Measuring Content Success Part 1: Gathering data

How many likes does a social media post need to be classed as successful?

Imagine an organisation with 1,050 followers on LinkedIn. A recent post of theirs got 20 likes and 5 reposts. Is that good? Bad? Amazing? Did it get more likes because there was a link? Or because of the time of day it was posted?

Measuring the success of content is proving tricky for your organisation. So how should you go about it?

Many people use general online advice as a yardstick, and that’s often based on other users’ engagement. If a similar organisation or competitor gets 40+ likes per post,  they set the bar just as high. But is that really the right method? 

Not according to our Head of Data Science, Oliver Paul: “Comparing data across customers is like comparing apples to oranges.”

“Some customers publish multiple times a day, Others have over a million followers. Some share original reports, while others post light-hearted memes. Success for one customer does not equal success for all customers.”

In this series we’ll be exploring how the collection, management and organisation of data can help give you a more accurate measure of success. 

Step one: gathering enough data. 

“We cannot draw conclusions from what we cannot see.”

Measuring success starts with having something to measure - enough of it to avoid being led astray by the occasional fluke. 

A one-off post might perform 10x better than your average post, boosting your average engagement. The less data there is to measure, the more likely it is that spikes in engagement like this could give you misleading results.

“Averages are less likely to be skewed where there is a large sample size.”

Imagine a study about the rise of vegetarianism. You interview 30 people in your local town, and find they are all staunch omnivores. Is that indicative of reality? Not really. A better approach could be to interview 500 people from 10 different cities across the country. Now you have a diverse and more comprehensive dataset to work with.

To explain this, we’ve taken the perennial question: when is the best time to post on social media?

Ollie mapped out some partners’ social posting data to see what times and days they were posting at the most:

“We can see we have very little posting activity on weekends and early mornings. It may be the case that these periods are great for posting, but we will never know unless we start posting at these times.”

We will be less confident in making conclusions before 8AM and after 3PM since we have much smaller sample sizes.”

The more content you post, the more data you have, and the more informed your next decisions can be. Testing the success of different types of content would follow the same process: publishing videos, carousels, text, images and more to gather enough data to ask questions. 

Struggling to post high volumes of high quality content? That’s our bread and butter. We help organisations grow, engage and understand their audiences better. Drop us a line if you’re interested in learning how. 

With a great deal of data comes a great amount of responsibility - and knowledge. Now you can spot patterns, draw conclusions, and ask more granular questions about what works and what doesn’t. 

That’s where Part 2 comes in: contextualising your data, A.K.A, normalising it to find answers. Stay tuned!