Review of Telecom API market, pre-conference workshop at the Telecom API Event, 7-9 October. Covering: Home Truths, Market Situation, Why Do We Need Telecom APIs, Reality Check: AT&T, Telus, Etisalat, Telecom Italia, Telefonica, Verizon, Turkcell, Mapping the Landscape, Building the Telecom Application Developer, Landscape, The Litany of Excuses
12. Structure
• Introductions
• Ground Rules
• Home Truths
• Market Situation (2PM)
• Why Do We Need Telecom APIs
• Reality Check: AT&T, Telus, Etisalat, Telecom Italia, Telefonica,
Verizon, Turkcell
o Likely take a break through these case studies (3PM)
• Mapping the Landscape (4PM)
• Building the Telecom Application Developer Landscape
• The Litany of Excuses
• Open Discussion (5PM)
13.
14.
15. Keep in mind throughout the workshop
• Web APIs not Device APIs
• Telecom APIs across the business
o Internal, Partners, Customers, 3rd Parties
• Standards where appropriate (UNI and NNI, everything else is ?)
o Especially if market unproven
o Especially if developed by people who have NO experience in APIs
o Note if anyone accuses me of being anti-standards – I led FSAN GPON, and worked in
ETSI, ITU, ATM Forum and IETF – I know what I’m talking about
• Telecom skills are no longer sufficient, need IT and Web skills / experience
o Multi-disciplinary problem
• Competition is global, competition is beyond other telcos, and we can only
survive as part of multiple ecosystems
• OTT is pejorative (like Redskin), use the term OSP (Online Service Provider) its
what they call themselves
16. No developed market telco
has successfully engaged
mobile application
developers with Telecom
APIs
19. APIs reduce business
friction. This means the
value is not ‘in the API’ it’s
in the service or data
delivered through the API.
20. Mobile Application Developers
ONLY care about direct access to
a large engaged customer base
that is prepared to pay.
Apple and Android fulfill this
need, Telcos are IRRELEVANT
21. Analysys Mason Prediction on Voice Usage
Telco voice usage will remain significant (74% of usage by 2018). New comms services will be
essential to maintain revenue given price pressure and data revenues not offsetting declines.
22. Analysys Mason Prediction on Messaging and ARPU
Telco p2p messaging will become niche, anticipate rapid revenue decline as SMS moves to data.
Analysys Mason is predicting a >25% ARPU decline by 2018.
Industry needs to find >$100B in new revenues within 5 years.
23. Easy and Economical 90%
Global comms clouds
Laggards 10%
Applications and Services
Customers
Telco
What if a Telcos continue to do nothing?
Telcos become the “path of last resort” as apps use “easy and economical” APIs for
90% of comms, and Telcos for only 10%
24. We’re Underestimating the Impact of when WhatsApp adds Voice
Page 24
The Western European Syndrome may be upon us sooner than we think
26. What is an API?
• http://www.telco.com/api.php?action=remove_friction
APIs reduce business friction by making it easy for software systems to work together
using existing well understood web technology that any IT person can understand
27. Why do Telcos need APIs?
Without APIs Telecoms will become irrelevant as Service Providers
because customers will expect communications to be embedded in their
experiences. — Alan Quayle, Independent Telecom Thinker
§ APIs will become critical to maintaining Telecom’s customer relevance
The money is not ‘in the API,’ it’s in the service delivered by the API. APIs
are simply delivering services more efficiently, which opens up new
business opportunities. — Jose Valles, ex-VP Partner Products at Telefónica
Digital
§ APIs are just a technology, its all about the services
An API strategy is becoming a must…in terms of speed to market with
new products, maximizing business development, and product
development opportunities. — Steve Kurtz, VP Business Development, USA
TODAY
§ APIs are a global IT trend across all industries
28. Why do Telcos need APIs?
1995 2000 2005 2010
Why do we need a
Web site?
Of course we have
a Web site
Why do we need
an API?
Of course we have
an API
InnovationUpsell
New business
Operational efficiency Increase footprintAccelerate internal projects
Extend products / services
Make churn harder Partner opportunities
New distribution Device and mobile support
Telecoms is the ‘vital spice’ of any successful business ecosystem
Process automation
29. What APIs can Telcos offer?
Capabilities Opportunities
Telco API
internal systems
billing, rating, charging
lower operational costs
payment services
identity and securitycalling
location
customer insight
customer profile
cloud call centers
enterprise cloud
messaging
device
CRM
VAS
personalization
hypervoice
communication enabled
business processes
directory
fraud management
31. Where’s the Money in External APIs?
Mobile payments revenue, source BI
Intelligence. Includes Apple, Android,
Square, Visa Mobile, etc.$244B by 2017
$157B by 2018 Total Telecom API revenue, source
Mind Commerce. Payment,
Communications, Identity, Cloud, etc.
