GAIT was originally an initiative of the Women Connected program, which mission is to reduce gender inequality in mobile internet and mobile money services.
Aramé Awanis
Developed by the GSMA as part of the Women Connected program, the GAIT toolkit for machine learning and subscriber usage analysis provides mobile and mobile money operators with valuable information on the barriers that prevent women from accessing these services. It is a lever for digital inclusion for women and growth for operators.
What is "GAIT" in short?
Targeting mobile network operators (MNOs) and mobile money providers (MMPs), GAIT is a toolkit for machine learning and analysis of mobile and mobile money usage in a gender-segmented approach.
GAIT provides mobile and mobile money operators with customer usage analysis that can help them better understand the difference in usage between the sexes. The aim is to better understand their customers' practices in terms of cell phone possession and use in order to better respond to their expectations.
It should be noted that the initial version of GAIT did not take into account mobile money data and had no usage analysis function. These two functionalities are now available in the toolbox, following the evolution of GAIT.
Why and how was this initiative born?
GAIT was originally an initiative of the Women Connected program, which mission is to reduce gender inequality in mobile internet and mobile money services. To achieve this, the program works with mobile operators and their partners to remove barriers that prevent women from accessing these services.
The team initially commissioned the GAIT program because of a recurring challenge it faced: the lack of subscriber data, segmented by gender at the MNO level. The lack of such data made it difficult to assess the size of the gap, to conduct a usage analysis based on accurate data, and to measure progress by operators after actions were taken. This is despite the fact that gender data is generally available at the country level through survey data23.
The expected benefits of GAIT
The first step in reducing gender app is to accurately identify the gender composition of operators' customers. By enabling MNOs and MMPs to better understand the gender composition of their subscriber base, as well as gender-disaggregated behavior patterns, GAIT aims to not only close the gender gap in cell phone ownership, but also encourage the adoption of services such as mobile internet and mobile money.
For us, this goal benefits all parties involved:
- Women: greater digital and financial inclusion of women leads to a wide range of social and economic benefits: it opens up access to life- enhancing services such as information, education and sophisticated financial services. In turn, improving women's access to money and mobile internet can improve the livelihoods and overall socioeconomic status of women, who are often disproportionately underserved compared to the general population.
- Mobile operators and mobile money: integrating more female subscribers leads to pure growth in revenues and market share. Beyond "simple" cell phone ownership, the use of sophisticated services, such as mobile internet and mobile money, also means higher average revenue per user.
How does it work on a technical level?
GAIT has two key functions:
Gender identification, designating each customer as female or male. This function is performed by machine learning. It is implemented in three main steps:
- Step 1: identify the gender of a representative subset of clients through a carefully designed telephone survey.
- Step 2: Build a predictive model based on the cell phone usage habits of this precisely identified subset of customers. The measures of cell phone usage used to compute this model include call detail records (CDRs, including voice and SMS), internet usage, and/or mobile money data.
- Step 3: apply the predictive model to the rest of the customer base, qualifying each customer as female or male.
Usage analysis: a platform with a set of data dashboards to better understand customer habits and uses and inform strategic decisions. These are personalized tables on the uses of GSM customers (airtime, SMS, data) and / or mobile money, segmented by gender. The platform includes three menus: GSM, Mobile Money and GSM vs. Mobile Money - the latter allowing users to observe correlations between voice, data and payment usage.
What are the planned evolutions, if any?
We have just updated the initial version of GAIT to add the following features:
- Mobile money data management, in addition to purely GSM data. These data allow us to measure in particular the gap between the ownership of a mobile money account and its actual use.
- Analysis of usage, thanks to a set of customizable dashboards.
- Simplification of use, which no longer requires advanced computer skills.
Is the development of data and AI impacting GSMA's practices?
GAIT was originally an initiative of the Connected Women program, which mission is to reduce inequalities between women and men in mobile internet and mobile money services.
To achieve this, the program works with mobile operators and their partners to remove barriers that prevent women from accessing these services.
The team initially commissioned the GAIT program because of a recurring challenge it faced: the lack of subscriber data, segmented by gender at the MNO level.
The lack of such data made it difficult to assess the size of the gap, to conduct a usage analysis based on accurate data, and to measure progress by operators following actions taken. This is despite the fact that gender data is generally available at the country level through survey data.
What about telecom operators?
Again, in the context of the pandemic, we have seen multiple initiatives by different operators to mitigate the impact of the pandemic, especially in developing countries.
1For more information about the first version of GAIT : https://www.gsma.com/mobilefordevelopment/wp-content/uploads/2018/09/GSMA-Gender-Analysis-and-Identification-Report-GAIT-August-2018.pdf
Extract from our whiter paper : Challenges and advancements in the era of data and artificial intelligence
