SMS Statistics Explained

There are a number of supply and demand side indicators useful for analyzing the SMS market. On the supply side, the International Telecommunication Union (ITU) has identified the number of SMS sent as a relevant indicator to collect. The definition is provided in the ITU's Handbook for the Collection of Administrative Data on Telecommunications/ICT. The key point is that the definition refers to SMS "sent from mobile handsets to other mobile handsets (peer-to-peer), both to national and international destinations." The data are supposed to exclude "SMS received, value-added SMS, premium SMS and SMS sent from web-based services." In practice a number of countries do not strictly adhere to this definition, hampering comparability. 

ITU SMS Sent definition (Click to enlarge)

Countries where receiving party pays (RPP) has been the norm (i.e., operators charge users for both outgoing and incoming calls and texts) typically report both messages sent and received. This is the case for example for Canada, the United States and Singapore. The number of messages sent should roughly match the number received so the figure could be divided by two. There might be an imbalance between international SMS sent and received but in practice, overseas SMS accounts for only a small portion of the total. What is a bigger issue is whether the total traffic represents an accounting or technical measure. If it is an accounting measure, then the received text messages sent within a network (on-net) might not be included since the operator receives no payment for them. In that case, dividing the total traffic by two would underrepresent the sent messages. 

ANACOM, Mobile Service (click to enlarge)

Most SMS messages are to and from mobile subscribers. Take Portugal where 99.2% of text messages are between domestic mobile subscribers with international, Value-Added, Machine to Machine (M2M) and the fixed network accounting for a small proportion.

Most countries use a technical rather than accounting measure to record SMS traffic. This is because text messages can be bundled into a plan. The technical measurement ensures that all bundled SMS are recorded. Some countries use other methods. For example in Turkey, text messages were counted upon their arrival at the message switching center. However this resulted in text messages in Turkish being counted as more than one message. From 2014, messages were counted according to billing. As a result, SMS traffic dropped 18.2% in the first quarter of 2014 (compared to a drop of 5.7% the previous quarter and -7.2% the subsequent quarter).

The number of text messages sent per mobile subscriber per month can be calculated by dividing the number of text messages by the average number of subscribers (see this post for the top 10 in 2014).

Another ITU indicator is messaging revenue. Here the definition includes both SMS and MMS. MMS accounts for just a small proportion of total text so it is not a major issue that the revenue streams are combined. What is problematic is that it is becoming increasingly difficult to account for messaging revenue due to growing bundling of voice, message and text services.

ITU messaging revenue definition, (Click to enlarge)

The revenue data allow the Average Revenue Per Text to be calculated. This is calculated for India where the underlying data is published by the Telecom Regulatory Authority of India (India). In the 2nd quarter of 2015, ARPT was 0.24 US¢.  SMS accounted for 2.4% of mobile revenues in India during that period.

Click to enlarge

ITU SMS price definition (click to enlarge)

Other pricing indicators include the actual tariff for an SMS. Here the ITU asks for the prices of an on-net and off-net pay as you go SMS. Prices for some African nations are shown in this post.

The SMS Termination Rate (STR) is the amount that one operator charges another to deliver a cross network message. In some countries, Sender Keeps All (SKA) is used whereby operators keep all of the revenue from and do not charge for delivering cross network texts. Other countries have STRs that are either agreed amongst operators or established by the regulator. The Body of European Regulators for Electronic Communications (BEREC) publishes STRs for their region (see this post).

Pew text messaging (click to enlarge)

On the demand side, the percentage of mobile users who send a text message reflects the popularity of the service. A few countries compile this. The Pew Research Center has compiled data on SMS users in a number of countries with the proportion of mobile users sending text ranging from 98% in the Philippines to just 37% in Pakistan. This indicator is useful for gauging demand and analyzing why SMS usage is so much more popular in some countries than others (e.g., pricing, literacy, etc.).The intensity of usage can also be obtained from surveys. For example in France, texters are asked how many they send a week (101 in 2014, see post).

SMS can be contrasted with other messaging methods. Ongoing panel surveys from ONTSI in Spain found that in the 1st quarter of 2014, only 5.5% of Spaniards 15 years and older sent a text message every day compared to 22% who used their mobile phone to send an email (ONTSI, 2015, “Las TIC en los hogares españoles”, http://www.ontsi.red.es/ontsi/es/estudios-informes/xlvi-oleada-del-panel-hogares-“las-tic-en-los-hogares-españoles”-4t2014)

The purpose of SMS could also be derived from surveys. A Canadian survey found that of those who text, 18% use SMS to contact family each day and 32% to contact friends (Statistics Canada, 2014, Canadians’ connections with family
and friends, http://www.statcan.gc.ca/pub/89-652-x/89-652-x2014006-eng.htm).

