 "Analytics is not just about giving insights: Sudeshna Datta & Suhale Kapoor")
Companies can use decision engineering to help CXOs implement key insights, Sudeshna Datta and Suhale Kapoor tell Ankita Rai
Considering the significance of pricing strategy and impact on brand equity, it is imperative for brands to adopt intelligent analytics when they put a price tag on their products. How can analytics help companies make the right pricing decisions?
Datta: A consumer's sentiment towards a brand and, hence, the purchase decision depends on her perceptions about a brand's identity, personality, and heritage. A brand's price is integral to the brand's equity. Conjoint analysis is used to measure preferences for product features, to learn how changes to price affect demand for products or service, and to forecast the likely acceptance of a product if brought to market. It helps marketers determine what features a new product should have and how it should be priced.
Datta: We even use conjoint analytics in our recruitment process. For example, if you ask a candidate whether she wants to work for a bigger brand, or desire more salary, the answer would be 'yes'. But these are not so straight-forward choices. No one would say they want lower salary. But if you break each attribute in terms of flexibity, proximity to the workplace, bigger brand, training etc each choice becomes a trade-off. For instance, 'high salary, small brand name and closer to home' can be one choice. Thus, conjoint analytics can also be used to hire the right candidate and how much the weightage she gives to salary, training, brand name, flexibility etc.
Look back on Flipkart's Big Billion sale fiasco. How can analytics help e-retailers plan these mega events?
Kapoor: E-commerce players can use analytics at two levels when going for a sale event. First is planning, which includes planning logistics, inventory, what SKUs to carry, duration of the sale etc. This is more strategic in nature. Here companies can use past trends and optimise them and decide what actions to take. The other part is more tactical in nature. Companies can use predictive algorithms to suggest related products, next best products etc. In fact, 30 to 40 per cent of Amazon sales come for such relevant suggestions.
So whether one is launching an end-of-season sale or discounts and schemes during festival, planning is very important.
Remember, however, that in analytics you cannot predict anything unless you have a history. It is all about data. Unless you have the data from the past, you cannot predict the future. This is very different from market research. It is more about survey data where you ask a number of consumers how likely they are to buy a product and predict sales. But analytics requires old data. If you don't have data about previous sale of a large magnitude, you cannot predict how much inventory to keep and how much traffic is expected and how many servers to keep .
Datta: Analytics can also help retailers plan the whole consumer chain. First is segmenting consumers into deal-seeking consumers, convenience seekers, price buyers etc. Second is customer acquisition through marketing and sales and once you have acquired the customer how to cross-sell and up-sell. Third is customer retention. Analytics can be used here to make customised loyalty programmes, offers and emailers to keep the customer engaged.
Freemium - a combination of 'free' to 'premium' - has become the dominant business model among start-ups and smart phone app developers. How can companies leverage analytics to drive revenues in freemium models?
Kapoor: Most of the subscription companies, such as Linkedin or Gaana.com, work on freemium (free to premium) models. If you pay, you get some extra features. These companies make money only from the paid subscribers and advertising. So analytics is very critical for them. The challenge for these companies is how to promote user acceptance for paid services. In such a case we use free-trial marketing strategy for consumers, in which a service is provided free of charge either with limited functionalities or with full functions for a limited time. Through the analysis of people who shift from free to paid, we predict who are the consumers who would upgrade and how they are different from others.
For instance, if there are 100 million subscribers out of which 10 million are paid, we analyse all the members and find out how the paid members are different from others. We predict out of the 90 million, who are most likely to start paying and give them relevant offers. Instead of sending an offer, say one month free membership to all the 90 million subscribers, we will give it only to people who are likely to upgrade. So the marketing budget will come down and conversions will go up.
Has data analytics itself changed in some way over the last few years - given the changes in markets and among consumers? What are the new trends one should watch out for?
Kapoor: Till now, most of the analytics companies were used to giving insights and developing models. Then it was up to the client to figure out what to do with these insights. Now the trend is towards decision-engineering. It is about helping CXOs in decision-making.
Datta: Instead of traditional analytics, where the analytics is done on a given data set, decision engineering involves working with key decision-makers, such as CMO, CTOs, CXOs to understand what are their pain-points and then figure out what are the various data levers we can apply to help them take the decisions. It is not a canned approach. So for the same problem you would need a combination of data such as market research, big data or may be large data sitting with a company's customer base. Decision engineering brings all the data in the company together to enable decision-making.
MEET THE ANALYSTS
Considering the significance of pricing strategy and impact on brand equity, it is imperative for brands to adopt intelligent analytics when they put a price tag on their products. How can analytics help companies make the right pricing decisions?
Datta: A consumer's sentiment towards a brand and, hence, the purchase decision depends on her perceptions about a brand's identity, personality, and heritage. A brand's price is integral to the brand's equity. Conjoint analysis is used to measure preferences for product features, to learn how changes to price affect demand for products or service, and to forecast the likely acceptance of a product if brought to market. It helps marketers determine what features a new product should have and how it should be priced.