$18B in 2016
Alan Quayle’s view is the revenues will
be $18B and dominated by payment
and communication service revenues
by 2016
Juniper predicts DCB (Direct Carrier
Billing) growing from $2.5B in 2012 to
$13B in 2017 (this assumes only digital
downloads, not goods and services)
36. Market Situation: API Bait and Switch
API
Roll-up, roll-up, buy my API
software, expose some APIs,
to gain fame and fortune just
like Apple and Google
What is sold
What happens in practice
38. Market Requirements: Why are operators spending
money on API?
• M2M to support provisioning and management
o Verizon (Hughes Telematics), Rogers, AT&T, Telefonica, Ericsson (bought
Telenor assets)
o This is generally treated as a silo, and not part of a broader API business
o And is getting wrapped up in the silly IoT hype
• Wishful thinking in building a developer community like Android and
Apple
o AT&T, Globe, BlueVia, Deutsche Telekom’s Developer Garden
o Market does not yet universally understand this is a failed strategy
• Support open innovation and work more easily with partners on new
business models and market opportunities
o AT&T – this really started only in the last year
39. Market Requirements: Why are operators spending
money on API?
• Support internal innovation, in some cases focused on specific market segments
like enterprise
o Verizon, Telecom Italia, Portugal Telecom
• Support open innovation with specific partners targeting specific market
segments
o Telus – targeting approach, no long tail engagement
• Experimenting in what APIs could means to their business
o SingTel, Rogers
• Build specific business opportunities like direct carrier billing (mobile payments)
o Telefonica and Telenor, Ooredoo, Etisalat, most telcos these days
• Laziness
o Brow beaten into doing something by their vendors
o Copying AT&T
o Following GSMA and OneAPI Exchange
41. Learning
• It doesn’t matter how much you spend, developers are rationale decision
makers
o Telcos continue to struggle in engaging long-tail developers, but developers will
happily turn up to events to take AT&T’s generous prize money at hackathons.
• Telcos have many APIs they can offer beyond voice, messaging, payments
and customer information, including speech processing, authentication,
identity, mHealth, M2M (Machine to Machine), really any service can be
exposed through an API.
o The hard part is building a business around the API, not offering the API.
• Long tail is being redefined as light-weight open innovation.
o That is not trying to build a developer community in competition to Apple and
Google.
o Instead, exploring with select partners, universities, and friendly developers new
ideas or unique capabilities to telcos, e.g. mHealth, connected car, and M2M, and
more (some of which will be at TADSummit)
42.
43. Telus Overview
• A continuing source of SMS revenue growth comes from the enterprise use of
SMS for alerting and notifications
• Telus has implemented a focused API program over the past 7 years – targeting
SMB with SMS alerting capabilities
o Through both direct sales people and local system integration partners
o Value to an enterprise is the ubiquity of communications with its customers in Canada
o Cost is irrelevant as business value far exceeds margin costs of an SMS
• Note SMS is both AT&T and Verizon’s largest by volume API
• Process is designed to enable Telus to launch more apps and faster with a focus
on SME (Small Medium Enterprise)
o Achieved a 4 to 40 annual service launch improvement
o Reduce cost by 75% in launching new apps
• Profitable within the first year of operation
• Now dominated by internal consumption of APIs
44.
45.
46.
47.
48.
49.
50.
51. Background
• The project started on 10th June, 2010
o Within 9 months there were 150 running applications in the market and 10600 registered users.
o Today the total number of network based services developed & available for consumption is >500
• Etisalat has two methods of revenue generation from the services: subscription and on-demand.
o On-demand basis works on the subscriber being charged for every transaction or messaging that they
receive from a service, for e.g. a subscriber sends a message requesting the current exchange rate of his
local currency and he gets a reply message from the service with the information and is charged for
that message only.
o The developer keeps 90% of the revenue.
• Both methods ensure recurring revenue for Etisalat as opposed to one time downloadable fees and
a focus on keeping the customer engaged
• The Developer portal enables the simplified creation of mobile apps for amateur as well as expert
software developers equipped with standard & open APIs, Software Development Kits (SDKs)
covering the major programming languages, sample apps & user guides to direct them.
