Tuesday, September 15, 2009

Augmented TV

Most internet-to-TV boxes today (roku, dlink m500, xbox360, ps3, appletv, windows media center) act as a separate video source to the TV. They get content from several sources
  • stream from a commercial source (e.g. Roku, Xbox with netflix, amazon vod, hulu)
  • play from a local or remote store (e.g. Media Center, Tversity)
  • download from a commercial source to local media store and play (AppleTV + iTunes)
  • stream from a UGC source (YouTube)
  • stream P2P (watch webcam on TV, ustream)
  • videophone conf (no real support)
Each of these use-cases has a different setup and interface, different devices are needed along with different remotes. And of course, making this all work with the PRIMARY use-case (regular cable TV) of the TV is harder.

To make this work seamlessly, here's how it should work
- I have a Wifi set-top box into which DishTV, DVD, Xbox inputs go in. Output goes to TV.
- The set-top box has a built in media-server (Tversity?) with a webpage
- I can login to the webpage online and configure my channels.
- Channels are either sat://dishtv.com/ch302 or http://youtube.com/channels/mychannel or http://webcam_ip/ or http://netflix.com/..
- Each channel has a schedule that is time-variant (dishtv schedule) or just a further selector (VOD) with search OR just a single program (daddy's webcam).
- Each channel consists of :
- an array of entries which could be shown in time-sequence (broadcast) or just folder icon-style with categories/search (games) or just a list with search/sort (mp3)
You can then just select a channel from the regular menu and watch the right video.

One of the channels in the TV can be a meta-channel that allows you to view a directory
of public channels. Eventually there would be a "channel creation" ecosystem (much like
the iphone apps directory). People will create channels and post them in the overall channel directory. Folks can stitch regular Youtube videos along with ads in between.

Wednesday, October 18, 2006

Shopify — A shop in minutes, a business for life.

Shopify — A shop in minutes, a business for life.: "You are unique. Your shop could be too.

Shopify lets you build shops with as much style and flare as you see fit. When you signup for Shopify, you can choose from one of our growing number of custom-made designs or get really creative and create your own designs."

Track your orders.

When you login to your store you are presented with a quick overview of your most recent orders.

Organize your inventory.

Organize your products yourself or let Shopify do it for you with Smart Collections.

How much does this cost me?

Shopify is free to use. There are no signup and monthly fees. We charge a 3% commission on successful product sales. If you sell more than $10,000 all further sales for that month will automatically be reduced to 2% commission.

If you don't earn anything we don't either.

Tuesday, July 25, 2006

Spotrunner and the Virtual Advertiser

Spotrunner and Youtube

Spotrunner is an advertising platform for video-spots that are cheap and can be run on cabletv and video sites (Youtube, YahooVideo etc.). Its Adsense for video-ads.

The ad system is
- self-serve
- targetable by category, media-markets
- can choose prebuilt videos and customize them
- provides budgeting, reporting etc.

Spotrunner buys the airtime in local TV markets. The professionally produced ad costs about $500, compared to several thousand dollars otherwise. The airtime itself is extra, of course, and Spot Runner takes a commission. For example, to run 21 spots on The Weather Channel in silicon valley between 6-9am, it costs $42 ($2 per spotrun). The same number of spots on ESPN between 7pm-12midnight costs $1900. SMBs have clear visibility and can make informed choices.

By reducing the friction for a local business to get on TV, Waxman envisions the service opening new doors for small businesses that maybe never would have considered TV advertising before. Their mission is to use the Internet to put the power of TV in the hands of everyone.

Spotrunner along with narrow-casting abilities of Comcast's video-on-demand, Tivo or Youtube, can be used to target an video-ad at a specific person or a specific demographic.
Imagine - a guy could go to Spotrunner today and propose to his girlfriend (21 times !) on Weather Channel for just $542.

This is a real game-changing business ! It is founded by the founders of PeoplePC and Firefly.