Cellphone owners rather than subscriptions key mobile access policy indicator

Mobile subscription data becoming less useful for gauging mobile penetration
73% of world population owned a cellphone in 2011

Given that mobile communications is widely acknowledged as being the main Information and Communication Technology tool in developing nations, reliable figures on the level of penetration is important. However multi-SIM card ownership, lapsed subscriptions that are still counted for a time and the growing number of machines connected to mobile networks is increasingly blurring the relevance of mobile subscription based penetration figures (See: http://www.ictdata.org/2010/07/more-mobiles-than-people.html).

By the end of 2011, there were 93 countries with more mobile phone subscriptions than people.

Wireless Intelligence, a consultancy specializing in mobile numbers, recently acknowledged the mismatch between theory and reality:

“The assumption that the number of connections reflects the number of individuals subscribed to a network is increasingly a misleading one.” https://wirelessintelligence.com/analysis/2012/10/global-mobile-penetration-subscribers-versus-connections/354/

Wireless Intelligence uses several assumptions to arrive at a figure for their estimate of unique mobile subscriptions for 2012: 3.2 billion (or only 45% mobile user penetration compared to a 95% mobile subscription penetration). Their research is based on 39 countries, representing about 75% of global mobile connections.

There is a more accurate way to gauge mobile phone penetration: using surveys asking individuals if they own a mobile phone. A World Bank project carried out a survey in 2011 of households in rural areas of three Chinese provinces. Some 85% of respondents reported having a mobile phone. Of those that did not, almost a third stated that the reason was because they did not need one. Just over one in five reported that they could use another person's cellphone.

Results of 2011 household survey in rural areas of three Chinese provinces, World Bank 

Recent cellphone ownership surveys have been carried out in almost 100 countries. Based on this data—representing some 90% of the world's population and mobile subscriptions—73 percent of the world's population (or 4.3 billion people) owned a mobile phone in 2011.

There are only 12 countries where 95% or more of the surveyed population owns a mobile phone.

Mobile phone ownership is ubiquitous in only around a dozen countries where 95% or more of the survey population claims to have a cellphone. These are mainly Middle Eastern and Nordic nations. Cellphone ownership was less than half the population in 16 countries where survey data is available.  The three economies where mobile subscriptions most outnumber actual cellphone ownership are Hong Kong, Botswana and Russia. The three where cellphone penetration is most understated based on subscription data are Haiti, Iraq and China. While cellular networks have had a huge impact on access to communications, there is still some distance to go to reach ubiquity.

Cellphone ownership, % of surveyed population, 2011

	Mobile	% population
	subscriptions	owning
	per 100 people	cell phone
Bahrain	150	99		15+
Qatar	130	99	*	15+
Albania	96	99		15+
Hong Kong	208	98	*	16-60
Iraq	78	98		15+
Haiti	42	98		15+
Finland	166	96	*	15+
Sweden	120	96		16-60
Spain	115	96		18+
Israel	121	95		18+
Jordan	120	95		18+
Latvia	103	95	*	15+
Italy	152	94	*	15+
Luxembourg	149	94	*	15+
Denmark	127	94	*	15+
Czech Republic	122	94	*	15+
Armenia	96	94		15+
Estonia	140	93	*	15+
Netherlands	119	93	*	15+
China	73	93		18+
Cyprus	134	92	*	15+
Austria	155	91	*	15+
Lithuania	153	91		18+
Macedonia	110	91		15+
Slovenia	107	91	*	15+
Ireland	107	91	*	15+
Uzbekistan	89	90		15+
United Kingdom	130	89		18+
Belgium	114	89	*	15+
Australia	109	89		18+
Germany	133	88		18+
Greece	108	88	*	15+
Iran	74	88		15+
Kazakhstan	140	87		15+
Morocco	114	87		12-65
Korea, Rep.	107	87	*	3+
Russia	180	86		18+
Croatia	116	86		15+
Japan	102	86		18+
Vietnam	143	85		15+
United States	106	85		18+
France	105	85		18+
Bosnia and Herzegovina	82	85		15+
Ukraine	122	84		18+
Portugal	115	84	*	15+
Azerbaijan	111	84		15+
Turkey	91	84		18+
Kyrgyz Republic	102	83		15+
South Africa	127	82		18+
Georgia	99	81		15+
Philippines	91	81		16-60
Turkmenistan	64	80	*	15+
Lebanon	85	79		18+
Poland	129	78		18+
Brazil	124	76		10+
Bulgaria	140	74	*	15+
Canada	75	74		18+
Kenya	66	74		18+
Egypt	105	71		18+
Moldova	104	71	*	15+
Nigeria	59	71	*	15+
Zimbabwe	73	70		15+
Thailand	122	69	*	15+
Afghanistan	56	64		15+
Botswana	157	62	*	15+
Ghana	87	59	*	15+
Mexico	86	57		18+
Indonesia	98	55		18+
Rwanda	44	54	*	15+
India	74	53		18+
Uganda	47	52	*	15+
Bangladesh	51	49		15+
Pakistan	62	48		18+
Ecuador	102	47		16+
Senegal	70	46	*	15+
Cameroon	50	43	*	15+
Sierra Leone	36	37	*	15+
Tanzania	61	35	*	15+
Cambodia	69	33	*	15+
Chad	35	32	*	15+
Burundi	15	32	*	15+
Liberia	45	22	*	15+
Mali	78	21	*	15+
Burkina Faso	51	19	*	15+
Niger	29	18	*	15+
Central African Republic	24	16	*	15+
Myanmar	2	6		15+
AVERAGE	87	73

Note: * = 2010.
Source: ictDATA.org adapted from Pew, Eurobarometer, BBG, Gallup and national government statistics.

