Root ideas, categories & adjective verticals - lessons for the Groupon era
Here are some notes as I tried to fit Groupon’s rapid growth into a pattern of some of the previous internet blockbusters we’ve seen.
The blockbuster models created ‘root ideas’ and were the horizontals in the new category which spawned ecosystems, copycats and derivatives that were adjective verticals.
Blockbuster - Category - Root idea
Google - search - keywords
Technorati - blogs - tags
Facebook - social network - profiles
Twitter - microblogs - streams
Groupon - daily deals - deals
The lesson here is to look beyond the ‘category’ to the root idea and than to understand how the ecosystem around this new root idea would form.
For example, google’s search is both horizontal and vertical in the big distinct categories i.e. images, shopping etc. leaving little room for true search verticals. However, the keywords ecosystem has spawned numerous ad networks, SEO, SEM and other companies and opened up numerous pay-per-click arbitrage models.
So, if one were to apply that to facebook, one would ask is the opportunity in building a vertical social network or in building an ecosystem around rich profiles. Quora has rich expert profiles, Linked-in rich resume profiles etc.
Streams similar to tags are like salt and have completely dissolved into the internet soup and so just as every object now is being tagged and we all ask technorati-who?, every conversation is now a stream. Twitter unlike technorati has profiles, i.e. rich press-release-profiles that will ensure their survival.
Examples of how the Press release is the metaphor to understand twitter.
- breaking news in a far-off and remote spot is essentially a press release by an observer
- as is the fact that the German chancellor is going to make a trip to the United states,
- as is the fact that techcrunch has just released a new blog-post)
Which brings us to today’s Groupon. The category is daily deals, they have infused the word ‘deal’ with specific meaning & characteristics much like google did with keywords.
A whole deal-centric ecosystem is forming and there will be multiple opportunities there. But deal verticals need to be careful, think of Amazon here as it moved out of books into different categories and created an Amazon marketplace. Groupon, LivingSocial and Gilt could become large Deal marketplaces like Amazon, eBay, etsy etc. Opportunistic players are going to find it hard to carve out separate verticals unless they have a unique niche.
As an exercise lets think of how StyleMint.com will evolve. StyleMint is a celebrity-endorsed commerce site recently launced by BeachMint.com. The tagline reads An exclusive t-shirt collection designed by Mary-Kate Olsen and Ashley Olsen.
To be succesful, besides its core competence i.e. celebrity-endorsements, t-shirt design it needs manufacturing, logistics, distribution, shipping. It needs an ecommerce platform, affiliate programs, online marketing, email lists and daily deals. It needs facebook, twitter integration, viral word-of-mouth and branding. All with a deal/shopping2.0 flavor
But since it is not a ‘regular’ ecommerce storefront it can’t use any of the services available such as the the Amazon marketplace, Big Commerce, ebay etc. It can use one of the new emerging storefronts i.e. payvment.com, shopify.com and the like but they are all emerging players and not yet established and feature-rich. So the deep-pocketed StyleMint builds out a unique platform.
However, during the 24 months that they are doing this, Groupon builds out a massive horizontal on daily deals with a huge userbase (70 million today - one can prorate it to about 300m in 24 months) and begins to start adding a ‘celebrity-endorsed’ daily deal add-on. 6 months later StyleMint decides to ‘partner’ with Groupon citing the ‘economies of scale’ in leveraging the superior technology & distribution & myriad other benefits that Groupon affords.
When the mop-up by Groupon of all verticals-of-interest begins to happen the leaders of the ancillary services that are being provided to the motley crew of home-grown verticals will stand to succeed in a big way.
What are those services. Deal re-sellers, Deal managers, Deal white label platforms, Deal ad-networks and other innovative ideas. In other words the deal ecosystem.
the nodejs bird flies over the appengine cloud whistling at Gulliver Google
Looking at the cloud computing landscape how does appengine maintain its value proposition in a world where nodejs provides dramatically better request response times and a slew of nosql databases wait in line to replace and take on BigTable.
