Sunday, 20 October 2013

Person to Person digital market places are broken

I was very interested by a number of video blogs this weekend concerning the problems with digital market places.  On Techcrunch Simon Rothman explained about Greylock's $100M dollar fund wanting to invest in this trillion dollar market.  Simon explains that eBay (where he used to work) is not broken but they have moved from their initial person to person buyer seller relationship to professional sellers selling to the public.  This leaves a gap in the digital economy; he says there is a secondary market in personal goods selling which is not being exploited properly.

Separately on the BBC's Bottom-Line, Francios Coumau of Ebay also explained how eBay does $65B of business per year but 70% of this is from professional sellers.  This means that $19.5B is still person to person selling but as Simon Rothman says the "experience selling secondary goods is broken. If you go around and ask your friends have they sold something on eBay recently; generally the answer is no". He estimates that there is a trillion dollars of "stuff you own" sitting in people closets which is not being sold.

Of course Craigslist is a supposed "successful" person to person digital market place.  But their UI has not changed in years and their mobile app (although functional) is extremely poor in its look and feel and integration with other parts of the internet totally non-existent.  In fact CraigsList is a very weak implementation which still gains traction simply because the demand from the public is there.  Craigslist is partially owned by eBay and is reported to have turned over $122M in 2010 and made $99M profit in the same period.

There are other more local person to person market places.  The UK has Gumtree (ultimately owned by eBay) which does a better job than Craigslist at giving users a good buying and selling experience. However, I wonder if eBay really wants these other market places to do that well as ultimately they compete with eBay itself where their $19.5B person to person market is still vastly larger than all their other market places put together.  By buying up these smaller market places, it restricts the chances of the success of new market places by other new startups.  But this restriction will only remain whilst these alternative market places are very boring and poorly implemented - particularly on mobile where new exciting opportunities exist.

Sunday, 6 October 2013

Why would cars need SIM cards - they don't!!

I was reading about the Audi A3 getting LTE (G4 data) connectivity using local SIM cards in the car.  It immediately struck me as complete madness.  Already cars use bluetooth to enable our phones to do voice calls to others.  Why would we do it differently for data.   Well it appears Audi and some other car manufacturers think this is a viable option and of course the operators/carriers would love to sell more SIM cards.  Well they've got as far as releasing the car, but I just don't think its going to go very far.  Of course it is possible for a machine to have a SIM card to enable data connectivity.  For example a lift might have a SIM card to tell its service engineers it needs to be repaired.  However, when a machine always has a human involved and that human is likely to have a phone with bluetooth then there is no need to a separate SIM card.  Of course it assumes the phone has a bluetooth connection and it has been paired with the car, but you do this to enable voice calls.  There is no technical reason why the same paired connection could not be used for an Internet connection...

So  in the Internet of Things there is no need for each device to have a SIM card.  In fact the economics are such that this is highly unlikely as operator contracts are relatively expensive.  The internet of things will be made up of  devices which communicate using bluetooth and wifi to central internet connected hubs; these hubs will have a data contract associated with them.  Where the hubs are related to a user they are likely to be that user's phone.


Wednesday, 18 September 2013

Ericsson's prediction about 50B M2M devices are flawed

A few years ago there was a prediction the world would run out of IP addresses.  This did not happen in the predicted catastrophic way it was thought would happen.  The reason was a local network was created using a technology called NAT which took the pressure off IP addresses. 

Ericsson have predicted that by 2020 there will be 50Billion M2M (machne to machine) devices and that this will cause a strain on the networks and I am sure they're hoping that companies will buy more Ericsson netwrk equipment as a result.

I think that Ericsson predictions about large numbers (50B or even more) of M2M devices are correct but the prediction that the networks will not be able to take the strain are similar to the IP address issue.  Its now becoming clear that local networks and intelligent devices will take the strain.  BLE (Bluetooth Low Energy) will be the equivalent of the NAT for networks, these BLE networks will integrate with intelligent devices to interpret and integrate with the full Internet.  We're now already seeing the initial versions of these devices appearing in the form of iBeacons but this is a very new and rapidly expanding area.  These beacons will grow into a whole new technology and Ericsson needs to consider re-writing its' predictions.

