Google for Cars

Would you recommend this product?
No reviews yet
Hi ProductHunt, thanks for the add. @greggfidan we have a lot under the hood that you may not realize. We do more than just price analysis, our rankings are more like Googles Pagerank. We predict user interest in cars not just helping you find the cheapest ones... We also have the most used cars (250k more than autotrader and over 500k more than the sites you listed). We basically reversed engineered every single dealer web platform so we collect the data directly and organically (Dealers do not pay us for the listings directly nor can they buy their way to the top like the other sites.). AU.TO is also text based search, so users can simply type in keywords giving you precision searches with a single input. We are a small passionate team competing in a huge space our goal is simply to help consumers find the best cars. We have lots of stuff coming, next week we announce our first mobile product stay tuned... Here is a short clip we put together that explains AU.TO
@google4cars - what's the best way to reach you? I'd like to ask you some more questions. Or, you can email me:
@realcartips @greggfidan emailed... I can also be reached @
I use this for buying cars - better than craigslist
@dilthemd oh, I know - it's game-changing for the car buyer and for the dealer as well. Great find, Dil.
As someone who covers the car buying space, I'm not seeing anything that differentiates this site with other used car sites. and have been doing used car price analysis for years (with graphs that look very similar).
@greggfidan Here are some of the answers to the questions you emailed me... 1. "We predict user interest in cars not just helping you find the cheapest ones" Can you elaborate on this? We have a 60 page patent that drills in detail, but basically we initialize the vehicles in the system with a price/mileage regression analysis. We also do a much better job at comparing apples to apples (We use machine learning and vin data to compare vehicles as you know all the data is not in the vin so we use machine learning to predict similar cars. (As a comparison you mentioned our regression analysis has 3x the correlation coefficient. We are not comparing base models against loaded ones) Next we use our 60 node openstack cluster (Over 1 million in hardware) to analyze all the user data we track EVERYTHING. So the car with the big black wheels will move up our list and the ugly green one nobody is clicking on will drop… we track carfax, owners, price history age on lot, any metric that would add value we measure. 2. "We also have the most used cars (250k more than autotrader and over 500k more than the sites you listed)" How are you calculating this? Because when I did similar searches on CarGurus and AutoTrader, I got more listings on those sites. If you look @ we are only used right now we have not turned on new yet. Autotrader has the 2nd largest @ 1.7M used and cargurus uses data with around 200k less than AT. I can show you how to query them all to get the accurate count. Also being that we aggregate our own data we have noticed that all the sites leave up sold listings. We touch ever car every 24 hours and issue a 24h TTL. So our content is fresh. Not going to name names but most of our competitors have around 10% stale sold listings ours get purged near real time. We also collect data all day long vs batch processing data nightly I have built over 500 dealer websites and know it takes about 48 hours for cars to get indexed. By processing our data all day long we find cars first, just this hour we have processed 1,096 new listings that will not be listed elsewhere till tomorrow or later. 3. Why should a car buyer use your site vs all the other used car sites? We are bigger, faster and offer more intelligent results, we stay completely transparent and take you right to the source. If you didn’t notice when you click the cars we take you to the dealer, eBay, or where ever we found the source of the listing. We don’t have some lame lead trap that users hate. Another thing all our competitors move like dinosaurs we have over 100 commits a week in an agile development cycle where users will see improvements every 2 weeks. We also have an awesome team with members from CERN, NASA, MIT, Microsoft, Autotrader, Dealix, iDealer and Autobytel. If you have any ideas on how we can make even better let us know. We have lots in the pipe but with limited burn things just take time. P.S. Europe Next!
I saw you guys at the Edmunds Hackamotive last year! Fun to see you on Product Hunt. Congrats on the massive progress. The site looks excellent.
@crixlet Thanks, They invited us again this year. We will announce our new mobile product @hackomotive
I was treated to a demo and let me tell you, I've seen nothing like it - to know if that car you're considering is prices right for the local market, for the country... and the UI is beautiful.