Lets take all their recent enhancements
- Always On, Warm Up, 10 minute task limit (all of these are improving internal constraints that don’t exist in your vanilla nodejs host)
- Channel api (nodejs was born to do this)
- Pipeline api (nodejs with flow control libraries do this at cpu speed as opposed to per-http-request taskq speed)
- High Replication DataStore (at an increased cost (3x) all of a sudden makes redundant and scalable riak on ec2 or rackspace more interesting)
So what can Gulliver Google do
1. Add a nodejs frontend to its BigTable datastore. (Google buys joyent.com)
2. Drop-in integration with the various google api properties and google apps making development on appengine a no brainer for consuming google api and other third party api services. (Google buys apigee.com)
It’s getting interesting in Brobdingnag-land Please pass the popcorn!
The shorturl squeeze and loosening of the 140-belt
t.co was announced today and with that the rapid growth of the shorturl the bitly-bubble has burst.
Twitter is ‘asserting control’ over its link dominion. Every link, repeat “every link’ will now be wrapped in a t.co link wrapper that will redirect the user to the original link posted and since t.co links are of a predictable size the 140-character computation will be based on the t.co link size and not the original link size which on the face of it means that tweets can be a bit longer than 140 chars but really means that you don’t need to use shorturls anymore to save tweet space.
The official blog post is at
http://blog.twitter.com/2010/06/links-and-twitter-length-shouldnt.html
The death of the hashtag! At #tanhf the hashtag was diagnosed with ‘annotanitis’ a terminal illness
At #tanhf the hashtag was diagnosed with ‘annotanitis’ a terminal illness, I’d give it 3 months said the tweet-doctor…
The storied hashtag that grew to prominence in a few short years has now been diagnosed with annotanitis.
This is a fatal illness that grows insiduously within the system. The first signs of annotatis are ‘attributii’ these are parasites that piggyback onto the main tweet body and carry deadly payloads of nams and vilues. Hashtags don’t stand a chance as their only defence is the hash that gets swamped by a deluge of nams and villi.
The scientific construct of this hashtag killer has been discovered and it belongs to the json family and the notation of the annotanitis parasite is:
# The attributi payload
attributii = [{“attributii”: {“nam1”: “vilue1”, “nam..n”: “vilue..n”}}]
# The parasite piggyback
”status” “text body host of terminal hashtags”,
”annotations”: simplejson.dumps(attributi)
All clients can expect to be bombarded with hordes of these parasites all attached to host tweet bodies. Resistance is futile, and because the attributi are arbitrary and can mutate at any time, selective accomodative wrappers are needed.
We need: A programmable [feature-pluggable app platform] (for all) Twitter client (s)
Unix had a shell language. DOS had a batch language. Lotus 1-2-3 had its macro language. Emacs is a programming tool as much as it is a text editor. We have gotten out of the habit of making programmable end-user products, but they are still just as important today as they were a couple of decades ago.
- In the near future where there are
- multiple shout servers (the subset of all 140char micro-meso-macro blog services)
- a few shout aggregators (BIN-GLE-TWIT) and
- a few hundred shout clients each with a clever set of features tailored for specific use cases
- What is needed is shoutware i.e. middleware that lets users and web services tie into the shoutstream
Abbreviated apps provide that middleware allowing one to condense the whole app interface into a command-line like shell script that can be invoked simply by typing commands in the shell (input into the textarea or click an icon for the gui-crowd) on a terminal ( a bookmarklet-decorated web page)
So, with a bit of poetic latitude here is an analogy between Unix and the Abbreviated App Web
- Unix shell scripting language ~= Web javascript.
- Unix scripts ~= Web <script =’text/javascript src=’myscript.js’></script>
- Unix binaries ~= Web services
- Unix pipes ~= Web http POST webhooks
- Unix terminal ~= Web page decorated by bookmarklet
- Unix shell environment ~= Web page (e.g. twitter.com homepage)
- Unix commandline ~= Web input box (twitter.com’s status update textarea)
Hello (Hackathon) World!
We’ve entered the era of the 2 day Hackathons, were you conceive an idea, form a team (or go solo), build an app and get it vetted by a savvy group, launch, go viral and start all over again 30 20 days later. We’ve moved from startups to experiments that yield results at a breathtaking pace and allow one to leverage trial and error to find and follow the rich veins. Startups need business plans, angels, vc’s etc., experiments need ideas, sweat and a Lab. 565 Labs is that kind of a Lab which tests out many experiments and picks, chooses those that show promise for further growth and development.
