The critical skills every performance marketer needs

Published Aug 8 2023 Last updated Apr 3 2024 8 minute read

Back in the 1980s, the burger chain A&W wanted to beat McDonald’s and its famous quarter pounder burger. Their idea was quite simple: launch a bigger burger, and thus the third pounder was born. To the firm's despair, it was a complete failure.

After some consumer research, the problem was identified: people could not understand why they were paying more for a smaller burger. After all, 3 is smaller than 4, and so a third pounder must have less meat than a quarter pounder.

This is not a post about burgers, but this true story illustrates an unfortunate truth: most of us struggle with numbers and math.

A reason why marketers could be struggling with numbers

Marketers are no exception. Some of us lack training in mathematics; some even avoided it altogether in their studies. That could be why our research shows that 33% of media buying and performance professionals report a gap in data skills.

I would like to believe that all of us know that 1/3 is greater than 1/4, but this is only the tip of the iceberg when applying math skills to data interpretation. Modern performance marketers, or at least the good ones, work with numbers daily. Data is the lifeblood of digital marketing, and working proficiently with it is vital if you want your business (and career!) to thrive.

The 3 most important performance marketing skills

So what are the 3 most crucial skills if you want to be a data-guided marketer?

1. Mathematics

To be great at marketing, you need to be good at mathematics. Now, no one argues that you should master inferential calculus or prepare a TED Talk on numbers theory, but there are some mathematical concepts you should at least be comfortable with.

Linear algebra

The first one is linear algebra. Yes, the same one you learned way back in high school. The one those pesky math teachers claimed you would “use every single day.” 

We’re still rolling our eyes at that claim, but it is actually true. 

Calculating conversion rates, click-through rates, or cost per acquisition. Simple examples, for sure, but they are all linear algebra. While you can go down a deeper rabbit hole that includes line planes, plotting data points in 3D spaces, and matrices, it all starts with linear algebra. In fact, many modern machine learning models have their roots in linear algebra.

In the day-to-day life of a digital marketer, though, linear algebra will help you to clean, transform, and scale your data. Cool, right?


Since I just talked about rates, it’s a good time to also mention percentages. Humans are actually quite bad at thinking about percentages, fractions, and decimals. One theory called 'whole number dominance' claims this is due to many objects we encounter are based in whole numbers.  

Think about it for a second. When was the last time you encountered 3.14 trees? As a result, we evolved to think in absolute numbers. 

But percentages are everywhere in marketing. Many key performance indicators in various marketing channels are percentages: click rates, engagement rates, conversion rates, retention rates, bounce rates, open rates, and so on. If you want to be a marketer that thinks as a marketing analyst, you must be confident with percentages.

Probability and statistics

Finally, we have probability and statistics – two areas that have been growing in importance for marketers, and the next steps once you have mastered the basics. 

These are no simple domains, for sure. But knowing the basics will help you create regression models to track campaign performance, explain what is happening, forecast more accurately, and predict customer conversion rates.

A concept like marketing mix modeling, for example, is pure statistics. And although smart tools help you with that, understanding the maths behind it will help you interpret and tweak models.

MMM or Marketing Mix Modeling helps you determine which marketing channels contribute to sales or leads. For more about it, watch the video here: Marketing Mix Modelling.


2. Spreadsheets

Some people think you need to learn R, Python or SQL to perform data analysis. They are not completely wrong either. If you want to be a data scientist, you should probably master at least one of them.

As a data-minded performance marketing manager, you don’t need to go that far, though. Good old spreadsheets can take you a long way if you are willing to go beyond the basics.

Pivot tables

One of the most crucial spreadsheet functions you will need to master is pivot tables. A way to envision a pivot table is when you turn a column into column headers, aggregating the values within them. Maybe a visualization can help you understand it:

In Google Sheets (my favorite) or Excel, you can create a table like the example above by simply selecting all your data, going to Insert, and selecting Pivot Table.

When you do this, you get an easy-to-use drag-and-drop interface that allows you to aggregate  and investigate your data in different ways. By selecting other pivots and aggregation methods, you could understand average sales per category over time, total sales per product, or the share of each category within overall sales. You could answer all of these without any complex formulas or functions.

But before you start pivoting, you might need to do some cleaning and merging data. And there is no way of doing it without some handy formulas. Here are some that you should keep in your toolkit at all times.


This might be one of the functions that I use the most. It stands for vertical lookup, allowing you to look for a value in one column and return a value from another one. It is extremely helpful to create categories, which you will then use in your pivot table.

Here you can see a classic example. Use the ProductID to look into a Dictionary Table and return the product category.In this case, “Swimwear”:

IF Statements

Working with conditionals is key for data analysis, so it’s only natural to use it in spreadsheets. First, be sure any data analysis will make use of the SUMIF() function, which allows you to sum a value based on certain conditions. It’s convenient when you are working with super granular data and want to group it. And remember the AVERAGEIF(), which is the same logic but for means, and the variation SUMIFS() and AVERAGEIFS(), which are for multiple conditions.

Of course, conditionals are not limited to sums and averages. You can perform extremely complex functions by using IF() combined with AND() and OR() functions.


