It all starts with a high-level business goal laid out by the CEO.
Grow your customer base. Increase customer lifetime value. Get more customers to use your online shopping tool or make recurring orders. Reduce churn. Reduce cost per acquisition.
These are the challenges that get passed down to your CMO. And then the question becomes how? How do I get more people to shop online? How do I get more recurring orders? How do I increase customer lifetime value or decrease churn?
From there, the how becomes a series of tactics. Perhaps you offer a coupon or incentive for signing up with your new online shopping tool. Maybe you make it simpler to sign up. Maybe you put your budget into making your best customers aware that they can shop online.
Once you’ve decided on tactics, more questions follow. Who are you targeting with this tactic? What’s the best way to reach them? What kind of content are they most likely to respond to? Which channels do they engage with? How can you get their attention?
Once the CMO and marketing team have answered those questions, the process moves from strategy to execution. There are content calendars and technical setups, channel managers and marketing employees at every level.
And here’s the thing: Execution is where the best laid plans often break down.
Execution is the marketing bottleneck.
The truth is that channel tools often don’t have the data to fully execute on your strategy. And so when strategy gets passed down to be executed in said tools, it also gets diluted (at best) or discarded altogether (at worst).
Let’s say, for example, you’ve got a new online shopping system for your store. The CEO has asked the CMO to get more customers to use the new tool. The CMO has told marketing to target customers who have not already signed up.
So far, so good, right?
Except that once the plan gets down to your email marketing specialist or your social media team or your SEM pro, chances are the tools they’re using don’t have the data to identify which customers have and haven’t signed up.
Which means either your process grinds to a halt or your channel managers individually have to do their best to try and solve the puzzle. The chances are high that this will lead to inconsistent targeting across your marketing channels and a scattershot approach to targeting the customers you’re trying to reach.
Here’s a surprise: The problem isn’t data.
Looking at that breakdown, it’s tempting to say that the problem is data.
If your channel managers just understood customers better, all these issues would fix themselves! If you had more data, you’d know who to target! If the tech team didn’t take so darn long to get you the data you need, email marketing could target customers who haven’t yet signed up for your online shopping portal! And so could Facebook and Google ads and every other channel manager. Right?
Even if you have a basic CDP or a database that centralizes your customer data, if that tool isn’t infused with data science that interprets the data and makes it actionable, each of your channel managers is now expected to act as a data scientist, interpreting the data for themselves and making decisions based on those interpretations.
And those interpretations aren’t simple.
Because your business might have 10 fields just for subscriber status. It might have fields that are only populated for people in Germany, others only for mobile users, still others only for customers in your CRM. It might have thousands or hundreds of thousands or millions of events that need to be interpreted.
And when individual marketers at varying levels in their careers and varying skill levels with data interpretation are expected to sift through those fields and make quick decisions about what is and isn’t relevant…chances are, those interpretations will be very different.
Which means you’ve still got a fractured customer experience. Not to mention that every single channel manager now has a more difficult job.
Relevancy is the real problem here.
So, if the problem isn’t data, what is it? How do we understand customers better? How do we target them better? How do we personalize 1:1? If solving the data problem won’t solve the marketing problem, what do we do?
The easiest answer is to go back to our goals. Data itself was never the goal. The goal was reducing churn or increasing subscriptions. It was driving tangible, measurable business results. It was taking the high-level strategy developed at the CEO- and CMO-led stages of your process and executing on that strategy.
In other words, the end goal is relevancy—matching the right users with the right content at the right times to achieve our goals. Without relevancy, data just creates more problems.
Let’s go back to our online shopping example. The CEO and CMO want more online shoppers. You’ve decided to send out coupons. And now the question becomes: How do we target relevant audiences?
Because if you send a coupon for vegan hamburgers that cost $9 a pound to your not-vegan budget shopping audience, that’s going to strike them as expensive and irrelevant. But if you identify a shopper who actually shops vegan, that offer is suddenly cool and relevant and interesting. It’s suddenly something worth signing up for.
Inverting the campaign model
And so the solution here isn’t to simply acquire or unify more data. The solution is an inversion of the campaign model.
