Moving Forward with Predictive Analytics
Wow, you made it; you’ve surpassed 90% of the companies out there. You’re up on stage and receiving rewards for innovative campaigns based on exceptional insights; that’s if your company bought enough tables or seats. No correlation there.
Stage 5 is a game changer. That’s when we stop looking back so much and start making predictions. Based on what we know about previous results, we can predict all types of things. How much traffic should we be getting? What the conversion rate should be? And more importantly, based on what we know about our customers, segments, or audiences – who is most likely to convert? – so that before they do so, we can adjust our bidding accordingly: bidding higher for audiences with a higher likelihood to convert, and passing on those that look like tyre-kickers.
Any tool or innovation can be used for good or bad, and it’s up to us to provide that principled lens to ensure we follow ethical practices, or we risk having our new toys taken away. When explaining level 4 or 5 of the Analytics Maturity Curve, it’s best to make clear that we’re only ever looking to understand a 330-degree view of our customers and only that data they permit us to use.
Marketers and brands must continue to move away from the perception that data is creepy by not being creepy and using data with permission for good. As data-fuelled marketers, we can do more than sell things. We can improve things for our customers and the world.
According to IBM, predictive analytics ‘is a branch of advanced analytics that predicts future outcomes using historical data combined with statistical modelling, data mining techniques, and machine learning. Companies employ predictive analytics to find patterns in this data to identify risks and opportunities. Predictive analytics is often associated with big data and data science.’
So, let’s look back a little; whatever happened to big data? According to Hyoun Park, the CEO and Founder of Amalgam Insight, it died: ‘The Era of Big Data passed away on June 5, 2019. The Era of Big Data is coming to an end as the focus shifts from how we collect data to processing that data in real-time. Big Data is now a business asset supporting the next eras of multi-cloud support, machine learning, and real-time analytics.
‘Ultimately, it was a stepping stone to where we are now. It wasn’t the answer, but it helped us ask the right questions and start thinking seriously about the data we were or could or should be gathering.
‘Big Data will be remembered for its role in enabling the beginning of social media dominance, its role in fundamentally changing the mindset of enterprises in working with multiple orders of magnitude increases in data volume, and in clarifying the value of analytic data, data quality, and data governance for the ongoing valuation of data as an enterprise asset.’
So big data was an important enabler, but we quickly realised the value actually lay in our ability to manage and use that data effectively at scale. Its value is in the predictions we can make and our ability to trust the data and insights that materialise.
As we get to this stage, the learning curve ramps up again, and we’ll likely need to add more specialists to our team. It all starts to seem too difficult, but it doesn’t need to be, and we can start making gains bit by bit. Here is a basic model used by mint.com and market director Noah Kangan. It shows how you can create simple, effective predictions to inform your decisions.
Noah Kangan is a successful serial entrepreneur and the founder of OkDork. Before this, he was the market director for Mint.com, and he was given a seemingly unachievable target: to acquire 100,000 users in six months.
It seemed like a nearly impossible task, but rather than shrinking from the challenge, Noah leaned into it, thinking practically about where the best use of his time was – he wanted to work smart.
As with our Analytics Maturity Curve, once he had his basic data, he worked on defining the KPIs and targets that would allow him to reach his set goal – setting benchmarks and milestones for new users, sessions, and recurring revenue and breaking the goal into quarterly, monthly, and weekly targets.
Noah created a spreadsheet with this information as the third step, and it looked like this:
Source: Noah Kangan
As you can see, it’s composed of the channels he used and goals for metrics that were important for Mint.
Noah then made a simple scoring system based on ease of implementation and potential impact. These numbers were added to create a final score, giving him the ideal combination of strategies and channels.
Source: Noah Kangan
Using this simple method, he brought in 1,000,000 new users – 10 times more than the target of 100,000.
What makes this story particularly interesting is that it highlights what we mean by predictive analytics and how even the most straightforward methods can deliver or at least support the delivery of outstanding results.
In this algorithm, he uses just two constant metrics multiplied together. That’s it. Noah is making the prediction himself based on his own experience. No AI is needed. The point is that you can do this, and it doesn’t have to be difficult to get great results.
As we continue working through how and why our data is F**ked, we’ll look at what is possible right now with predictive analytics and some key personalisation strategies that are accessible with the right mix of data, maturity, and tools.
We’ll go on to look at what the future might bring, from Amazon predicting when we might want a product and stocking it before we purchase it to insurance companies using our data as consumers to assess our potential health risks and the price we pay.
By now, you have gained a comprehensive understanding of the Analytics Maturity Curve and the essential requirements for advancing at each level. You may be thinking about how missing some levels may have contributed to your own F**ked data. You will clearly see what other organisations have achieved at each level.
Before we go further, let’s take a step back and look at The Power of One: why we’re doing this, how not to sell bananas, and what we want to achieve beyond our data being unF**ked.
