Press "Enter" to skip to content

Real-time analytics in E-commerce (Part-2)




In our earlier blog post, we introduced why e-commerce companies needed to have a continuous and live view of their business as opposed to the more prevalent method of having a historic view. In some cases, even a 5-minute delay in insights can mean an unhappy or lost customer.

In this blog post, we’ll try to look at a few compelling cases for the use of real-time analytics across the customer life cycle for a typical e-commerce setup.

The web/mobile driven nature of e-businesses means numerous opportunities for a brand to engage with a prospective customer. However, this also presents a few challenges:

  1. A company’s time with its customers is much lesser than what a traditional retailer gets
  2. Businesses have to get the interactions right EVERY single time at each stage
  3. One bad experience can take customers away to another competitor who is only a click away


It is for these challenges that an e-commerce company needs to have continuous visibility of their business – starting from its acquisition campaigns, customer buying process, and fulfillment – and be able to respond to situations in seconds and not hours. This not only helps in providing great customer experience leading to revenues but also optimizes the use of available resources.


The rest of this article will talk about potential opportunities that real-time analytics opens up for an e-commerce business at various stage of the customer lifecycle.


Customer Discovery

Customer-Discovery-IQLECTCampaign  Optimization:

Current Situation:

Digital marketing makes up for a very large chunk of the marketing budget at an e-commerce company. Many a time, companies are running multiple media campaigns each day (to target different groups or test effectiveness of each campaign). However, the performance metrics of these campaigns are only available towards the end of the day



With Real-time Analytics:

  • Track key metrics (# of clicks, page visits, sessions, cart size, etc) for each campaign
  • Track campaigns yielding the best results
  • Test with numerous campaigns to arrive at the most optimal mix


Business Outcomes:

  • Maximize traffic to website/app
  • Optimization of campaign spends


Customer Engagement


User Behaviour Analysis:

Current Situation:

There are numerous tools today that track user activities on a website/mobile app but most of these only provide aggregated information towards the end of the day. Only a handful of tools is able to churn out insights more rapidly (at intervals of 10 minutes) to enable a company to act on unfavorable incidents.



With Real-time Analytics:

  • Segment customers based on historic and current behavioral data to make customized offers, product recommendations when the customer is actively engaged
  • Identify the occurrence and reason for drop offs and immediately respond with an incentive for user to return (doing this after a few hours means giving the user sufficient time to buy elsewhere)

Business Outcomes:

  • Improve customer experience
  • Increase average cart size of customers


Promotions and Offers:

Current Situation:

Promotions and offers are the most widely used technique to increase sales as well as drive customers towards brand loyalty. The prevalent practice today is to extend a blanket offer to all customers without the knowledge what motivates a customer to buy. This leads to a situation where – i) a user is given an offer although s/he didn’t need one to complete the transaction OR ii) an offer is not sufficient to incentivize an ‘on-the-fence’ user.


With Real-time Analytics:

  • Make real-time customer-specific offers based on their price elasticity for various products. Sometimes, you may not have to extend an offer for a customer to complete a transaction
  • Block fraudulent usage of offers by detecting usage pattern in real-time

Business Outcomes:

  • Improve profit margins
  • Reduce losses from fraudulent transactions



Current Situation:

Like with traditional retail, the customer experience in e-commerce is tied to the overall infrastructure (e.g. Website or mobile app response times in the case of e-commerce). It is commonly observed that the traffic to a website drops drastically when the infrastructure performs poorly but this is only verified in hindsight.


With Real-time Analytics:

  • Decisions on infrastructure could be taken based on a combined view of visitor and infrastructure statistics
  • E.g. When higher drop offs at payment page occur and server utilization is over 90% over the past few minutes, one may need to add infrastructure to handle the high loads. The converse could also be true where you could afford to reduce infrastructure when the traffic is not too high.

Business Outcomes:

  • Improve customer experience
  • Reduce lost revenue opportunities
  • Optimize infra costs


Sales Conversion


Payments Analytics:

Current Situation:

Unlike a lot of aspects of e-commerce operations, online payments are one area which is very complex, critical and outside the control of the business. This complexity arises from:

  • Multiple payment channels (like Net banking, cards, wallets, etc)
  • Multiple stakeholders for a transaction (like customer, merchant, issuing bank, merchant bank, payment gateway, processor, etc)

While an e-commerce company may not be privy to some of the information across this value chain, they are also unable to utilize the available information to reduce payment failures, drive better bargains with banks, etc.


With Real-time Analytics:

  • Knowledge of transaction failures and frauds for each channel or geography during peak loads can help the company to resolve these issues immediately with the payment vendor (payment gateway, wallets, banks)
  • Knowledge of preferred payment modes of customers can help create mutually beneficial partnerships with payment vendors to improve sales

Business Outcomes:

  • Reduce transaction failures
  • Provide greater customer value

While a common argument we hear is that all of these are being done today using various tools, just probing deeper has revealed that most of the tools are not able to support “real-time-ness” that we think is needed. Like we mentioned earlier, getting these insights even 5 minutes late could mean a prospective customer has already made the purchase with a competitor.


In a subsequent article, we will look at the current technology setup at e-businesses today, the challenges/gaps with it and options to implement real-time analytics for their operations.

If you want to try out our Real-Time Full Stack Analytics Platform, please Sign Up here.

Leave a Reply