Have you seen the documentary The Social Dilemma on Netflix? I strongly recommend watching it. It’s about companies armed with unparalleled weapons. It’s about machine learning algorithms that are six times more likely to hit their target in terms of presenting the right offer to the right customer. Which means they can sell more. Why leave such a growing competitive advantage to Facebook and Google? This superpower is more accessible than you might think.
By entrusting machines with tasks that humans find hard to achieve, or which are so repetitive and alienating that no person could achieve them, companies will derive several advantages.
This could equally be applied to the progression rate through a company’s marketing funnels.
By determining the right price, more sales can be made at the regular price to customers who are prepared to pay it, and offers or rebates can be presented only to customers who require them in order to take action (they are more price sensitive). Data and automation enable the creation of dynamic promotional rules that would be impossible for humans to deploy. This contributes to increasing the average value of transactions and the number of transactions.
…or unforeseeable and even counterintuitive opportunities. Integrating the data stream enables the discovery of similarities or common characteristics amongst customer groups that have a considerably larger likelihood of becoming customers in the future. This particularly enables you to determine the most promising markets or segments, or even specific product attributes to highlight for a cluster of customers.
Focusing your employees on creative, interesting tasks they find fulfilling will have real effects on their happiness at work and your ability to retain them, in addition to creating more value for the company.
To benefit from the advantages mentioned above, which when taken together represent an undeniable competitive advantage, you need to start somewhere.
We often talk about the importance of having a data-driven culture. While it’s certainly easier to say than to effect, it’s essential to make it happen. A data-driven culture is based on a few pillars, such as:
If you would like to learn more about data-driven culture, this article from Tableau discusses it in different terms, but in a similar spirit.
It’s important to start with data that already exist at your organization, such as posted transactions, existing contacts, or even data on the frequentation of existing sites. Too many projects begin with ambitious plans to collect new data and are then abandoned before even having created a small amount of value due to cost. It’s a common mistake. Beginning to create value using existing data enables you to better determine what new data you need to collect and organize in the future. This both limits costs and evolves them in relation to the value created (your CFO and accountants will love that), thereby reducing risks related to massive data collection (your lawyers will love that). Remember: Prototyping is your ally.
Patterns exist in company data that could accelerate the discovery of opportunities. My advice is to trust the conclusions that emerge from the data. Let me explain. If an algorithm recommends targeting customers who like skiing because they are 68% more likely to convert than another customer, you need to test that—even if it might seem like something that’s totally coming out of left field. If the data points reveal this tendency and that an interesting amount of volume can be achieved, you should go with it. That may not feel intuitive for a human, or explainable, but if this cluster of customers has a higher propensity to be converted, you’ll be achieving more with your advertising budget. You’ll present the right attribute to the right customer segment and sell more, even if the data points may be hard to rationalize to your CEO.
My last piece of advice is to create a space where the data can be explored in complete security, without compromising customer data. That’s important for customers, but also for the company, its reputation (you might recall a certain financial institution in this regard), and its capacity for innovation and longevity! These data can and should be depersonalized for this purpose—the value of the insights and automation that will emerge from them will not be in any way reduced.
Finally, to ignore automation, algorithms, and artificial intelligence is to forgo superpowers that others will end up employing themselves. They allow you to spend more wisely, and therefore less, on advertising, and above all allow you to think about and develop your product and the argumentation surrounding it based on data you probably already have.