In the past, marketers would create campaigns based on demographic data and preconceived content. We’d assign target segments based on those demographics. And we’d cast a wide net.
Now, inverting that campaign model, we can use behavioral data to create the right content for our users and predict which users will respond to that content (and how). Where data used to be a general guideline, now it can actually help you personalize content to users at a 1:1 level–making sure vegans get vegan coupons and pizza-lovers get pizza coupons.
This is the reason giants like Amazon, Google, Netflix, and Spotify are knocking their marketing out of the park. They’ve moved away from the traditional campaign model and toward 1:1 personalization that’s automated—with data science, AI, and machine learning—to match the best content with each individual user.
But wait…don’t we need data to match content to users?
I’m glad you asked. The short answer is yes. Yes, we need data to match content to users. The point here is that data isn’t enough. Data isn’t the centerpiece of your goals. It isn’t what keeps marketers up at night.
What keeps us up are the results. It’s the matching. It’s the relevancy. It’s the way that data drives real business outcomes.
Where data has value is in its ability to improve our matching capabilities. Its value is in moving us past the old marketing model of matching content to users based on demographics to matching based on individual behaviors and interests.
Instead of advertising handbags to all female customers, now we can advertise handbags to people who consistently buy handbags. Instead of assuming women between the ages of 28 and 40 are moms, we can identify who reads articles about motherhood on our sites. Instead of targeting everyone between the age of 16 and 20 with college admissions ads, we can identify who’s researching colleges and who’s interested in our college’s specific programs.
Just having the data won’t get us there, though. We need correlation. We need a way to predict the likelihood to convert based on the data. We need, in other words, data science, AI, and machine learning.
Without machine learning, predicting behavior is a demanding, time-consuming, manual estimation process for the marketer. And the less data science background a marketer has, the more the results will feel like guesswork.
With machine learning, though, predictions become stronger. Machines can find correlations between the 75+ customer attributes Lytics uses to segment by behavior. Machines are faster and better at seeing these patterns and predicting the true likelihood of conversion for each individual customer.
Even better, machine learning doesn’t get tripped up by missing pieces of data. It correlates things based on what it has and doesn’t sweat what it doesn’t have. This means data sparsity and gaps don’t affect the machine’s ability to predict and segment your customers.
So, how do we fix the marketing workflow?
80 – 90% of marketing dollars are generally spent coordinating, deciding, reconfiguring, and aligning all the content across channels. But what if we could streamline the coordination? What if the decisions came down into the channels already made? What if your Facebook intern and your email team didn’t have to liaise three times on each campaign to make sure they weren’t overlapping efforts and irritating customers?
How much money and time could you save? How much more personalized would your experiences feel to customers? (After all, 86% of those same customers say personalization impacts their buying decisions.)
Today, the leap from strategy to execution is where marketing’s good intentions are hitting a wall. It’s where personalization breaks down, workload increases at the channel management level, and strong strategic decisions weaken and often disappear.
And so this is where we at Lytics focus our energies.
Does our CDP centralize data? Of course it does. We were one of the first to do it and we’re still committed to a strong data platform today.
But we also understand that you need more. You didn’t wake up this morning itching to solve a data problem. You woke up with marketing goals in mind. Data is only as valuable as its ability to support those goals.
This is how Lytics differs from other CDPs on the market today. We’re not just handing you data and expecting your channel managers to be data scientists. We’ve built data science into our platform. By the time our data gets to your channel managers, all the strategic data-driven decisions have already been made. The audiences have already been identified and filtered. Your channel managers don’t have to figure out who the right customers are or what content to match them to. They already know the answers based on data science, AI, and machine learning.
This is the key challenge facing marketers today. Not where and how to get more data—but how to truly harness it to reach marketing goals.
Can a CDP help you reach your marketing goals?
We know it can. If you’re curious about the power of a best-in-breed CDP, let’s chat. We’d love to give you a demo, answer your questions, and talk to you about the epic marketing results we’re seeing from customers as they implement this new way of thinking about and using data.