Vidteq is a Bangalore based startup that has launched video maps based driving direction for Bangalore.
For instance, you can get the driving directions from forum mall to indira nagar – in text format as well as via video.
There are subtle features implemented in the product – for instance, you can click on the specific direction route and get the video for that (i.e. need not wait for the entire route video).
The monetization is thru’ ads that are contextual to driving direction.
Overall, VidTeq seems interesting – but here are some bigger questions for them.
Data collection is of course manual – they have a video capture team whose job is to go and capture videos of all the roads (Main and Internal) in Bangalore city.
We have a camera mounted vehicle which gets used for this purpose. What they get is the raw video. The raw video then goes through stabilization process before it gets delivered to the Media team for editing. Once edited, these videos get geo-tagged and stored. We have built a whole set of proprietary tools to handle the complete process of geo-tagging the Video clips. This is true with Images as well.
Video Maps: Fancy or Need
While the experience of Vidteq is good, the bigger question is whether their video maps is solving a bigger need or it’s just a fancy thing to do?
The challenge with Indian roads is the lack of structured data and hence not a single entity in this space has got the lat/long/data right. Routeguru has implemented landmark based driving direction for NCR – but these are highly operational intensive operations – and that’s why scaling up is a challenge.
The point that I am trying to make is whether it makes sense to invest in building video maps – instead why not geo tag the metro cities (navigational as well as PoI data)? Why not focus on getting the landmark-based-driving-direction first?
Do video maps really solve a bigger pain? I would like to believe that, but my inclination is towards well implemented text-based driving directions.
What’s your opinion?
Also check out: latlong