F U L O N

CRM+Email marketing for Amazon Sellers

Introduction

In this blog, We will give you practical ways to use Predictive Analytics using the E-commerce Data you have from your own websites, social media and even the data from third party Marketplaces like Amazon, Walmart etc.

You will see this is the future Game Changer and using Artificial Intelligence and Machine Learning will not only be simplified by absolute must. We are going to combine these power of Data Analytics with User Experiences and Journeys to identify the and eventually solve the pain points of the customers to find the deep hidden value in ways.

We will cover

  1. The phases and paths necessary to start gathering data and generating predictability
  2. How to gather and extract relevant data
  3. Identifying the user behaviours using the data insights. Basically, know as much about your customers like what they really want, why they buy and when etc.
  4. Identify the critical paths which is actually impacting your sales.

Here I will be exploring with you the phases and stages to generate predictive analytics that can deliver useful insights into the customer behaviours.

Predicting Analysis using the data will allow you to answer core responses, on your Customers, Product and Services.

NoteIt is not a fool proof magical system which will 100% accurately predict the future but can give you the likely and unlikely of some events and predictions

What is Predictive Analytics

It all starts with finding the main problem you are trying to solve

Here is some drill down questions for E-commerce brands –

  1. Need More Sales
  2. Reduce the costs
  3. More New Customers
  4. More repeat buyers
  5. Keeping only those products which will sell more etc

There are three Core Process –

    1. Identifying the Information
    2. Analyse the Information using the models to validate the accuracy
    3. Implement the results based on the insights predicted.

1. Identifying the information:

  • Define the E-commerce business problem which is required to be solved.
  • What data is required to answers those questions/problems.
  • Look at the quality of Data. Are we capturing sufficient enough data?

2. Analyse the Information and Selecting the Model:

  • Choose the right data modelling algorithm. Start with small sample of your E-commerce data and validate the results and model.
  • Making sense of the results.

3. Visualise and Implement:

Results from previous exercise then will be used to build the deep insights and actions for the business team to identify the process and areas of growth.

As I said earlier, it has to start with finding the right questions. This requires you to be –

  1. Understanding your Brand
  2. Have a curiosity and open to all sort of questions
  3. And good skills in understanding the process

E-commerce Applications like Fulon allows giving the deep customer insights and actions using the actual sales data from Marketplaces like Amazon.

How to identify the real problems – Our Fictional Company NutriKaSupplements

Let’s look at the example from a fictional company in Health and Supplement business for Men and women say Brand is NutriKaSupplements

What is the business problem of NutriKaSupplements? Of course, generating more revenue

NutriKaSupplements like other Brands is doing okay handcrafted but Brand owners always have in mind that they could do better. Now, unlike a lot of businesses, Brand owners are smart enough to realize that they need to take a look at themselves at what they’re doing, how their customers are behaving before they start implementing the changes make a radical change. They know they’re

products are great but their sales seem to be tied to certain times like New Year when many customers think of health. But they need to unlock a way to get more consistent revenue for the rest of the year.

To increase sales, we will need to dig further down. Profits are related to costs and expenses plus sales and any other legal liabilities etc. E-commerce is a margin business so lower your spending will increase your revenue as well.

What are the core problems:

  • NutriKaSupplements doesn’t have enough buyers. More buyers mean more sales.
  • They are not getting enough revenue out of our existing customers. Few repeat buyers.
  • They are not efficient about what they manufacture, which means launching products with less demand.

They are not able to spend on sales and marketing and ad spending on the actual sales that are happening

Note - You can adapt the same process for your E-commerce Brand to find the legit problems. Plus, you can further drill down to the core of each and identify the areas to focus on. You need to combine intuition with observation. You may use all the latest data tools and mining techniques. But still, the best way of finding stuff out is the oldest one of all; that is, talking to customers, listening to what they say, developing empathy for what these customers are dealing with.

How to find the right Data for Predictive Analytics

We’re going to explore how to find the relevant data sources for predictive analytics for your E-commerce Brand (in this case NutriKaSupplements). So how to identify the data that we need? Simple, ask questions, which then leads to a question – what kind of questions to ask?

Let’s look at NutriKaSupplements (our fictional company) and for a small E-commerce Brand, we will be limiting ourselves to data that directly relates to the buyer’s journey.

Core Questions:

There is core five W’s + 1: Who, What, When, Where, Why, and How

Segregating your questions into different groups will help you start to identify what’s important.

For example, the main question when it comes to buyers is Who are your Buyers.

Check out more details on 5W+1H:

https://marketingland.com/5w1h-marketing-analytics-5w1h-framework-86043

https://www.smartinsights.com/managing-digital-marketing/planning-budgeting/5ws-h-business-planning/

Who:

In the Who group, questions are mainly demographic information like:

  • age,
  • gender,
  • education,
  • marital status,
  • income level,
  • job title, hobbies,
  • interests,
  • who they rely on for information?