• Their interaction with the solution is through easy to use interfaces enabled by Web Services that
expose the operator service and network capabilities to allow developers to concentrate only on
developing the app without concerning about the complexities of network protocols.
• Mobile application developers, both amateur and professional, now enjoy a simplified process of
Application Development, deployment and commercialization with mChoiceTM Cloud TAP’s
Mobile Application Developer Portal. Sample Applications, Simulators and Guides help enhance
the application paving the way for the creation of a series of sought after mobile applications.
52. Etisalat Sri Lanka, Emerging Market Telco App Store
USSD API for 100% reach in $1.25 ARPU market
• Top 5 apps subs 3 months after launch:
• Yalu (anon chat) 40,668
• Sinhalalen (local) Jokes 33,787
• Fun Facts 12,807
• Technology News 8,262
• Word Puzzle 5,554
• Business models supported are per
message and subscription
• Accounts for 3.5% of ARPU (2012 #)
55. Key Points
• Developing markets are different – messaging still matters, USSD is a
massive untapped potential for infotainment services in developing
markets.
• Local matters – content local developers, not localized, locally
originated are proving successful with customers because they identify
with the content.
• Size doesn’t matter, Etisalat Sri Lanka only has 3 million subscribers
with low ARPU. Yet they engaged developers creating sticky application
and differentiated their offer in the market.
• The supporting technology is run on a cloud, lowering costs and
allowing greater flexibility.
57. Summary
• Program started in 2008, offered SMS APIs to their content partners.
• Focused across the demand curve, success remains in using APIs internally and
within Telecom Italia’s existing ecosystem of partners / customers.
• Revenue of $250M per year made attributed to the platform, this does not
include the revenue retained by the Lines of Business. Roughly 1B transactions
per month. >1.5B transactions now
• Generally considered within the Telecom community to be the most successful
API implementation of any operator.
o They are Oracle’s largest OCSG (Oracle Communications Services Gatekeeper, API
platform) deployment.
• New division formed in July 2013 called “Service Delivery Platform and NetAPI”,
this appears to consolidate some of the lines of business that were taking most of
the revenue generated by the API platform.
o It is rumored the new division’s annual revenue is >$2B (figure needs to be confirmed).
58.
59.
60.
61. Learning
• Demonstrates the money is not in long tail developers, rather use API
with partners and internally.
• Highlights the importance of using APIs across all possible business
models, internal innovation, partner innovation, not simply focusing on
the “long-tail”.
• Importance of an integrated platform that enables Telecom Italia to
efficiently support high transaction volume partners as well as internal
services enables it to support the relatively high priced solution of
Oracle licenses and Accenture professional service.
• API standards are purely a baseline; they are adapted to meet specific
needs, which means API standards are useless.
o Processes around defining new APIs are key for rapid innovation.
74. Summary
• Long history starting in 1998 of attempting to build developer communities, 2-3
year cycle of launching and closing developer initiatives. BlueVia is its longest
running initiative.
• Internal operational efficiency and long tail developer. Current focus is on the
payment API in cooperation with Telenor and 2 other telcos, targeting partners
and large accounts, not long-tail developers.
• BlueVia has turned around its API business by focusing on the payment API.
• Developers consider the BlueVia program to have failed like every other long tail
telco program in the past.
• Jose Valles, head of BlueVia recently promoted to VP Partner Products at
Telefónica Digital, and now out of the business. Focused on build the payment
service business, adding new APIs particularly in communications, but focus has
moved solidly internal.
75. Learning
• “The money is not in the long tail.” Jose Valles, Head of Blue Via, presented at the SDP Global
Summit . The money for BlueVia is with partners, building powerful partnerships on Telecom
APIs.
• SMS revenue share failed as a business model as the market moved to buying bundles of SMS, so
customers did not want / expect to pay a full price SMS when using the services built on BlueVia
APIs.
• Focus on the core capabilities of a telco, the advertising APIs did not work as TEF lacks the
credibility, inventory, and ability to build a business in this domain.
• Long tail is about light-weight innovation, not trying to build a developer community in
competition to Apple and Google. Instead, exploring with partners, Universities, friendly
developers new ideas / unique capabilities to telcos, e.g. Arduino GSM Shield for M2M over
mobile networks.
• TEF Digital is going to have to re-invent itself again as its filling with corporate politicians. For
example, the TU Go application is being forced on OpCos based on the Jahjah platform it bought
for $100M, this requires expensive integration for the OpCos so they are pushing back.
• And now all the external API stuff is dead
84. Learning
• Copy Tropo or Twilio if you want to compete in the US in communications APIs.