Root ideas, categories & adjective verticals - lessons for the Groupon era
Here are some notes as I tried to fit Groupon’s rapid growth into a pattern of some of the previous internet blockbusters we’ve seen.
The blockbuster models created ‘root ideas’ and were the horizontals in the new category which spawned ecosystems, copycats and derivatives that were adjective verticals.
Blockbuster - Category - Root idea
Google - search - keywords
Technorati - blogs - tags
Facebook - social network - profiles
Twitter - microblogs - streams
Groupon - daily deals - deals
The lesson here is to look beyond the ‘category’ to the root idea and than to understand how the ecosystem around this new root idea would form.
For example, google’s search is both horizontal and vertical in the big distinct categories i.e. images, shopping etc. leaving little room for true search verticals. However, the keywords ecosystem has spawned numerous ad networks, SEO, SEM and other companies and opened up numerous pay-per-click arbitrage models.
So, if one were to apply that to facebook, one would ask is the opportunity in building a vertical social network or in building an ecosystem around rich profiles. Quora has rich expert profiles, Linked-in rich resume profiles etc.
Streams similar to tags are like salt and have completely dissolved into the internet soup and so just as every object now is being tagged and we all ask technorati-who?, every conversation is now a stream. Twitter unlike technorati has profiles, i.e. rich press-release-profiles that will ensure their survival.
Examples of how the Press release is the metaphor to understand twitter.
- breaking news in a far-off and remote spot is essentially a press release by an observer
- as is the fact that the German chancellor is going to make a trip to the United states,
- as is the fact that techcrunch has just released a new blog-post)
Which brings us to today’s Groupon. The category is daily deals, they have infused the word ‘deal’ with specific meaning & characteristics much like google did with keywords.
A whole deal-centric ecosystem is forming and there will be multiple opportunities there. But deal verticals need to be careful, think of Amazon here as it moved out of books into different categories and created an Amazon marketplace. Groupon, LivingSocial and Gilt could become large Deal marketplaces like Amazon, eBay, etsy etc. Opportunistic players are going to find it hard to carve out separate verticals unless they have a unique niche.
As an exercise lets think of how StyleMint.com will evolve. StyleMint is a celebrity-endorsed commerce site recently launced by BeachMint.com. The tagline reads An exclusive t-shirt collection designed by Mary-Kate Olsen and Ashley Olsen.
To be succesful, besides its core competence i.e. celebrity-endorsements, t-shirt design it needs manufacturing, logistics, distribution, shipping. It needs an ecommerce platform, affiliate programs, online marketing, email lists and daily deals. It needs facebook, twitter integration, viral word-of-mouth and branding. All with a deal/shopping2.0 flavor
But since it is not a ‘regular’ ecommerce storefront it can’t use any of the services available such as the the Amazon marketplace, Big Commerce, ebay etc. It can use one of the new emerging storefronts i.e. payvment.com, shopify.com and the like but they are all emerging players and not yet established and feature-rich. So the deep-pocketed StyleMint builds out a unique platform.
However, during the 24 months that they are doing this, Groupon builds out a massive horizontal on daily deals with a huge userbase (70 million today - one can prorate it to about 300m in 24 months) and begins to start adding a ‘celebrity-endorsed’ daily deal add-on. 6 months later StyleMint decides to ‘partner’ with Groupon citing the ‘economies of scale’ in leveraging the superior technology & distribution & myriad other benefits that Groupon affords.
When the mop-up by Groupon of all verticals-of-interest begins to happen the leaders of the ancillary services that are being provided to the motley crew of home-grown verticals will stand to succeed in a big way.
What are those services. Deal re-sellers, Deal managers, Deal white label platforms, Deal ad-networks and other innovative ideas. In other words the deal ecosystem.
the nodejs bird flies over the appengine cloud whistling at Gulliver Google
Looking at the cloud computing landscape how does appengine maintain its value proposition in a world where nodejs provides dramatically better request response times and a slew of nosql databases wait in line to replace and take on BigTable.
