In various roles in my career, I’ve had to ask myself: “What drives revenue growth for my business?” I think this is explicitly a math question, but I don’t often use math to answer it…
My first instinct is to take the path of least resistance and find answers using qualitative anecdotes. I think it’s human nature to rely on recent events to help explain things that are otherwise hard to explain. There’s even a term for it – recency bias.
But when I’m feeling more ambitious, I pull out some basic mental math. I usually identify a correlation between two easy-to-access business measures.
My brain is decently suited for these types of analyses. Though I admit, I usually don’t come to any meaningful conclusions. I mean let’s be honest, complex problems usually require more mathematically sophisticated solutions.
With the advent of three major technology shifts, I think this is all about to change. How? In what ways? Read on!
Digital Transformation is creating Big Data
According to IDG, 91% of organizations have adopted or have plans to adopt a digital-first business strategy. By transitioning from offline to online business solutions, companies will produce incredible amounts of data.
Web apps, mobile apps, online transactions, online communications, and internal offerings like digitized workflows. All of it producing data. More data with business signals than a company has ever had available before. And likely more than most companies know what to do with.
Think about what happens when a repair technician’s once paper-based checklist gets converted into a mobile app. Certainly data is captured about which tasks have been completed. But maybe more importantly, meta data is created. Like,
- when each task was completed,
- where they were completed, and
- in which order each task was completed.
Now imagine this fine-grained data getting created for all your internal and external business processes. Though it might seem a little daunting, the opportunity it creates for the business are endless.
The Modern Data Stack makes data accessible
Truly, the scope and scale of data captured by most businesses today provides great promise.
We can now answer business questions with precise evidence, not conjecture. But to realize this potential, we need to wrangle data to meet our analytics needs.
A whole industry has popped up to help us better manage our data. We can now easily take care of tasks like integrating, transforming, and pre-processing data (ETL). We can also persistently store and secure data for quick analytics (The Data Warehouse).
These tasks roughly make up a space known as The Modern Data Stack. The modern data stack’s primary goal is to get data standardized, quickly accessible, and easy to analyze. In other words, it makes data accessible to us.
Ok, so Digital Transformation ensures we have lots of data. And the Modern Data Stack makes data accessible. What’s next?
Machine Learning makes data useful
Thinking back to my freshman Stats 101 class, the concept that has stuck with me by far the most is correlation. It has helped me answer questions like:
- Does a grad school degree really lead to a higher income?
- Is my diet directly related to my blood pressure?
- What is my chance of getting COVID if I wear my mask all the time?
I've applied the concept of correlation to problems at work as well:
- If I increase the speed of our production line, what impact does that have on quality?
- Can investing more in influencer marketing generate more sales leads?
- Do more frequent customer success reviews lead to increased customer adoption?
These correlation examples are easily interpretable by our human brains. Because, well, they are simple.
They ask us to examine the interdependence between a single input and output. We know, however, that a real-world outcome is rarely explained by a single input.
For example, when trying to predict my lead volume, I know that influencer marketing must be a factor. But it’s safe to assume it’s not the only factor. Brand awareness, market conditions, consumer trends, and hundreds of other variables play a role as well.
The interplay between multiple variables is called multi-variate correlation. Decidedly, our human brains are just not well suited for this type of analysis.
We have a hard enough time understanding curves plotted on 2-dimensional planes. Imagine thinking about the interdependence of 10 dimensions!
Machine learning presents the perfect solution for the problem. We finally have the tools to run a holistic analysis on volumes of data across hundreds of variables. This means we can finally understand the drivers of revenue.
Usually, the relationship between business signals and resulting revenue isn’t even linear. But a machine learning model can identify impactful correlations.
For instance, machine learning can identify that 3 report downloads per month with 10 or more web logins per week leads to higher revenue spend. However, more than 6 report downloads per month leads to a decrease in revenue.
In an increasingly product-led growth world, it is expected that product usage patterns alone will indicate a customer’s propensity to buy. Yet a customer’s journey doesn’t just start or end with product use.
So, a robust analysis should include product use along with business signals from marketing, sales, customer success, and customer support. This type of complexity demands a sophisticated analysis that only machine learning can provide.
Building truly useful machine learning models is hard. It requires the right set of inputs. And the models should be customized to fit your unique business if you expect insights specific to your business.
It's time to accelerate business growth!
All that said, I still think the future is bright for unlocking what factors truly drive business revenue. And thanks to Digital Transformation, the Modern Data Stack, and Machine Learning, solving the puzzle won’t be nearly as hard as it has been historically.
Call me a nerd but using complex math to solve real business problems gets me jazzed up! In fact, it’s why we started Tingono.
I’m excited to build a solution that is making it easier to determine which activities are actually responsible for revenue retention and expansion. If this interests you, reach out to learn more about what we’re up to.
