Making the most of your data: how to use statistical analysis to improve your marketing

Statistics help you to make sense of your data which, in turn, helps you make better marketing decisions.

How do you know that your marketing is having the impact you think – or hope – it is? To truly understand the progress your business is making, it’s essential to analyse your marketing campaign data. Data shows you what is happening, but also why it’s happening.  

But there’s so much data out there that it can be hard to know where to start. In this blog, we reveal where to find your data, how to analyse it, and how to use these insights to guide your marketing decisions and strategy.

Where to find your data

The first step to making sense of your data is to collate the data you need into workable files. Google Analytics is a great place to start. This tool can monitor all aspects of your website and track users, pageviews, and even sales performance. 

Data from your social media platforms is easily accessible through platforms like Hootsuite or Iconosquare. LinkedIn even provides a year’s worth of data on your company LinkedIn page. This allows you to monitor the performance of your posts and reveals how users interact with your content through metrics like user engagement and click-through rate. These insights can guide future marketing decisions. 

Both Google Analytics and LinkedIn allow you to export your chosen data to Excel, for example, which gives you more freedom to engage with the data in ways most useful to your business. This may mean exporting data from a specific time period, like during a new marketing campaign. Or maybe you’re only interested in views on a particular webpage and not on the rest of your website? Exporting your data will allow you to concentrate on the data that matters most to your marketing.

The tools of statistical analysis

Now that you have all this data, what can you do with it? Currently, your data will look like rows and columns of dates and numbers. Each individual data point is meaningless on its own – when you’ve got so many figures laid out in a row, it can seem hard to discern its meaning. This is where statistics come in. 
    
There are countless tools at your disposable to make statistical analysis as easy as possible. We like to use Excel, but other tools like Minitab or Google Sheets can also be used. Completing statistical analysis in Excel is as simple as clicking a button. All you need is the right tools to interpret the results it gives you – no complex calculations are necessary.

The basics – graphs, trendlines and averages

Graphs
There’s no better way to begin your analysis than with a graph. Being able to visualise how your business is progressing over time is critical to understand if your current marketing strategy is effective. Plotting sales or website visitors over time will immediately highlight anomalies, successes and areas for improvement. Maybe the outcome of your analysis shows increasing numbers of visitors to your website but no increase in sales. This gives you the opportunity to think about why that would happen and how to resolve it.

Tip: Hover your mouse over the data point you’re interested in to identify it.

Trendlines
If you’re using a lot of daily data, your graph may look quite erratic, and you still may not know where to start. If this is the case, insert a moving average trendline to get a smoother curve that still represents your data. 

To get an idea of the trend over the whole graph, add a linear trendline, also known as a line of best fit. This will help you to quickly establish answers to key questions you may have about your marketing ROI. For example, are you growing engagement on social media? Are sales increasing at the rate you expect?

Averages
You can also learn how engagement with your website and social media changes from month to month or week to week. Or maybe you want to know if there’s a day of the week when posting to your LinkedIn will get the most traction? Averages are a great way to easily compare segments of your data. 

In Excel, use the AVERAGE function to quickly calculate averages from your selected data. These simple but useful tools should help you to understand your data. Being able to see trends and changes over time will ensure you make informed and effective marketing decisions.

Correlation versus causation

Do you want to take your analysis one step further? If so, you can use correlation and regression analysis to understand how your marketing affects your business. 

Tip: To do this in Excel, you’ll need the Analysis ToolPak add-in.

Correlation refers to how strong the relationship is between two variables. For example, if you know that publishing your company blog on a Tuesday results in more views than if you publish it on a Friday then there’s a correlation between the day you publish and the number of views that blog gets. Mathematically, Excel calculates what is known as the correlation coefficient. This is a number between -1 and 1, with -1 being a perfect negative correlation and 1 being a perfect positive correlation.

The closer your correlation coefficient is to 1, the stronger the relationship between two variables. For example, you may want to examine the relationship between your LinkedIn views and website views. Is there even a connection there at all? If your correlation coefficient is under 0.6 you can assume that, while there may be a correlation, it’s not a very strong one. 

But what about regression analysis, and how is it different from correlation analysis? There are multiple types of regression but, broadly speaking, regression models show how changes in one variable (or sometimes multiple variables) influence changes in another variable. Correlation is part of this, but correlation doesn’t equal causation, and further regression analysis will help to distinguish the difference. 

When regression analysis is done in Excel, it presents you with a large table that may look a little intimidating. But the R Square value should be easy to find, and this is what you need. R square will tell you what percent of the change in your dependent variable (AKA, the effect) can be explained by your independent variable (AKA, the cause). For example, imagine you want to know how the amount you spend on advertising affects your sales. Say your R square is 0.642. This would imply that 64% of the change in sales can be explained by the cost of your advertising. 

In other words, there is causation, not just correlation.

Tip: To check how statistically significant (AKA reliable) your R square is, find the Sig F value in your Excel regression table and check that it’s less than 0.05.

Bringing the data findings together

We’ve been working with a long-time client to evaluate trends in their online presence over the past year. One of our aims was to assess engagement with their LinkedIn page and see its impact on visitor numbers to the client’s website. 

We began by looking at website users over the past year and created several graphs using data from Google Analytics to track users, pageviews, average time on page and bounce rate. Apart from a slight drop in January 2022, trendlines showed us that all these metrics have either increased or remained consistent since August 2021. 

Bounce rate can be a particularly interesting metric to investigate. It tells you what percentage of users visited and left your page without interacting with the rest of your website. This can be a good or bad thing, depending on what you want your visitors to do. A higher bounce rate could imply that your website is easy to use, and information is easy to find. 

LinkedIn provides engagement statistics per day and per post, allowing you to see when people are looking at your page and what posts perform the best. This was our next step. Again, we created several graphs to track post performance over time. LinkedIn even calculates an engagement rate for you by looking at likes, shares and clicks versus impressions. We were pleased to see that impressions have been steadily increasing over the past year. However, engagement and click-through rates have decreased slightly. This sparked follow-up questions: why is this? Is this due to an increase in impressions but not likes, shares or clicks? And if so, what do we need to do to rectify this?

Our last piece of analysis involved a look into the relationship between the client’s website and LinkedIn. Correlation and regression analysis showed there was little to no correlation between the two. Again, this raises questions: are people just not clicking on the blogs posted to LinkedIn? How can we make sure that more people do click on the blogs?

Using statistical analysis to guide marketing decisions

Now that we have our findings, we can use these findings to guide and improve our marketing decisions. The result of our data analysis was real insights into website and LinkedIn performance. Without statistical analysis, we may have concluded that more impressions are a good thing and looked no further. Now that we know people are not clicking through from the LinkedIn posts to the website blogs, it gives us the chance to think about why and, hopefully, lead us to improve. 

And you can do the same. With the help of handy tools like Excel you can gain meaningful and actionable insights from the data you’ve collected. Data analysis is essential to understanding the what and the why of your marketing and can help you achieve higher ROI on your marketing investment. Data analysis can show you what works and what doesn’t work in terms of engagement, conversion and brand awareness, and is critical to marketing performance and business success.

What next? Context

Context is just as important as the data itself. Check out our blog on the value of context when analysing data for more information. 

This step-by-step guide (with images) to using the Analysis ToolPak may also prove useful. 
 

If you need support to write engaging, audience-driven content for your B2B brand or agency, contact Copestone today.