This is not a particular function, although it can be used within some. REGEX stands for regular expression, and it’s a powerful way to find patterns in text. For example, if you are a bit inconsistent with your naming conventions, you can use REGEX to find patterns and categorize your data. When you master REGEX, you can use things like REGEXMATCH(), REGEXEXTRACT(), and REGEXSUBSTITUTE() to clean your data like a Pro.

There is more. Spreadsheet software has been around for more than 30 years and is so versatile that there is no way I could cover everything. However, I believe that by acing your pivots, checking your lookups, and managing your ifs, you are already a step ahead and ready for great data analysis. 

This takes us to the next key area you should focus on to be a data-driven marketer.

3. Data visualization

Data insights are worth nothing if they don’t lead to action, and to get the ball moving you will need to explain your findings to others. This is where data visualization becomes an essential skill for marketers. It’s a complex field involving design, psychology and (again) mathematics. You can find some awesome resources that we have published on this already:  

But I would like to give you some things to consider about “data viz.”

Keep it simple

When building a visualization, you are working with six elements: position, shape, size, color, line width, and line type. In theory, you can add a different dimension to each, but things can easily get messy. For example:


This chart is trying to communicate too much at the same time:

Each point is a unique product

  • The X position shows how many units sold
  • The Y position shows the average price
  • The color shows in which campaign it was sold
  • The shape which category it belongs to

It doesn’t matter how long you look into this. You probably can’t answer simple things like: Which item generated more revenue? Which campaign sold more products? Which category has the highest average sales price?

Always remember that a good visualization is easily understood and delivers the main point quickly. To get there, you should always start your visualization by answering: what do I want people to take away?

Make conscious choices

When building a good visualization, every choice you make matters. The first choice you need to make is which type of chart you will create. The following flowchart can help you with that. But notice how it all starts with knowing what you need to get across. Is it a comparison? A relationship? Without answering this first, no matter how much time and effort you put into your design, your data visualization will achieve nothing.

After picking the chart, choosing the right colors is essential. Not every visualization needs to be colorful either. The example above is black and white, but still makes the point. Color can help your information stand out, though. 

When picking the right palette, there are some things you should bear in mind.

First, some colors have strong associations. The classic example is “red = bad, green = good.” People are just conditioned to read things certain ways, depending on the context. 

For example, take a quick look at the image below. Are the results good or bad?

This image from Facebook Ads Manager shows a clear rule: if the percentage increases, it’s in green, while if it decreases, it’s red. But as a general rule a lower CPC is better while a higher CTR is better. So, in this case, both change (%) columns should be green, right?

You need to pick a suitable palette to reinforce your point. 

If you want to show how different two data points are, pick colors with great contrast. If you want to show unity, keep the shades closer to each other. Also, when choosing colors, think about accessibility and choose color-blind friendly palettes. It’s a lot to think about, I know. But luckily there are tools to help you, and it’s wise to rely on them.

Finally, keep it consistent once you have picked a set of colors. If you are building a presentation where you will show the same breakdown several times (for example, performance per country or platform), use the same color for the same category on all charts. It will help people understand the overall picture.

4 more tips for performance marketers

If you've been working in digital marketing for a while, you may have heard some of this advice. But just to be on the safe side, here are 4 tips to help you become a successful performance marketing manager.

1. When using bar charts, always start the axis at zero

Starting at other points of the scale might mislead the reader and make any differences seem bigger than they really are. For example, these two charts show the same data. On the left, the bar chart starts at 77 years, on the right at zero. On the left, we get the impression that Iceland's average life expectancy is considerably higher than Greece's - when the truth is they differ by 2.3 years.

2. Time goes from Left to Right

We are conditioned to visualize time on a horizontal scale, and going the same direction as our writing (which means if you are building for a right-to-left reading audience, you can flip this advice). But please, don’t ever do a timeline going from top to bottom; it just doesn’t make sense.


3. Use directions that make sense

This is on the same line as the point above. People tend to have some assumptions as to how the data should be shown. For example:

  • The horizontal axis goes from small (left) to big (right)
  • Vertically, small is on the bottom, and big is on the top
  • In a Scatterplot, the explanatory variable is on the horizontal axis, and the outcome is on the vertical

4. Build for your audience

As I mentioned, the initial point of any visualization should be answering,“What do you want people to get out of it?” This is true. But a large part of that question is “people”. Remember that everyone has a different context and frame of knowledge, so you must build for the audience. That means explaining acronyms, using annotation to reinforce points, and giving background explanations when necessary.

Wrap up

Marketing is a numbers game – there’s no way around it. These tips are in no way exhaustive, and there are whole fields of knowledge to explore. In that case, you may want to dive into a robust statistics course or even learn R or Python. Meanwhile, if you want to ramp up your data viz skills, you may want to start building a few different dashboards.  

My point is that getting acquainted with these fields can make you a better marketer: one that can look into numbers, quickly identify trends and patterns, gain insights, and turn them into action plans.

If you want to become a better marketer, you can still learn more about social media marketing, offline and online channels, consumer psychology and even search engine optimization. But make sure you don't forget to also develop the 'technical skills' outlined in this post.

Grow your business, and your career

I hope this was inspiring and motivating. Because the future of marketing will be full of data, and data skills will be increasingly necessary. By sharpening up your skill set you will be able to understand better where to optimize your campaigns, react faster to market changes, and, by the end of the day, grow your business and career. 

Want to work smarter with your marketing data?
Discover Funnel