotta love that Google Maps API, right?
Thanks, yes people love the maps, especially the local data on the maps. Big thanks to the team at Google for making such an amazing tool. We recently launched a tool that lets any webmaster add our maps to his or her webpage with cut and paste
Any ideas you (or your readers) have on what we could do better are very welcome. The UK has a great community of people experimenting with geodata, and we’ve been lucky to work with and learn from them.
Chris Lake is editor at E-consultancy.com (firstname.lastname@example.org).