Sunday, 15 September 2013

"IndoorLocal" a new Category to add to "local mobile social"

We've heard a lot about the growth of "Local Mobile and Social" over the past few years.  I think a new category will now be added "Indoor".  I think for short it will be called "IndoorLocal".

Since its now evident that Apple is on the cusp of creating a new eCommerce eco-system based on their iBeacon technology which will surpass and subsume all the work being done in NFC simply because it does everything NFC does and yet does so much more too.

IndoorLocal is a whole new area which will explode in the net few years.

Friday, 28 June 2013

PhoneCount now predicting all the way to 2050

Predicting what's going to happen next month and even next year is difficult in this fast moving world but predicting what it will be like in 2050 is nigh on impossible, isn't it?  But, we at like tackling the impossible.

We've built a model of population increasing and phone/devices increasing and put the two together to give us an eye on the future.

One thing you'll notice in our figures is that the number of devices per person in 2050 reaches on average 3.7 devices per person.  You might think this unreasonable but we think it accurate.  We think people will have lots of mobile devices per person, from in-car, in house etc and back to the humble communication device(phone) too.  In fact the Internet of Things will be driving up the number of devices hugely.  For this reason in the model after 2020 we change the word "phone" to "device".

Obviously there is lots to disagree with here so let us know what you think and have fun playing with it.

Friday, 31 May 2013

Tim Cook makes big mistake and will make the history book of the future for the gaff

Famous people need to be careful about predicting technology future and Tim Cook stepped into the history books this week with the statement about Google Glass

about Google Glass "likelihood of a broad range of appeal hard to see" says Tim Cook
in 1977 Ken Olsen said

"There is no reason for any individual to have a computer in his home."
 and Bill Gates quote is now thought to be a myth

"640kb ought to be enough for anybody."
Time will tell whether or not this statement of Tim's becomes truely false, but even if the current Google Glass does not take off, it's clear to me at least that a close derivative of glass will become a truely successful mass consumer product and Tim's quote will be added to Ken Olsen and Bill Gates future predictions.

Saturday, 18 May 2013

Opinions on Google IO 2013

I’ve watched quite a few hours of the videos from Google IO last week and here is my summary
1) 900 Million android phones and 48B Google Play for Apps downloads   (Apple Announced 50B Appstore downloads the same day- obviously Android will surpass Apple soon)
2) Search is moving to personal context understanding based on user info like my calendar and previous search requests.  Example of this given using voice search where the word “it” based on previously defined nouns are used.  All very powerful stuff.
3) Android Apps can now have 100 Geofences per app per user.  This means an App can internally with efficient battery usage determine when a Geofence is reached and cause a notification
4) Lots of Google Glass stuff.  Glass screen is better than I’d thought with a resolution of 640x480.   Simple APIs connects using REST apis and connected html cards where there are simple menus and small amounts of data at any one moment. But the whole thing is very early... they’re still working on much of the functionality... I’d say another 2 years before it really takes over the world (so long as the privacy people don’t suffocate it) – but then it will really take over the world.  But of course the iWatch will also be around by then as well.
4) Big advances in web technologies.  All very impressive stuff and I’d say very signficant.  They are using new techniques called Web Components which allow the grouping of HTML/CSS/Javascript fragments(within this are things called Shadow DOM and HTML templates).  In this way you can define your own HTML tags with lots of functionality.  Example is Adobe have a single tag which is a PDF viewer (so a web page can add this very easily in part of a screen with a single tag). 
Very good talk by the Javascript team explaining about the massive improvements in performance of the past few years but they’re close to the limit.  They know they need native app performance speeds and the way they’re going to achieve this is by creating a new language called Dart.  Dart is a cross between Java and Javascript but without the performance bottlenecks of Javascript.... they think they can achieve the same speed as Java;   Dart is already working in Chrome and for other browsers it regresses back to Javascript so always works.  Of course Microsoft have something similar called TypeScript but I think its this Google open standards one which will get somewhere.