More From This Section
Kapoor: Conjoint analytics is used for finding the optimal price-pack architecture. It enables you to create a market advantage through redesign of your product portfolio. Consider this example: Faced with rising input costs, a leading chocolate manufacturer in India had two options, either to raise the prices of the chocolates or introduce smaller pack sizes. Now market research cannot be used in this scenario because it is not easy to understand how much extra a consumer will pay for a different-size pack of the same brand. These are not rational decisions. Customers take several attributes into account when making choices - such as brand, colour, price quantity etc. It is about trade-offs. In conjoint, you put all the attributes together and let the consumer trade off and make the choice.
Datta: We even use conjoint analytics in our recruitment process. For example, if you ask a candidate whether she wants to work for a bigger brand, or desire more salary, the answer would be 'yes'. But these are not so straight-forward choices. No one would say they want lower salary. But if you break each attribute in terms of flexibity, proximity to the workplace, bigger brand, training etc each choice becomes a trade-off. For instance, 'high salary, small brand name and closer to home' can be one choice. Thus, conjoint analytics can also be used to hire the right candidate and how much the weightage she gives to salary, training, brand name, flexibility etc.
Look back on Flipkart's Big Billion sale fiasco. How can analytics help e-retailers plan these mega events?
Kapoor: E-commerce players can use analytics at two levels when going for a sale event. First is planning, which includes planning logistics, inventory, what SKUs to carry, duration of the sale etc. This is more strategic in nature. Here companies can use past trends and optimise them and decide what actions to take. The other part is more tactical in nature. Companies can use predictive algorithms to suggest related products, next best products etc. In fact, 30 to 40 per cent of Amazon sales come for such relevant suggestions.
So whether one is launching an end-of-season sale or discounts and schemes during festival, planning is very important.
Remember, however, that in analytics you cannot predict anything unless you have a history. It is all about data. Unless you have the data from the past, you cannot predict the future. This is very different from market research. It is more about survey data where you ask a number of consumers how likely they are to buy a product and predict sales. But analytics requires old data. If you don't have data about previous sale of a large magnitude, you cannot predict how much inventory to keep and how much traffic is expected and how many servers to keep .
Datta: Analytics can also help retailers plan the whole consumer chain. First is segmenting consumers into deal-seeking consumers, convenience seekers, price buyers etc. Second is customer acquisition through marketing and sales and once you have acquired the customer how to cross-sell and up-sell. Third is customer retention. Analytics can be used here to make customised loyalty programmes, offers and emailers to keep the customer engaged.
Freemium - a combination of 'free' to 'premium' - has become the dominant business model among start-ups and smart phone app developers. How can companies leverage analytics to drive revenues in freemium models?
Kapoor: Most of the subscription companies, such as Linkedin or Gaana.com, work on freemium (free to premium) models. If you pay, you get some extra features. These companies make money only from the paid subscribers and advertising. So analytics is very critical for them. The challenge for these companies is how to promote user acceptance for paid services. In such a case we use free-trial marketing strategy for consumers, in which a service is provided free of charge either with limited functionalities or with full functions for a limited time. Through the analysis of people who shift from free to paid, we predict who are the consumers who would upgrade and how they are different from others.
For instance, if there are 100 million subscribers out of which 10 million are paid, we analyse all the members and find out how the paid members are different from others. We predict out of the 90 million, who are most likely to start paying and give them relevant offers. Instead of sending an offer, say one month free membership to all the 90 million subscribers, we will give it only to people who are likely to upgrade. So the marketing budget will come down and conversions will go up.
Has data analytics itself changed in some way over the last few years - given the changes in markets and among consumers? What are the new trends one should watch out for?
Kapoor: Till now, most of the analytics companies were used to giving insights and developing models. Then it was up to the client to figure out what to do with these insights. Now the trend is towards decision-engineering. It is about helping CXOs in decision-making.
Datta: Instead of traditional analytics, where the analytics is done on a given data set, decision engineering involves working with key decision-makers, such as CMO, CTOs, CXOs to understand what are their pain-points and then figure out what are the various data levers we can apply to help them take the decisions. It is not a canned approach. So for the same problem you would need a combination of data such as market research, big data or may be large data sitting with a company's customer base. Decision engineering brings all the data in the company together to enable decision-making.
MEET THE ANALYSTS
- Sudeshna Datta leads the corporate development, human resources and marketing functions at AbsolutData
- Prior to co-founding Absolutdata, she was a co-founder of globedecor.com, a home decor company. In the past, she has held various positions at Pfizer, New York, and later Kraft Foods Chicago
- She has an MBA in marketing from Cornell University
- Suhale Kapoor has been involved in building the India operations of the company from the start-up stage and currently looks at finance, technology and business development
- Prior to co-founding AbsolutData, he was with Mitsubishi Corp
- Kapoor holds an MBA from IIM Ahmedabad