Hi marketers! You made it. You finally made it. 👋

From now on there’s no excuse not to use no-code machine learning to level up your marketing. 

No-code machine learning is beneficial for:

• Marketing teams that want to avoid going to the data science team for repetitive predictive models.
• Autonomous marketing analysts who want to make predictions and share with their bigger team to make better decisions.
• Marketing departments looking to get insight into a problem they’ve been experiencing. 
• Project managers looking to make their project AI-driven without needing a data science team or technical knowledge. 
  

As a marketer you probably know the future marketing trends everyone is talking about in 2020 such as being data-driven, using AI, personalization, and more. What’s missing in these discussions is how to actually implement AI into your workflow. 

This is the purpose of this blog post. This is meant to inspire what you can do with machine learning at every stage of the customer journey. 

But First, Why Collecting Marketing Data Is So Important

With customer data, you can begin to find patterns in attributes that relate to customer behavior. Having customer data is very powerful. If you have the right data, you can predict the probability a customer will perform actions such as cancelling their subscription, how much they will spend, their life-time value (LTV), and more. 

Machine Learning for Top of Funnel Marketing: Predicting LTV

Top of Funnel Marketing (TOFU) = Use machine learning to attract the right customers to your product.

Marketers should increase brand awareness to increase leads.

How to Use ML for TOFU Marketing

Once you’ve collected enough data on your customers (minimum of 1,000 rows x 5 columns), you can start making meaningful predictions with your data to inform marketing decisions.

For TOFU marketing, it’s useful to predict Lifetime Value (LTV) to gear your marketing campaigns around the customers who have the highest LTV. This will avoid wasted spend and focus on the serious customers. 

What Kind of Data Do You Need to Gather

In this age of digital marketing there are so many ways to collect and use data. The obvious ones are Typeform or POS systems like Square or Stripe. With these you can collect data easily and export them to a CSV file or a database. 

Here’s an example of what columns are good for predicting LTV.

Read more about collecting data here. 

How to Take Action From Your Predictions

When you predict LTV, you will instantly see the relationship between attributes. You can see the top drivers attributing to LTV in the Drivers bar. The dataset we used recorded marketing campaign metrics for an insurance company. You can find this dataset in our Data Store.

With the Number of Policies being the number one Driver, you can take a deeper look into what this means. 

Here you can see the customers with the highest LTV have 2 policies. At this point you can begin to shape your marketing campaigns to get your customer’s on two policies. 

Other attributes you can explore are state, income, gender, etc. to create personas for your campaigns.

Machine Learning for Middle of Funnel Marketing: Personas

Middle of Funnel Marketing (MOFU) = Use machine learning to predict customer actions and learn how to engage.

Transitioning to personas, this is one of the most underrated and looked over features of no-code machine learning. 

From the prediction we just made, you can click the “Personas” tab to save personas and predict customer behavior.

Here you predict the range of the customer’s LTV to see what attributes relate to the highest or lowest LTV. This will help you immensely in marketing campaign personalization. 

We have talked about churn a lot in previous posts as a use case, but creating personas that predict which customers will churn is also incredibly important to directing marketing campaigns. 

Read more about personas here.

Machine Learning for Bottom of Funnel Marketing: Cross Selling and Up Selling

Bottom of Funnel Marketing (BOFU) = Use machine learning to retain customers and predict probability of more purchases.

Retaining and developing existing customers is much more cost-effective than acquiring new ones, and marketers that effectively encourage purchases from existing customers get the most lifetime value from each customer.

Predicting cross-selling and up-selling is another amazing way to increase customer LTV. In SaaS predicting in-app purchases is a hot commodity to guide features and design as well. Using customer demographic and purchase data, Obviously AI can identify which customers will be interested in other products and which customers will be interested in upgrading existing products, allowing marketers to determine which touch points will likely result in the desired action.

What Kind of Data You Need to Gather

The same goes for churn and LTV, purchase and personal customer data is always useful for predicting cross selling and up selling. 

To predict cross sells, you’re predicting what related products the customer is most likely to buy. The data useful for predicting this is website data and purchase data. With this you can see the web pages most visited after a purchase and see how much each customer spent on a product. 

To predict up sells, more personal data is required. Because you’re predicting if a customer is going to buy an upgraded version of your product, the process is more complex and takes a more personalized approach. Taking examples from this post on upselling, “you could offer an upgraded subscription plan for a mobile phone to a woman who regularly exceeds her Internet data usage, or a better computer to a man who has already looked at a number of different gaming laptops in a store.”

Taking Action From Cross-Sell and Up-Sell Predictions

When you predict an in-app purchase decisio n or a customer buying another product, you can begin to recommend products to your customers. This can be done through email campaigns, content, or advertising. You can even build something more sophisticated like a recommendation engine used in applications like Uber Eats or Amazon. 

Here’s a great post on how large companies use machine learning to predict cross sells and up sells. 

Hopefully You Have a Better Idea How to Insert No-Code ML Into Your Marketing Funnel

You can read more about marketing with no-code machine learning here. We’ve also created several use cases for marketing, including:

• Growth Hacking
• Personalization
• Dynamic Pricing
  

If you wish to read more marketing-focused use cases, feel free to reach out!

Peace! ✌️