Recommended citation:
ictDATA.org. 2013. "Cellphone owners rather than subscriptions key mobile access policy indicator." http://www.ictdata.org/2013/01/cellphone-owners-rather-than.html

More Mobiles than People

At the end of 2009, mobile phone subscriptions exceeded the population in 76 economies; in two economies—the UAE and Montenegro—there were more than two mobile phone subscriptions per inhabitant.

How can the number of mobile phone subscriptions exceed the population?

1. Having more than one subscription. Some users opt to have more than one subscription perhaps because they want one personal number and one business number, or they have one voice and one broadband subscription or they want to avail themselves of cheaper calling rates for on-net calls.
2. Double counting. Some users may have stopped using one mobile network and switched to another but the original operator continues to report them as active.
3. Machine subscriptions. A growing number of devices ranging from Automatic Teller Machines to beverage dispensers use mobile networks to transmit payment information.

Using mobile subscription data to compare countries can be misleading. Take Europe, where a special survey was conducted in 2008 to determine the number of people between the ages of 16-74 who use mobile phones. [1] Iceland and Norway occupied the top two positions yet they only rank 26th and 24th respectively in mobile subscription penetration (out of 31 countries). On the other hand, Greece and Italy, which rank 1st and 2nd in mobile subscription penetration, only rank 26th and 15th in mobile user penetration.

Click to enlarge

It is undoubtedly better to compare mobile performance using the percentage of people who use a mobile phone but most countries do not compile this data. Another alternative is to use the percentage of households with a mobile phone where the availability is better but still not as complete as mobile subscription data.

EUROSTAT. "Special module 2008: Individuals - Use of advanced services: Use of mobile phone" in Information society statistics at http://epp.eurostat.ec.europa.eu/portal/page/portal/information_society/data/database.

Redefining broadband

The United States Federal Communications Commission (FCC) has redefined its definition of broadband. [1] The FCC had used 200 kilo bits per second (kbps) as the threshold and has now increased this by a factor of 20 to 4 Mega bits per second (Mbps). India has proposed that 3-4 Mbps be used as the broadband threshold in a consultation paper arguing "that many bandwidth hungry applications are not getting developed as they see no business model due to restrictive capacity of the Internet in India". [2]

This trend towards higher thresholds for the definition of broadband suggests that the speed used by international organizations—256 kbps—is becoming out of touch with national goalposts and should be reexamined. [3] Nevertheless, most countries do not provide an explicit speed for broadband in their statistics and instead use general criteria such as "always-on" Internet connections or service categories (ADSL, cable modem).

This post will be updated to add the broadband definitions used by different countries as they become available.

See earlier post about this subject.

[1] "In determining whether broadband is being deployed to all Americans in a reasonable and timely fashion, this Sixth Report takes the overdue step of raising the minimum speed threshold for broadband from services in “excess of 200 kilobits per second (kbps) in both directions”—a standard adopted over a decade ago in the 1999 First Broadband Deployment Report ... As an alternative benchmark for this year’s report, and given that this year’s inquiry was conducted in conjunction with the National Broadband Plan proceeding, we find it appropriate and reasonable to adopt instead the minimum speed threshold of the national broadband availability target proposed in the National Broadband Plan. The National Broadband Plan recommends as a national broadband availability target that every household in America have access to affordable broadband service offering actual download (i.e., to the customer) speeds of at least 4 Mbps and actual upload (i.e., from the customer) speeds of at least 1 Mbps. This target was derived from analysis of user behavior, demands this usage places on the network, and recent experience in network evolution." See: FCC. July 20, 2010. SIXTH BROADBAND DEPLOYMENT REPORT. http://hraunfoss.fcc.gov/edocs_public/attachmatch/FCC-10-129A1.doc

[2] "Extrapolating this trend and considering that a household connection is generally used by 3 to 4 persons, the bandwidth requirement per connection is expected to be minimum of 3 to 4 Mbps per household in very near future to support emerging applications." See: TRAI. June 10, 2010. Consultation Paper on National Broadband Plan.

http://www.trai.gov.in/WriteReadData/trai/upload/ConsultationPapers/202/consultationon10june10.pdf

[3] The ITU and OECD define broadband as 256 kbps. See: ITU. March 2010. Definitions of World Telecommunication/ICT Indicators. http://www.itu.int/ITU-D/ict/handbook.html and "OECD Broadband Portal" at: http://www.oecd.org/sti/ict/broadband