Marketing Analytics in Emerging Markets: Evaluating the Impact of Access to Analytics for Micro Entrepreneurs in Rwanda

The key idea of our research project is – “How can knowledge of analytics have an impact on small scale entrepreneurs and their business?” Specifically, we seek to improve the growth of small firms in emerging markets by addressing a significant constraint they face: access to marketing information & marketing analytics. Few micro and small enterprises formally record their marketing data (e.g. sales, products, customers) which hampers their ability to extract meaningful insights from business information.

We have designed an easy-to-use mobile application tool that leverages a web-interface to increase an entrepreneur’s ability to track, access and take action on business intelligence – which includes activities such as: recording accurate data, storing and retrieving information, performing analytics, reporting relevant metrics, and tracking and comparing performance over time. For most entrepreneurs in low-income countries, having access to this type of business intelligence represents a novel capability and, thus, could lead to lasting improvements in formal record keeping. Moreover, by making salient the critical business metrics and benchmarks of one’s own firm, this business intelligence tool may also nudge entrepreneurs to think and behave differently. To the best of our knowledge, researchers have not yet explored the impact of providing entrepreneurs with tools that improve access to information on their own marketing practices and performance. The few related ‘access to information’ studies (Conley and Udry 2010, Jensen 2007) do provide critical insights on the role of learning, but these papers do not touch on the growth of small businesses or the role of ICTs (information communication technologies) in enhancing an entrepreneur’s usage of information about her own firm.

We conduct a randomized controlled field experiment (RCFE) to rigorously evaluate the impact of improved access to business intelligence for a sample of growth-oriented micro and small business owners in Rwanda. We evaluate the impact of the “MARKETmanager” information technology tool: this tool will focus on metrics related to marketing in the business (e.g. sales, products and customers, as well as policies/practices related to pricing, promotions, selling, etc.). Each entrepreneur in our sample is provided a secure web profile where she can regularly view summary reports of her current business information, along with basic ‘analytics’ that compute performance ratios and compare this information to performance in previous periods. (See Appendix for mock-ups of the “MARKETmanager” tool.) By manipulating entrepreneurs’ access to information on the marketing business function, we are interested in determining whether our main outcomes of interest (including firm survival, employment, sales, profits and return on assets) are affected differently, as well as the process mechanisms through which these differential effects occur (i.e. theory of change). The results will be beneficial to marketing managers, entrepreneurs as well as policy makers interested in growth of firms. Another major contribution of our work is that we look at analytics from a small firm perspective. When it comes to driving success in emerging markets, large companies often depend on small businesses for driving sales to the end consumer (e.g. retailers, mom and pop shops etc.) and one of the key contributors in this regard that could drive success is the ability to leverage data and analytics well.

2019-07-01

2020-05-31

Main effect is measured on - sales and profits (firm growth), in addition we will also try to identify the mediating mechanisms leading to the effect (if any).

Our sample consists of 550 growth-oriented entrepreneurs. The baseline data was collected and verified for this set of firms (n=550),and the participants were randomly assigned into one of three experimental groups:

1. Group 1 (treatment): 250 entrepreneurs receive a smartphone with the “MARKETmanager” tool, a mobile app tool that is focused on tracking key business metrics related to entrepreneurs’ management of marketing activities (e.g. products and customers, and related policies/practices).

2. Group 2 (pure control): 250 entrepreneurs form a counter-factual group that does not receive any intervention during the project (note: they will be given access after the project ends).

3. Group 3 (placebo): 50 entrepreneurs form a second comparison group that receives only a smart phone and data plan (they will not be given any marketing analytics apps, tools, intervention, etc.)

Our sample consists of 550 growth-oriented entrepreneurs. The baseline data was collected and verified for this set of firms (n=550),and the participants were randomly assigned into one of three experimental groups:

1. Group 1 (treatment): 250 entrepreneurs receive a smartphone with the “MARKETmanager” tool, a mobile app tool that is focused on tracking key business metrics related to entrepreneurs’ management of marketing activities (e.g. products and customers, and related policies/practices). The tool prompts entrepreneurs daily or weekly to complete a series of simple and straightforward questions about their business. Entrepreneurs are visited weekly by a data analyst who can assist with usage of the tool. On a monthly basis, entrepreneurs are also provided with ‘marketing analytics’ reports specific their firm, that compares current to past performance. Overall, this solution aims to increase an entrepreneur’s ability to pay attention to, and act on, key marketing and sales information.

2. Group 2 (pure control): 250 entrepreneurs form a counter-factual group that does not receive any intervention during the project (note: they will be given access after the project ends).

3. Group 3 (placebo): 50 entrepreneurs form a second comparison group that receives only a smart phone and data plan (they will not be given any marketing analytics apps, tools, intervention, etc.). This is to rule out potential alternative explanations in our study results due to exposure to smartphone or the internet.

Randomization done in office using stata

The randomization unit is a firm.

No

550 firms, 3 experimental groups - Treatment, Control and Placebo
The design is not clustered

550 firms, 3 experimental groups - Treatment, Control and Placebo (The design is not clustered)

250 firms in Treatment group, 250 in Control group and 50 in Placebo group