In this post, we will help you better understand the potential of the technology and take a look at the power of predictive analytics in marketing and why marketers should be taking note!
“Predictive analytics is a form of data mining that uses machine learning and statistical modeling to predict the future. Based on historical data.” – David Arnoux
Applications of predictive analytics in action today are all around us. For example, banks have been using predictive modelling to approve or decline your credit cards and personal loans.
Has your bank ever notified you of suspicious activity on your bank card?
Well, it’s likely that a statistical model was used to predict your future behaviour based on your past transactions. Serious deviations from this pattern are flagged as suspicious. And that’s when you get the notification.
Predictive Analytics is also used for weather forecasting, recommendation engines, spam filtering and fraud detection, and that just scratches the surface!
But why should marketers care?
Trends in Predictive Analytics
Let’s start by demonstrating the level of interest in the field.
We can see from Google Trends that search volume for the topic has been rising over the last 5 years. So, if you haven’t already, you should probably be taking note!
Predicting & Optimising Conversions
Predictive analytics in marketing can be used to optimise conversions throughout the funnel and to identify the best way to move those leads down the funnel.
If you can predict your customers’ behaviour along the funnel, you can also think of messages to best influence that behaviour and reach your customer’s highest potential value.
This is super-intelligence for marketers!
Imagine if you could not only determine whether a lead is a good fit for your product but also which leads are the most promising, the most likely to purchase in the first place. This would allow you to focus your team’s efforts on leads with the highest potential ROI.
Predictive Analytics in Marketing allows you to do just that!
This insight can then be used to identify which leads to focus more time on.
“It’s not about knowing how many leads you can attract… It’s about knowing how many good leads you can engage!”David Arnoux
Predictive Analytics in Marketing
Organisations are already feeding predictive models 1000’s of data points on each customer, gained from their online activity. This makes it possible to send personalised product recommendations.
A financial service provider can use thousands of data points created by your online behaviour to decide which credit card to offer you, and when.
A fashion retailer can use your data to decide which shoes to recommend as your next purchase, based on the jacket you just bought.
Let me demonstrate…
Netflix – Predicting Demand
As an online platform, Netflix obviously collect a lot of data, but it’s what they do with the data that matters.
From the insights they gain from the data they collect, Netflix is able to (a high degree of certainty) know what content it’s customers will want next.
You may not know, but Netflix analyses all the descriptions you read, how long you read them for, how much time you spend browsing titles before deciding on what to watch and even how long you watch it for.
When deciding which types of new content to invest in, this data is invaluable.
For example, the show House of Cards was born from predictive analytics. Netflix could see the audience had a strong interest in work by director David Fincher, the actor Kevin Spacey (at the time) and the original British series.
In statistical terms, it was a strong correlation.
Netflix can start producing content before the audience even knows they want it! They can anticipate the market demand, before the market demands it!
Amazon – Dynamic Pricing
Amazon are the masters of Dynamic pricing, a strategy that adjusts prices based on demand, historic buying behaviours, competitor prices and market trends.
Amazon adjusts prices on its website over 2.5million times per day!
As stated in this HBR article:
“Using a single price is economically inefficient because part of the demand curve that could be profitably served is priced out of the market”.
Dynamic pricing is now very common in e-commerce and has also been used by airline companies and flight comparison sites for years. Flight companies use factors such as date, time of day and historic online behaviour to display dynamic prices.
Tip: Next time you book a flight try opening the same deal in normal and incognito mode, you’ll likely see a different price!
Walmart – Supply Chain Optimisation
Walmart takes data instantaneously from its systems and incorporates it within its forecasts. They can then assess which products are likely to go out of stock and which have actually underperformed.
Combined with behavioural insights from its customers online activity, this provides a huge amount of data to help Walmart prepare for an increase, or decrease, in product demand.
Forecasting this, allows Walmart to personalize its online presence. They can target customers with specific products based on their predicted likelihood of making a purchase.