 

And if your E-commerce Brand is for B to B (Business to Business), you’d probably tend to focus a lot more in business data and information points for example about their jobs, etc

 

What:

In what group, we find questions like,

  • What products buyers are buying?
  • What would buyers want to buy that our Brand cannot fulfil? (This will help to find new product fit in your catalogue)
  • What other products did they buy?
  • What is their most preferred way of buying?
  • What don’t they like (Lookout for Negative reviews on Products and Brand), that turned them off to not buy anything?

 

When

Here we have time-specific questions.

  • When are they buying or not buying?
  • When did they spend the most time on our marketing messages and ad campaigns?
  • When they preferred to be contacted and when they do not want us to contact them?

 

Where:

This are all location-oriented questions.

  • Where are they when they’re buying from our online store?
  • Where they buy from, is it our online store, or Amazon, Walmart, etc
  • Where they make the most purchase from (Home, Office, Mobile device, etc)
  • Where they reside? In certain geographic locations, states, or close to some business centres?
  • Where are they buying from when they’re not buying from us?
  • Where their products are getting delivered?

Why:

This is the most important aspect where we can really find the behaviours of our buyer’s actions and motivations.

  • So why are people buying, say in case of NutriKaSupplements – why they are buying supplements
  • Why are they buying the amount that they’re buying?
  • Why don’t more people buy from us
  • Why they only respond to some of the communications and why rest of it fails?
  • Why does our stock sell well sometimes while it won’t sell and in the storage?

How:

    • Here we are thinking more from the process side
    • How are they paying us?
    • How do they like hearing from us?
    • How much really buying at a time?
    • How much are they paying?

To figure out questions related to your own e-commerce brand, you can check below websites –

http://customerjourneymarketer.com/

https://customerthink.com/

https://www.xminstitute.com/

It is really critical as Brand Owner to deeply understand and get the answers to the questions because without it, your actions will be based on wrong hunches and not validated by the real data

What Data Sources and information you need?

Once we have identified the questions, we need answers for, we need to identify the places where we can get the relevant data.

This exercise of identifying the data sources can be bit overwhelmed. Too much data is as bad as too little data. We need the right inputs.

We will be focussing here for a small to medium E-commerce brands, like our NutriKaSupplements health supplements

Let’s look at few data sources –

    1. Email Analytics – Emails may seem like a outdated but still it is one of the best medium to grow sales. You will look at open rate, click through rate, time to purchase. With tools like Fulon and Curvv (you can drill down further to each buyer level on when they buy, how much they buy from Marketplaces like Amazon). However, not all your buyers will be in your email list so still will be missing out your potential buyers who are in different medium.
    2. Pinterest Analytics – It is especially important for sectors like food, health and fashion, and if you have marketing campaigns there, as our Supplement store does. You can bring in other kinds of data, such as what kinds of pin boards are getting attractions? What’s bringing them to our store? What kind of ad are they responding to? What other things are customers interested in. Are they just health freaks, or is there something else that’s driving them?
    3. Facebook/Instagram Analytics – This is where you get insight into your customers, personal lives, their desires, likes and dislikes and fears. You will get demographic information, behavioural information and relationship status. It might be relevant when your campaigns are based on certain life events of your buyers like Marriages etc
    4. Google Analytics – which is almost an industry standard on websites. Email and other social medias are just marketing channels, that drives customer to your store or marketplace store. Now there is a big brother Google Analytics. We can get all the data on how they’re visiting our store. How many visitors came to the store? How many of those then visited the products and how may read the reviews and buy from us? plus, much more you can find all these things out in Google analytics. We can also find demographic data from Google Analytics. You can build a segment that focuses in on how many people who visited the testimonials page went on to buy, or a segment that looks at who gave the bulk order versus just a single purchase.
    5. Order analytics – One of the most important from various sources like your own stores and marketplaces like Amazon, Walmart, E-bay etc. Eventually these are your real buyers who are converted and can tell you all kinds of stores based on their purchase behaviours. With Analytics of Fulon, you will be able to deep dive into the customer minds. Actual buyers will give you the outputs on the customers which will then fee to the inputs to your marketing campaigns, products and new launches.

Now that we’ve identified the information sources, next step is how to start gathering the data, refine it, remove the noises and start making sense of the analytics.

Summary – So far we have discussed about what are the main goals, what questions for which we need answer for and then applying the data aggregation and models to predict.

There are lot of various Models and Tools and Programming libraries which can be used to train the model using the E-commerce data. However idea of this blog is to give an idea from the business perspective.

Related Post

Leave a Comment

Say hello and welcome to the new data-driven and customer-focused Amazon FBA business

Access

Follow Us

Email: info@fulon.co