Crowdsourcing disrupts retail

A mobile app is using the power of crowdsourcing to target problem areas in retail - and creating jobs at the same time.

Johannesburg, 18 Aug 2017
Read time 3min 40sec
Andrew Dawson, Commercial Director, Solutions in Hand.
Andrew Dawson, Commercial Director, Solutions in Hand.

Using the power of the Internet to source groups of people to perform certain tasks is becoming commonplace. Crowdsourcing uses the power of connected groups of private individuals to accomplish tasks that would normally take teams of employees a lot longer to complete. It leverages the power of the Internet and has access to a massive potential user base as a result.

As a phenomenon it's particularly suited to Africa, with its massive unemployment and geographically disparate populations. According to the GSMA mobile economy report, Sub-Saharan Africa accounts for nearly a tenth of the global mobile subscriber base and is expected to grow faster than every other region over the next five years.

Innovators are capitalising on this rapid growth in smartphone adoption to address a wide assortment of problems faced by businesses in all sectors and across the continent.

One example of this is the use of a mobile app to enable FMCG retail outlets situated in Africa to recruit people to conduct independent audits. Andrew Dawson, Commercial Director at Solutions in Hand, says: "Using a crowdsourcing app enables data collection by registered individuals who reside in a specific geographical area."

People become registered users by downloading the app and registering themselves on it, providing their geographical location. Users can then be geographically ring-fenced and recruited to submit surveys, photos, and other intelligence for retail outlets in their area. Remuneration is fixed per survey and is paid on a daily basis using vouchers and e-transactions. Data usage is reverse-billed to the customer.

This type of crowdsourcing app offers the unemployed an opportunity to generate a revenue stream in return for approved submissions. It also provides opportunities to conduct surveys in their spare time for those needing to generate additional income to make ends meet. A typical survey can generate anything upwards of R20, so could have a significant impact for someone doing between five and eight surveys a day.

The benefits to retail operations that implement a crowdsourcing approach to gathering information are varied, says Dawson. "Any successful business needs access to market intelligence that's current, accurate and measurable. The ability to crowdsource users who are spread out geographically means you can conduct countrywide, live, focused intelligence gathering as and when needed."

This means that the business gets data delivered to it in a measurable and accurate way and can make business decisions that are a direct reflection of the information at hand.

The feedback is submitted on forms that are designed to align with the retailer's key performance indicators (KPI). The data is collated and represented on a KPI dashboard for ease of reference.

How crowdsourcing an FMCG survey works:

* The user downloads the application and registers, providing a residential location.
* A radius is calculated - outlets within that radius will be assigned and rotated on a regular basis.
* The system allocates a call cycle of visits to match the data gathering requirements of the client.
* The user is notified of the list of available outlets the day prior and can decline or accept the route.
* The required questionnaires are downloaded onto the user's device.
* Once the user begins their cycle, there's a set time frame in which to conduct the allocated surveys.
* The user earns a fee per outlet visited and survey completed.

The types of information that can be captured like this include photos, text input, numeric input, barcode scanning and GPS tagging. Various measures can be implemented to ensure that the data collected is of the best quality possible. Dawson explains: "Users can be rated over time according to the quality of data they typically generate. Users with higher ratings can be allocated more work going forward, and users who return bad quality data or incomplete results can eventually be removed from the system, following the appropriate remedial measures."

See also