‘Target’ the Right Customers at the Right Time
Back in 2012, the US department store Target revealed a teenage girl’s pregnancy before her father even knew!
By analysing the historical purchase data of customers signed up to their loyalty program.
In short, the company used a predictive model to analyse what new moms where buying and looked back to see what they bought in the past. By using this data to train the algorithm it was eventually able to predict when a customer was likely to be pregnant.
As stated in the original NYT article:
“As Pole’s computers crawled through the data, he was able to identify about 25 products that, when analyzed together, allowed him to assign each shopper a “pregnancy prediction” score.”
Target then used this data to send promotional offers and coupons to women the algorithm predicted to be pregnant, a target group with a high spending potential.
The predictive model was correct!
While this story may be quite alarming, or funny, it does demonstrate the power of predictive analytics in marketing. It can be used to empower marketers with the ability to target customers with the right message, at the right time.
Predictive Analytics DOES NOT Require ‘Big Data’!
Ok, so I know what you’re now thinking.
These companies are huge, multinational goliaths, with massive data-sets! How can I possibly imitate?
Great questions, but you’ll be pleased to know that predictive analytics in marketing is now accessible to businesses of all sizes, and can be successfully leveraged with much smaller data-sets!
Let me share an example!
Harley Davidson (NYC) – Leads Generation up 3000%
The New York branch of Harley Davidson recently used the power of predictive analytics to increase lead generation by nearly 3000%!
Prior to implementing predictive analytics tools, the company would develop the traditional buyer ‘persona’ and lead with ‘gut feeling’ to identify the target audience.
By using predictive analytics tools Harley was able to identify new target segments they didn’t know existed, therefore, would have targetted.
Harley initially thought of their target audience in broad, buyer-persona based segments (gender, age, location) and projected only 2 percent of New York’s population to be potential buyers. The AI tools they used identified new audiences by correlating online behaviours with their likelihood of conversion and autonomously targeting them. – via GeoMarketing
Essentially, they used predictive analytics to find ‘lookalike audiences’, of which would be more likely to convert.
Remember, this was just for Harley Davidson New York, the data sets were not that big! Yet the results were huge!
Leveraging Predictive Analytics in Marketing
The evidence of the predictive analytics in marketing revolution is already around us. Every time we type a search query into Google, Facebook or Amazon we’re feeding data into the machine.
The machine thrives on data, growing ever more intelligent as it receives these feedback signals.
1. Ask the Right Questions
Simple, you need a sound hypothesis to test.
You need to know which questions you are trying to answer, you need to know which metrics you are trying to forecast and you need to know which future behaviour you are trying to predict!
2. Gather the Right Data
We’ve come a long way in terms of data availability. In Fact, it is said that 90% of the world’s data has been generated in the last two years!
But, we do still need complete (not large) data sets to arrive at plausible conclusions.
It’s important for you to figure out what data is available to you and whether it will be sufficient to answer your questions convincingly.
3. Use the Right Technology
Many leading analytics software companies have already launched predictive analytics tools, but they do vary in their methodologies.
To decide which solution is best for you, it’s crucial to have a team in place that has the right skill set to assess whether or not a particular software is for you.
4. Work With the Right People
Without the right people, it’s impossible to pose the right questions.
For all the talk of A.I. replacing people, it’s really only made clear that the focus should be on getting the right people!
Predictive analytics in marketing is now relatively mature. As we can see, companies like Target have been using it effectively since 2012. This maturity has also led to an array of great tools becoming available for organisations to use to build predictive models.
This accessibility puts predictive analytics into the ‘must have’ category of tools for marketers. If you’re not utilising it, you can bet the competition is!
Just remember, to leverage the technology effectively, you need to know what you are looking for, have the right data available, use the right tools and hire the right people.
If all four of these elements can be fulfilled, your business will be set up to become a predictive analytics powerhouse!
Learn A.I. for Marketing & Growth
Has this article peaked your curiosity? Interested in learning more about how A.I. can be applied to your business?