retail. Wooing the kirana store

Just like it’s true that India lives in its villages, it’s also true that much of middle-class India lives by its kirana store. While modern retail has made inroads into city life, there is tremendous opportunity in the large mom-and-pop-store market that envelops much of this country. They offer conveniences such as home delivery and monthly credit. There are an estimated 11 million kirana stores that sell grocery items in India.

However, building intelligence to supply kirana stores is a big challenge FMCG companies face. Their salesman has to visit about 40 stores a day, and can only spend 5-6 minutes to persuade the owners to stock more from the company. Big Retail uses a loyalty card or other technology that can capture customer data, but how would a company deal with issues of supply and demand to and from kirana stores? To complicate matters, India poses multiple challenges – seasonality, holidays shifting between months, peak shopping season differing from region to region.

To add more complexity, companies run multiple sales promotions to give dealers incentives to stock their brands. Another important challenge in FMCG retail is that products have to travel. Ensuring that products are available at the right location on time yet not overstocked, as that causes the inventory to age and adds to costs, is the crux. Like in any other sector today, artificial intelligence is being used in retail too to take the guesswork out of a manpower-intensive industry.

Information overload

Hemant Rupani, Director - Sales, Mondelez India, says AI and machine learning are crucial to the retail of tomorrow. The company is working on a pilot project that uses machine learning to integrate its internal data with big data on the markets, location, retailer’s past history, and other demographic details. Improving field resources’ productivity is crucial in an increasingly talent-constrained environment, and technology can step in to provide more knowledge.

Jayaraman Krishnamurthy, General Manager (Strategic Initiatives) of software solutions firm Capillary Technologies says his firm tries to leverage technology to make that small window of 5-6 minutes the salesman has effective. “India is typically a push market. There is not much loyalty. Availability is key. The salesman has to have all the knowledge at the tips of his fingers to be able to get his brand seen and sold in the stores.”

Gunjan Shah, Head - Sales, Britannia, says the amount of information salespersons have to deal with “has been a problem for ages”. Britannia in the last five years or so has been supplying salesmen with a device – which many have traded for an app now – that contains software that synthesises the historical data of that outlet. A store near a school would carry a product mix very different from one near an office. The software also suggests products that an outlet should sell. For instance, if its profile shows that a lighter kind of biscuit, say, Marie Gold, sells well, it would suggest that a similar biscuit like Nutrichoice would do well to be stocked there. It also prompts salesmen to tell the shopkeepers to stock more of the product if a promotion is on.

Capillary uses techniques such as collaborative filtering, personalisation, trends and early stage predictive new product introduction to give companies a more efficient list of stores. “We use machine learning to isolate and identify granular trends from complex sales signals, predictive models to solve multiple use cases. For instance, which retailers are most likely to buy a new product, recommendations for replenishment and trade promotions personalised to each retailer.”

Provides training too

Mondelez and Britannia are building technology in-house. It includes content to train the salespeople too. There is information about new products in the form of audio and video clips. There is a centralised team of experts that answers questions from salesmen about, say, whether a product is vegetarian or contains eggs, via a messaging platform. Orders can also be placed this way, with a chatbot functioning like a virtual salesman and transferring the order back to the stockpoint.

Britannia’s Shah says most companies are using technology to make the supply chain more efficient and better equip salesmen to deal with the pressure. “The difference lies in the level of adoption. Once we cross 60-70 per cent adoption, it makes an impact.”

Mondelez India’s Rupani says his field force and distributors have digital tools that provide all information for every retailer that they serve, including details of their location, past performance, channel – portfolio gaps, promotion performance and so on. Mondelez has a large cooler base of 3 lakh units, and uses technology to keep track of it using QR code scanning, as they expand distribution into smaller towns and rural areas.

Video technology is also pressed into use to check planogram compliance – to ensure that the brands’ display is being done as agreed. Many are the instances where brands, say, in the dairy sector, install coolers at stores, only to find that they are being used to store dosa batter and flowers! Heat maps are used as well, to learn how many customers are in the various areas of the store, whether they are single or in groups, their gender. Smart Store technology gives additional insights such as time spent in store and sections, trial and purchase intent and section/category preferences. All these help in customer identification and personal recommendation. “We are examining if some of these advanced techniques that have been traditionally used in modern retail can be applied to kirana stores,” says Krishnamurthy.

Britannia’s Shah says one-way communication to salesmen needs to be addressed. While brands are able to give them much information – feedback, suggestions and inputs, how the competition is faring, news about which promotion is doing well – collating it into qualitative information is a challenge the industry is grappling with.