The key value-creators are:

1. By making TV spots purchaseable online, they've converted the closed world of TV-ads into an open-market. Online-video will be great when it arrives, but they first have to own the TV-market. Local SMBs want TV ads, because their customers watch TV.

2. The biggest problem with TV ads is that SMBs do not have the skills, time and money to create TV ads. Production costs of TV ads is usually far more than the airtime cost. By templatizing TV-spot-creation using technology, they have made the cost affordable for SMBs.

The big advertisers are not Spotrunner's immediate customers - they are too concerned with brand image to accept templatized low-cost ads. SMBs (for whom price overrides branding concerns) are Spotrunner's target, and the customers of these local business watch TV more than Youtube..

The only issues I can see is that:

1. TV spot inventory is limited today, and that puts an upper bound on their market. Spotrunner cannot compete with the big media-buyers for the best airtime. Hence, Spotrunner will eventuall need online-video to take off. The potential route - Youtube users become channel-viewers, instead of video-viewers, and Spotrunner 10-second spots run in between videos. That could be a great business model for both.

2. Spotrunner's business model uses fixed rate-cards, instead of a bidding process. Local business advertisers are most comfortable with fixed rates and want to know what their money is going to get them (your ad is going to run on History Channel at time X). Bidded models where cost and ad runtimes are unknown is disconcerting to them. Hence I am not surprised by Spotrunner's pricing model. But fixed rate-cards do lead to an inefficient market - overpricing at the less-desirable segments, and under-pricing at the more-desirable segments.

3. Value to be extracted - For advertisers, the value is that they can run an ad at much lower cost than before. For TV channels, they have more buyers, so they can charge more, and sell TV times that were unsold before.

So, who loses ? The ad-producers - advertising agencies, creative directors, TV-marketers and advertising salespeople - all of whom create friction and costs in the process of getting a spot from the business to the TV screen. Also, to a slight degree, viewers lose - they get templatized ads, rather than more creative ads.

Regarding quality of the ads, it is clear that Spotrunner's templatized ads have excellent production quality, but lack orginality and brand specificity (e.g. tonality and color themes of the brand).But does this necessarily have to be the case ?

Can we actually create BETTER ads more efficiently online ?  Can the creative process be automated or outsourced or leverage the community (or all of the above) for better ads ?

Maybe we can create a huge repository of ads - with ideas, messages, images,
and videos - and provide tools that allow creatives to be brought together into an
ad in better ways than what an individual creative director can do by themselves.

Can we create a virtual creative director ?

Along the same line of thought, are there automated tools for creating the best
Adwords marketing text ?

Friday, July 21, 2006

Where's my twin ?

Where's My Virtual Twin ?

Consider a user A. Say I know everything about A - i.e. I know all the situations (s) that A has faced and what choices A made in each situation (links clicked, job applied, items bought, etc. ). This past behaviour of A can be called - P(a,s). Say I have similar information about a large set (N) of users. The set of users is Un and their past behaviour is modeled as a matrix P(n,s).

Now, can I use this past information P to predict A's behaviour in a new situation x ?

This requires a function f, such that F(a,x) = f ( P )

The existence of such a function, assumes :
(1) that user's behaviours are correlated - see Arrow's Impossibity Theorem and Social Choice theory.
(2) that a LOT about user A is known - i.e. the user A's row in matrix P is well populate and that a lot is known about situation s. Many other users have faced situation s, so the column s in P is also well populated.
(3) that f is practically computable.

How big does N need to be and how densely does the P matrix needs to be populated to make good predictions ?

The online portals, search-engines and purchase sites already know a lot about P. How much P does they really keep, analyse and use ? Does any one of them already have a useful subsection of P ? If not, can they share the parts of P that they each know without revealing identities ? Is P valuable enough to drive consolidation in the industry ?

If there is such a computable function f (and there may well not be), it leads to interesting possibilities.

1. You have a virtual twin. Your twin can predict what choices you will make in each situation. It might even tell you how you are going to feel if you click on a particular news story.

