top of page

The Real Social Media Deal

Social media was once the great new thing. On social media anyone could publish anything at no cost and audiences would grow through following and getting a follow-back in return. 


A few years later the ‘how it started’ and ‘how it’s going’ pictures are miles apart. It is still true that anyone can publish anything for free but missing is a large, delivered audience, specifically a large delivered audience comprising a business’s target market.


A delivered audience is the number of impressions a piece of content receives. Algorithmic sleight-of-hand is the villain of the piece. Crafted at social media HQs, crafty algorithms ensured that business users, diligently having grown audiences of one thousand, ten thousand or one hundred thousand followers, had been following a road to marketing disappointment, because their delivered audiences today are a small fraction of their number of followers. 

…How Can This Be?

 
[1] Platform behaviours
For social media platforms active users are their oxygen. Success is related to the number of active users. Their strategic plan included creating an ecosystem with each user’s number of followers publicly visible to other users. 


For users, business and personal, having more followers was the performance indicator above anything else. A large number of followers being better than a small number was soon a universally acknowledged truth. 
 

Attracting users by being free yet requiring costly world-scale cloud infrastructure means the platforms have to generate income from advertising. The prospect of free marketing was the mirage that would lure businesses to become users. 

Never intending marketing to be free, the social media platforms embedded sliding-scale slippage into the algorithms governing the relationship between an account’s number of followers, and the number of followers seeing the content being posted.

Separately, the ability of most posts to go viral is stomped on by hard shrinkage of the extra views generated by shares and retweets. 


But through using a platform its users have opted into seeing advertising. That is the social media deal – even for private accounts. 


Detail on the inner workings of social media algorithms has been hazy. Googling the subject brings up plenty of remarkably similar blogs and opinion pieces conspicuously unruffled by actual data.

Wanting to know the facts, we designed a research project to collect structured empirical data at scale, and from which we reverse-engineered the core formulae governing the main social media algorithms. This is what we discovered:
 

Inter-row picture of forage maize
Impressions delivered by social media algorithms

Social media apps predicting the views that a campaign hashtag (e.g. #farm24) might have received from posts, retweets and shares don’t factor into the reported numbers the minimised reality; showing instead epic exaggerations. Many of the attractive and expensive social media dashboards have hard-wired a version of this mistaken principle of how algorithms function into their results screens. 

n the same way, the relative influence possibly wielded by influencers doesn’t equate to the number of followers. 

As the number of followers increases the sliding-scale algorithms stomp ever more heavily on the free reach brake pedal. As examples, a video recently posted by a well-known farming title with close to 90,000 Twitter followers achieved under 700 (i.e. 0.8% follower reach) video views. On Facebook a video from another well-known farming title, with over 220k followers, received 1k views (representing 0.5% follower reach). 

[2] Target audience behaviours

So far, all followers have been regarded as equal. Another picture appears if we consider that for every account the followers are comprised of a mixture of target audience and not target audience. For businesses in agriculture the target audience is nearly always those involved in farming and their influencers. Across the accounts of the hundreds of agriculture businesses we're tracking the central median is a ratio of 20% target audience to 80% not target audience. One in five followers has any value to the business.
 

This important and often unknown imbalance has multiple drivers: friends come first and few people follow businesses, TV and sport faces are popular with some, and of course farming folk favour following fellow farmers ~ but independence of thinking is highly valued; an individual's ability to influence lessens when perceived as posting for money or payment in kind. 

Or the target audience members a business wants may never see the account to be able to choose to follow it – low ‘findability’ is a high hurdle when more than fifty thousand other social media business accounts also want to reach UK farmers. Yet they are exactly who needs to be reached. 

 

The next table shows the monster target-audience crunching effect when the platforms and farmer target audience behaviours are combined.

Case tractor spreading out hay
Targt audiencevies delivered by socil media algorithms

So, posting content on an agricultural account with ten thousand followers may reach, on average, somewhere around 380 of the desired target farmer audience.

 

Houston, we have a problem. Most of what has been passing as social media marketing has been imaginary marketing.

How Twitter, Facebook and Instagram algorithms work
bottom of page