One particular part of the IAMAI-IMRB research looked particularly questionable to us. The Internet industry association-research agency combine, last week, leading up to its annual India Digital Summit, released data projecting that:

1. India will have 213 million Mobile Internet users by June 2015, and there were 173 million mobile Internet users in India in December 2014.
2. The average monthly spend on Mobile Internet is now Rs 235, while the average monthly bill is Rs 439.
3. The proportion of this amount spent on Mobile Internet was 45% last year, and has now increased to 54% this year.
4. 63% of the mobile Internet users spend between Rs 101 and Rs 500 monthly on their mobile connection while 26% spend between Rs 501 to Rs 1000, and 7% of the users spend less than Rs 100 every month.
5. 33% of the Mobile Internet users avail rental plans with limited downloads, followed by 21% who subscribe to unlimited rental plans.
6. Data reveals that 26% of the non-working women access rental mobile Internet plans which allow unlimited access to Internet. A similar plan is also availed by one-fifth of college going students and Working women.
7. 1/4th of school going kids and one-fifth of the non-working women use mobile Internet on Pay per Site basis.
8. Nearly 40% of the young men, working women as well as older men access mobile Internet plans which entail limited access to Internet.

Why we’re questioning this data

1. Average monthly spends on data is clearly inaccurate: Rs 235 being spent monthly on Mobile Internet, even by Mobile Internet users, is incorrect. Bharti Airtel and Idea Cellular, two of India’s top 3 telecom operators have reported a Data ARPU (Average Revenue Per User) that contradicts this. Data ARPU for Airtel’s data customers was Rs 150 at the end of the September 2014 quarter, while for Idea, it was Rs 119. In addition, the IAMAI-IMRB figure of 54% of mobile billing spends on Mobile Internet appears hard to believe, but telecom operators haven’t released such a split for Mobile Internet users, so we don’t have the data to back up our opinion that it is inaccurate.

Mobile-internet-expenditure

mobile-internet-plans-india

IAMAI-IMRB’s response: “This figure is based on the survey done among active mobile Internet users in top 35 cities and are not projected to the all India urban universe. The ones reported by the Telcos are based on the entire universe.”

MediaNama’s take: The report, however, does extrapolate its data to the entire universe. This means that the IAMAI research cannot be extrapolated to the entire universe of Mobile Internet users, and should only be seen as data pertaining to its actual sample size (more on that later).

idea-data-arpu

2. The research doesn’t distinguish between WAP portal usage and Mobile Internet usage

Does an average user know the difference between the Internet and a WAP portal?

IAMAI-IMRB’s response: “As mentioned earlier, we ask the respondents if they access the Internet at least once a month. So, we do not distinguish between operator WAP users or Mobile Internet Users but only capture their Internet usage on mobile.”

india-mobile-internet-users

MediaNama’s take: This is important, because a significant number of users access WAP portals and websites, which is essentially incidental usage. Users who access an Airtel Live or a portal via a WAP Push (message sent with a WAP portal link) may not know the difference between open Internet access and a walled garden like Airtel Live. This would be misleading, in the context of the wider Internet.

A few quarters ago, both Airtel and Idea Cellular, removed ‘incidental and accidental’ users which used less than 1 MB a month, which led to a decline in its reported mobile connection base. Here’s data for Airtel, and for Idea.

*

The point about higher costs stands out for us, and we would recommend that our readers use this data with caution: it shouldn’t be treated as being indicative of Mobile Internet usage in India. The inaccuracy of sampling based data raises question of how the research was conducted and hence about the research itself.

Update: Deepak Abbot, in the comments, also raises some pertinent questions, based on what is in the report.

As per this report, there are 210mn overall Internet users and 159mn mobile internet users (Page 4 – Figure 2)

Why I feel this number is wrong:
There are roighly 60mn 3G subscribers (Not users)
There are 90mn 2g subscribers (not users)
There are 20mn broadband connections at home
Roughly 10mn enterprise users have access to Internet

All these put together give us a figure of 180mn subscribers.

Even at 60% active (active = meaningful & intentional use of Internet), we arrive at 110mn at max. If we multiply home broadband connection to 3 users then we can add another 40mn to this which still makes it 150mn and not 210mn.

Out of this mobile internet users are not more than 80-90mn

It’s a paid report, which we (MediaNama) don’t have access to, and our take here is based on information shared by the IAMAI via a press release. 

Apart from this,


We sent IAMAI-IMRB some other questions as well on their methodology, among other things:

MediaNama: Would appreciate an outline of the methodology involved in the research, since, as far as I understand, there is no clear way of determining the actual number of mobile Internet users because of multi-SIM use. Would appreciate information on sampling methodology, number of cities, towns and villages, sample size per city, town and village, and process of data collection and estimation.

IAMAI-IMRB: The study was conducted through a random exercise taking into account the electoral rolls as a sampling framework. We cover a total 35 Cities which can be classified as – Top 4 Metros, Next 4 Metros, Small metros, non-metros, Small towns. Cities covered under the town class are given below:

IAMAI-IMRB-Research-Methodology

As far as the sample size is concerned, we have covered an average sample size of nearly 1000 households per city in top 4 metros, 800 households per city in next 4 metros, 500 households per city in small metros, 400 households per city in non-metros and small towns. Care was taken to ensure even geographical spread in identifying the starting addresses across the cities selected. The syndicated research for the rural segment is based upon a primary research survey, where we interviewed 3,500 households and nearly 15,000 individuals of various age groups, across SECs and genders from the states of Assam, Maharashtra, Orissa, Tamil Nadu, Andhra Pradesh, Rajasthan and Uttar Pradesh.

We have made the selection of the states taking into consideration different population levels, varying literacy rates, per-capita income and population of disadvantaged groups.

MediaNama: When the IAMAI-IMRB refers to mobile Internet users, does it mean ‘ever active users’ (which the IAMAI-IMRB has used in its reports in the past), annual active users, monthly active users or daily active users?
IAMAI-IMRB: The mobile Internet users in both Urban and Rural India have accessed the Internet at least once a month – Active Internet Users

MediaNama: On what basis are you projecting 11% quarter on quarter increase in user base over the next two quarters?
IAMAI-IMRB: We conduct syndicated survey every year – regarding Mobile Internet we have been collecting data since 2010. Given the trends that we have identified in our past years, we are able to estimate the future growth that is likely to happen, which is more likely a reflection of historical growth witnessed.

MediaNama: What have you done to ensure that there isn’t double counting with respect to multi-SIM and multiDevice usage from individual users?
IAMAI-IMRB: We define a mobile Internet user as someone who has accessed the Internet on their mobile at least once a month. This number is not dependent on number of SIM or the number of Mobile Devices used by a user but on his/her usage.