2. You can model macro-behaviours - If you can model everybody's behaviour and their reactions to each other's behaviour, then you can predict what the best marketing strategy should be for a given situation or what the reaction of the population will be to a given story.

Monday, July 17, 2006

ComparisonEngines.com » Blog Archive » Shopping.com, Shopzilla, and NexTag BEWARE of Google Checkout

ComparisonEngines.com : CSE's - BEWARE of Google Checkout:

The complex online-traffic arbitrage game

To get traffic, online merchants pay listing fees to the CSEs and per-click fees for higher rankings.

Comparison Shopping Engine's (CSEs) pay Google for SEO traffic, then must convert that traffic to clicks to their merchants.

Large merchants also bid on Adwords, so they also drive up the advertising costs for the CSEs.

CSE's are also run Adsense ads, so they also get paid by Google (or rather its advertisers - usually, the same merchants).

To put this in context, follow the money :
1. Merchants have an ad budget.
2. This money goes to a) Adwords b) CSE c) Adsense
3. The CSE money also goes to Adwords.

Thus merchants are paying money to the agents, to compete
against themselves and drive up their advertising costs. At the moment, with current CPC rates, this is still better than paying for print/TV advertising. However, eventually, at some CPC rate, the economics will self-stabilize (at a macro-level).

Charlene Li's Blog: Google Checkout Supports Its Core Search Business

Charlene Li's Blog: Google Checkout Supports Its Core Search Business

GCheckout gives merchants
1. easier checkout, so more conversions
2. more trust - can compete with more trusted online merchants
3. discount on ads
4. icon on adwords ads, drives more CTR, so higher ranking
5. cheaper transaction fee than paypal and visa

For small online merchants, its a no-brainer. They will ALL offer it. Whether consumers will adopt it or not, is the open issue. At the very least, by nailing the merchants, Google Checkout has a much better chance for success than previous attempts like PayDirect etc.

Monday, September 12, 2005

Diskless laptops

It wont be long before laptops have no disks.
It is now becoming much more economical
to manufacture multi-gigabyte flash chips.
Flash chips are far more compact and less
power-hungry than hard disks, making them
prime candidates for laptops.

For example, consider a laptop with 10 4G Flash
chips vs a laptop with a 40G disk. The flash laptop
would be about $700 extra in storage costs, but
would be much lighter (no disk, much smaller battery),
would boot up and operate much faster and it would
be easy to plugin extra storage.

Samsung is manufacturing millions of these 4G
flash drives for Apple's iPod Nano, so economies
of scale will also drive manufacturing prices down.
Storage per $ in Flash chips is expected
to double by 2006 and quadruple by 2007. At that
point, the 40G flash laptop will only be $200 more
expensive than the hard-disk option. What will
become of the hard-disk manufacturers then ?
They will pretty much have to focus on servers
at that point.

At some point, one must expect that servers that
host transactional applications (database servers,
webservers) will also have a substantial flash component
to boot up quickly and provide quite write capabilities.

Monday, August 08, 2005

FarCaster: SilverFever

FarCaster: SilverFever

Scott Shattuck, a big smart guy who I've had the pleasure of working with
at E-speak, has written a great article on the differing roles of software
architects, evangelists and programmers, drawing a great analogy with
the roles on architects and craftsmen in the construction industry.

Architects in both industries need to be generalists, aware of
all the tools, materials and techniques that can be employed, and must
avoid dogma. Programmers, like craftsmen, need to be specialists,
and need to be able to bring their deep experience in a particular
area to overcome hurdles in that arena. They need to be devoted
to their specialization and promote their tools over others, so that new
uses of these tools and their limits can be discovered.

Another wonderful analogy is the comparison of programming languages
to structural materials. C++ is like steel, great for load-bearing, but
inflexible and hard to work with. Java is like wood, with less load-bearing
capability, but more flexible and easier to work with. HTML and sripting
languages are like putty, even more light-weight, but most flexible.
A good architect must be aware how best to use each of these materials.