• Don’t chase long tail developers in the back yard of Apple and Google
o Focus on businesses that need to run on Telecom APIs, or using APIs within your
existing ecosystem (Internally and with partners)
• Operators cannot credibly engage the long-tail, need to focus of specific
technologies or domains, e.g. in M2M with relevant M2M companies, it’s a more
a focused open innovation model.
• OneAPI is not relevant to long-tail developers, and looks archaic compared to
modern well-written APIs.
• Money is not in the APIs, 1c per location dip required 100 million dips a month
to make $1M and no developer will pay 1c location these days.
o The money is in the services and solutions enabled by APIs.
85.
86.
87.
88.
89.
90.
91. Mapping Telcos across the API Implementation
Landscape
Internal APIs External APIsBoth
Experiment
Broad
Focused
BusinessUseofAPIs
Organizational Focus of APIs
Likely
Evolution
Path
Telecom Italia does not have everything right, for example, they lack the focus on
building API-enabled businesses, but its closer than most.
92. mapping telcos across the API implementation landscape
Internal APIs External APIsBoth
Experiment
Broad
Focused
BusinessUseofAPIs
Organizational Focus of APIs
93. Mapping vendors across the API landscape
Cloud /
BOSS
Assets
IT /
Service
Assets
Network
Assets
IMS
Assets
Transactional APIs
e.g. call control
Informational APIs
(e.g. customer profile)
Developer
Community
Developer
Portal
API
Management
API
Services
Network
Gateway
API Publishers
Tropo, Twilio,
etc.
API Management (including API Security)
Intel Software (Mashery), CA (Layer 7), Apigee
94. Real World Complexity
Cloud /
BOSS
Assets
IT /
Service
Assets
Network
Assets
IMS
Assets
Transactional APIsInformational APIs
Developer
Community
Developer
Portal
API
Management
API
Services
Network
Gateway
Intel (Aepona),
Ericsson, Huawei,
Oracle, Open Source
Apigee, Intel (Mashery),
Layer 7 (CA), SOA Software,
3Scale, IBM, Open Source
2600Hz, Aculab,
Apidaze,
Bandwidth,
hSenid Mobile,
OnMobile,
OpenCloud,
Plivo, Restcomm,
Solaiemes,
TelAPI, Tropo,
Twilio, etc.
95. Fixed
Voice
($325B)
Mobile
Voice
($615B)
Fixed
Data
($275B)
Mobile
Data
($275B)
Regulated
Services
($1.5T)
Un-‐
regulated
Services
($650B)-‐5
to
-‐7%
5.5
to
9%
3
to
4%
0 to
2%
0-‐2%
Total
Telecoms
Services
($2.15T)
3-‐6%
+ =
Over
the
Top
Messaging
hits
SMS
growth
Mobile
substitution
of
fixed
broadband
with
LTE
OTT
substitution,
saturation,
competition
Mobile
and
OTT
substitution
Sources:
operator
averages
across
developed
and
developing
markets,
supplier
estimates,
Alan
Quayle
1-‐3.3%
Threats
to
Revenue
There’s just 2 things we need to focus on
100. Work with Developers to
embed communications
into applications, services
and business processes.
Make Comms the ‘essential
spice’ of every business
ecosystem.
$40B by 2018 source
Mind Commerce
Work with Developers to
create new services and
applications using
communications
capabilities. Innovate in
communications else watch
revenue decline.
$35B by 2018 source
Ovum
So What Should Telcos Do?
Telecom Application Developers are now essential to addressing the revenue decline
in communication services.
Previous efforts have failed, corrective action is required urgently.
101. Defining what is meant by Telecom App DeveloperRevenue
Product
Internal Telco Developers
Partner Developers
Telecom App Developers
Mobile App Developers
Long Tail Developers
102. This Is A More Accurate RepresentationRevenue
Product
Internal Telco Developers
Partner Developers
Telecom App Developers
Mobile App Developers
Long Tail Developers
103. ‘Cut and Paste’ web developers using
web-scripting and graphical tools on
app platforms
(10Ms)
IT/Web Programmers building
on FOSS, telecom app
platforms and telecom APIs
(1Ms)
What do we mean by Telecom App Developer?