Lets take all their recent enhancements
- Always On, Warm Up, 10 minute task limit (all of these are improving internal constraints that don’t exist in your vanilla nodejs host)
- Channel api (nodejs was born to do this)
- Pipeline api (nodejs with flow control libraries do this at cpu speed as opposed to per-http-request taskq speed)
- High Replication DataStore (at an increased cost (3x) all of a sudden makes redundant and scalable riak on ec2 or rackspace more interesting)
So what can Gulliver Google do
1. Add a nodejs frontend to its BigTable datastore. (Google buys joyent.com)
2. Drop-in integration with the various google api properties and google apps making development on appengine a no brainer for consuming google api and other third party api services. (Google buys apigee.com)
It’s getting interesting in Brobdingnag-land Please pass the popcorn!
The shorturl squeeze and loosening of the 140-belt
t.co was announced today and with that the rapid growth of the shorturl the bitly-bubble has burst.
Twitter is ‘asserting control’ over its link dominion. Every link, repeat “every link’ will now be wrapped in a t.co link wrapper that will redirect the user to the original link posted and since t.co links are of a predictable size the 140-character computation will be based on the t.co link size and not the original link size which on the face of it means that tweets can be a bit longer than 140 chars but really means that you don’t need to use shorturls anymore to save tweet space.
The official blog post is at
http://blog.twitter.com/2010/06/links-and-twitter-length-shouldnt.html
The death of the hashtag! At #tanhf the hashtag was diagnosed with ‘annotanitis’ a terminal illness
At #tanhf the hashtag was diagnosed with ‘annotanitis’ a terminal illness, I’d give it 3 months said the tweet-doctor…
The storied hashtag that grew to prominence in a few short years has now been diagnosed with annotanitis.
This is a fatal illness that grows insiduously within the system. The first signs of annotatis are ‘attributii’ these are parasites that piggyback onto the main tweet body and carry deadly payloads of nams and vilues. Hashtags don’t stand a chance as their only defence is the hash that gets swamped by a deluge of nams and villi.
The scientific construct of this hashtag killer has been discovered and it belongs to the json family and the notation of the annotanitis parasite is:
# The attributi payload
attributii = [{“attributii”: {“nam1”: “vilue1”, “nam..n”: “vilue..n”}}]
# The parasite piggyback
”status” “text body host of terminal hashtags”,
”annotations”: simplejson.dumps(attributi)
All clients can expect to be bombarded with hordes of these parasites all attached to host tweet bodies. Resistance is futile, and because the attributi are arbitrary and can mutate at any time, selective accomodative wrappers are needed.
We need: A programmable [feature-pluggable app platform] (for all) Twitter client (s)
Unix had a shell language. DOS had a batch language. Lotus 1-2-3 had its macro language. Emacs is a programming tool as much as it is a text editor. We have gotten out of the habit of making programmable end-user products, but they are still just as important today as they were a couple of decades ago.
- In the near future where there are
- multiple shout servers (the subset of all 140char micro-meso-macro blog services)
- a few shout aggregators (BIN-GLE-TWIT) and
- a few hundred shout clients each with a clever set of features tailored for specific use cases
- What is needed is shoutware i.e. middleware that lets users and web services tie into the shoutstream
Abbreviated apps provide that middleware allowing one to condense the whole app interface into a command-line like shell script that can be invoked simply by typing commands in the shell (input into the textarea or click an icon for the gui-crowd) on a terminal ( a bookmarklet-decorated web page)
So, with a bit of poetic latitude here is an analogy between Unix and the Abbreviated App Web
- Unix shell scripting language ~= Web javascript.
- Unix scripts ~= Web <script =’text/javascript src=’myscript.js’></script>
- Unix binaries ~= Web services
- Unix pipes ~= Web http POST webhooks
- Unix terminal ~= Web page decorated by bookmarklet
- Unix shell environment ~= Web page (e.g. twitter.com homepage)
- Unix commandline ~= Web input box (twitter.com’s status update textarea)
Hello (Hackathon) World!
We’ve entered the era of the 2 day Hackathons, were you conceive an idea, form a team (or go solo), build an app and get it vetted by a savvy group, launch, go viral and start all over again 30 20 days later. We’ve moved from startups to experiments that yield results at a breathtaking pace and allow one to leverage trial and error to find and follow the rich veins. Startups need business plans, angels, vc’s etc., experiments need ideas, sweat and a Lab. 565 Labs is that kind of a Lab which tests out many experiments and picks, chooses those that show promise for further growth and development.
