December, 2009

...now browsing by month

 

Scrapping Features is Hard (Very)

Thursday, December 3rd, 2009

Today at Yahoo! I had the pleasure of hearing Marty Cagan talk about building awesome products and a good question came up.  What companies do a good job at scrapping (aka retiring) features?  Building features is relatively cheap and easy.  Scrapping them after afterwards is very hard.  Netflix attempted to retire their Profiles features (allowing users to create profiles and set controls around on how many movies and the types of movies each profile can rent).  This did not go over well with their community.  Two weeks later they came back and decided to keep it.  Now the feature is buried deep into the Account Settings page.  This is the right way to do it if you absolutely cannot retire the feature. Here is how the sequence of events played out on the Netflix blog and site:

[Update March 18, 2010:  Another example of how hard it is to remove features from NetFlix.  I love that they have the fortitude to remove features even if a small and vocal set of users complain.]

Mobile Data: Ads, Carriers, and Devices

Thursday, December 3rd, 2009

Lots of new mobile data coming out over the past few days.  Since we are operating in an emerging category one has to make certain assumptions about how the mobile market will unfold.  Each new data point can validate or debunk ones assumptions.

Mobile Advertising

  • Mobile coupon redemption will hit $6B by 2014 according to Juniper Research [source].  Unless there is a new compelling scenario (e.g. eBay scenario helped create PayPal) this will be a very slow road.
  • Mobile will grab 11.7% of all digital ad spend by 2014 [source].  Mobile ad monetization is actually an area where lots of product innovation needs to happen.

Carriers and Connections

  • Mobile carrier wireless data revenues grew 27% y/y in Q3 2009 hitting $11.3B. While AT&T (with exclusive iPhone) is growing their data revenues fastest 6% q/q, but Verizon and Sprint are not far behind 5% q/q.[source]
  • T-Mobile (33M customers) even with some nice G1 Android phones lost 77,000 customers for the first time in history [source].
  • WiFi Hotspots hit 1.2B connections growing 47% in 2009 [source].  This is obviously good but big opportunity to provide centralized authentication and billing.  Its crazy that you need to create an account and enter credit card for each WiFi Hotspot provider.

Mobile Devices

  • In 2010 over 1B mobile devices will connect to the Internet vs. 1.3B PCs that are connected to the Internet. Mobile devices growing 2.5x faster then PCs so it won’t be long before more mobile devices are connected to the Internet then PCs. Happening incredibly fast.  [source]
  • 200M “Smartphones” (iPhone, Android, etc.) will ship in 2010 [source].  One smartphone generates as much data usage as thirty feature phones so this is analogous to shipping 600M feature phones [source] In my opinion this divergence will increase dramatically as the content (apps and web services) improve over the coming years.

Monetizing Mobile: You Got Options

Tuesday, December 1st, 2009

On the web; search is king of monetization. Not so in the mobile world. There are a myriad of open (do not require negotiations with operators) and viable monetization tools to choose from:

1.) Upfront License Fee:

  • The most common due to the early success of the iPhone App Store infrastructure.
  • Easy to grok for consumers – you pay $X you get this product.
  • No trial period limits addressable market – some people need to try before buying.
  • In the App Store there has been downward spiraling of prices towards $0.99 – without a runaway hit (1 million downloads) it will be very hard to make a living.

2.) Paywall (aka In-App Purchases)

  • Another form of a license fee, yet a delayed one that encourages some level of trialing.
  • Exposes additional content to the user e.g. a new level of a game, increased functionality, and more content (e.g. more article)
  • Early data indicates a 1.9% conversion rate [source]

3.) Virtual Goods

  • User purchases goods within the service e.g. more poker chips, equipment in a mob game, car accessories in a game, etc. These work well within role playing games where humans dress-up their personalities with paid virtual clothing. Game-makers also intentionally devise ways to slow down key actions within the game triggering an impulsion to purchase goods to save time.
  • Zynga reported <3% conversion rate across all games [source] (I would suspect that mobile games have higher conversion rates given the tight purchasing integration)
  • Com2U reported 35% conversion rate for their HomeRun3D app which far higher then the industry average [source]

4.) Search & Display Ads

  • Most advertisers are simply extending their search & display ads to mobile without customizing them [source]
  • With mobile search volume increasing dramatically (30% q/q and more in countries like Japan) and improved ad customization tools (AdMob has made good progress) advertisers will start to customize their mobile ads [source]
  • If you believe the pundits mobile search ads will dominate 70% of mobile advertising by 2014 [source].  If this happens I believe this is due more to the familiarity advertisers have developed with the search product and convenience of extending their existing ad campaigns to mobile.  Recall, display ads dominated Web 1.0 simply because they felt like ad products advertisers were used to buying in the offline world.

6.) Leads into Brick n’ Mortar Shops (Going in the opposite direction)

  • Most ads aim to connect users to a virtual event – a web page to buy, download, or register for something.  This goes in the opposite direction; it is an ad category based on connecting consumers to a physical presence in order to complete a transaction in a brick n’ mortars establishment.  Other then OpenTable that has built a profitable business in this space everything else is experimental:
    • Click-to-call (Google, Yahoo, and AdMob support yet oddly value a call the same as a click. Calls are less prone to fraud beyond other benefits)
    • Nearby product inventory set-aside (Krillion)
    • Restaurant reservation(e.g. OpenTable)
    • Redeemable coupon/deal (Foursquare, Loopt, and many others)
  • With the exception of OpenTable determining the probability that a consumer completes a transaction in a brick n’mortar shop after clicking or calling is ad-hoc, at best.
  • Apps like FourSqaure and Loopt are experimenting with redeemable coupons that simply require a user to show their coupon (from their phone) at the point of purchase.  This is a viable boot-strapping solution in the short term, yet over the long term it does not provide the level of data accuracy or freshness needed to build profitable campaigns.
  • Getting these ads to scale will require frictionless ad management tools and accurate and real-time data that provide merchants their true customer acquisition costs, otherwise their ad spend will be guess-work.  With retail store operating margins of 2% [source] too much guess work is lethal. Getting to scale means providing advertisers accurate and fresh data that only happens with direct integration with point of sale systems.
  • With millions of OpenTable reservations happening via mobile devices OpenTable can transition from a flat $1.0 per reservation to a dynamic pricing model.  Consider the following scenario:  if a user is on their mobile, a few miles away, looking for a table for three in a few hours when 35% of your restaurant is expected to be empty, and the average diner spends $20/head isn’t that reservation worth more then a $1.0?

7.) Promote Myself Ads (aka User-Based Advertising)

  • A form of a personal classified ad that allows the user to promote the thing they care most about: themselves.  So far, there is nothing inherently mobile about this product yet early data suggests that users are more willing to spend money on this type of product when using their mobile device: a Flirtomatic mobile user is 3 times more likely to spend money with Flirtomatic then a web user [source].
  • Flirtomatic’s user-based-ad products encourages a user to submit bids in order to promotes their profile to the top of the search results for a 4 hr period. The CPM rates for this user-based-advertising are 3-4 times the CPM of brand ads [source].

8.) New Ad Formats: Mobile Video

  • Beyond text and simple display ads there is a market for mobile video ads that provide a deeper sense of the product or service a user is considering.  If your getting off an airplane in JFK after a long flight the last thing you want to do is sit in a smelly town-car for one hour in traffic.  A 15 second video ad showing you the cars and drivers, the number of available cars that could be waiting outside in 15 minutes, and ratings from recent customers would help you decide which car-service to use. Furthermore, it allows the car-service to compete on something other then price — service.