Hardcore
Telco Software
Infrastructure
(10ks)
All based on IT / Web Technologies and Development Principles
Developers that recognize
they build telecom apps
today
104. ‘Cut and Paste’ web developers using
web-scripting and graphical tools on
app platforms
(10Ms)
IT/Web Programmers building
on FOSS, telecom app
platforms and telecom APIs
(1Ms)
TADS is about Building an Ecosystem
Hardcore
Telco Software
Infrastructure
(10ks)
All based on IT / Web Technologies and Development Principles
Developers that recognize
they build telecom apps
tomorrow
105. Examples of Telecom App Dev happening today!
For more details: http://alanquayle.com/category/startups-to-watch/
116. Middle Managers, Desk Jockeys, PowerPoint Politicians, Full-Time Standards
People, People who believe Management Consultants, Wage-Slaves, People
waiting to retire or made redundant, The Risk Adverse, Yes-People, People
who do what the boss tells them regardless, Corporate Anti-Bodies, etc.
Innovators, risk takers, small companies, technology leaders, independent
thinkers, creatives, people working at intersections of industries, etc.
TADS comes from the grassroots
117. The Web has won,
so let’s get with the program
118. BUT no ‘black and white’ thinking!
There’s space for both,
Telecom reaches 6B+ people
120. Developer Needs/
(Responsibilities)
Telco Needs/
(Responsibilities)
Vendor Responsibilities/
(Needs)
Modern and easy to use
tools, docs & support
(Communicate Needs)
Go-to-market support
Viable business models
Meet Dev
tech needs
Support and protect
the telco network
Brokering
Safely expose assets
Help from vendors in
building the necessary
go-to-market support
and business models
Innovation
Community of
developers
(Publicity)
Process that supports
hundreds of new services
(Publicity)
Industry-wide
Best Practices
TAD Manifesto
130. No. We tried a
similar service in our
market and it failed,
and we’re never ever
going to try again
What do you think of this service idea?
131. No. It will not work in
our market. Because
I’m a 50 year old guy
who understands all
my customers better
than they know
themselves.
What do you think of this service idea?
132. No. A feature of your
service overlaps with
an existing.
What do you think of this service idea?
133. No. We have a
similar service
launched, and are not
going to experiment
to make it better or
address other
customer segments.
What do you think of this service idea?
134. No. Our network can
not support such as
service, even though
such services are
going over the top
today.
What do you think of this service idea?
135. No. It looks a bit like
Joyn, we’re not sure
about it, but because
it looks a bit like
something we may do
in the future we’re not
going to do it.
What do you think of this service idea?
136. No. It must work
across all devices,
even though most
devices will never use
it.
What do you think of this service idea?
137. No. We need
additional (random)
features included
before we could
consider it.
What do you think of this service idea?
138. No. It must work on
IMS (even though it
doesn’t need to).
What do you think of this service idea?
139. No. It must work
across all our
customers from day
one, even though
most will never use it.
What do you think of this service idea?
140. No. It must conform
to our process and
design norms. But
we’re not going to tell
you what they are.
What do you think of this service idea?
141. No. It must integrate
with all our existing
platforms, even
though it can work
fine in the current
configuration.
What do you think of this service idea?
142. No. It must be
delivered through our
preferred SI or NEP,
who will copy / kill
the service
immediately.
What do you think of this service idea?
143. No. You must work
through our app
store / portal, which
we’re in the process of
closing.
What do you think of this service idea?
144. No. We can only focus on
4 service launches per
year. We only back major
successes like Video
Telephony, Mobile TV,
Push To Talk, See What I
See…
What do you think of this service idea?
145. No. We just don’t
have the bandwidth,
to do our job.
What do you think of this service idea?
146. No. We have a
network lock-down as
we launch LTE so
cannot do anything
for the next 6-9
month.
What do you think of this service idea?
147. No. Bob has left the
business and we’re
waiting on his
replacement, who
never comes.
What do you think of this service idea?
148. No. We’re waiting on
annual budgets to be
confirmed, sometime
in the next 6-12
months.
What do you think of this service idea?
150. No. Someone in the
organization doesn’t
like such services.
What do you think of this service idea?
151. No. That cannot be
implemented without
changing our IN /
product catalog /
CRM / billing /
network.
What do you think of this service idea?
152. No. We cannot bill /
sell services under $5
per month.
What do you think of this service idea?
153. No. We have a
backlog of 24 months
on billing updates,
even though the
service doesn’t need
to be in that pipeline.
What do you think of this service idea?
154. No. You must work
through our
innovation group who
we all hate and ignore
as they’re parasites on
our business.
What do you think of this service idea?
155. No. You must talk with
Bob who will then pass
you to Bill, who will then
pass you to Mary, who
will then pass you to
Paul, who will then pass
you back to Bob.
What do you think of this service idea?