Decoding algorithms for marketers
- 01 June 2018
- 267 views
Algorithms are arguably the most significant technical development to hit the marketing world in recent years. It’s important for marketers to know what they are, how they work, and how they impact the profession.
What is an algorithm?
From global financial trading and search engines, to dating apps and medical diagnostics, algorithms underpin the modern world. Experts predict an exponential rise in their use into the next decade as they become more sophisticated. But what are they, and why do marketers need to know about them?
The answer is surprisingly simple: algorithms are a set of instructions for doing a particular task. In the digital age, these instructions are typically carried out by automated computer programmes, designed to make consistent decisions, after processing reams of data, without the intervention of human beings. Some aspect of ‘machine learning,’ whereby the programme rewrites and optimises itself as it works, is usually involved.
How do marketers benefit from algorithms?
Marketers and brands have been quick to see the potential of this technology. The theory is that one computer algorithm can predict the future wants, needs and actions of potentially millions of customers, by examining the data trail left by their online shopping activity, browsing habits, and social media posts. The goal is to run personalised, targeted campaigns ‘at scale’, pushing the optimal communication (with the right creative) or special offer to the right customer at the right time in the right place. On the face of it, algorithms save enormous amounts of human labour, and predict behaviour far more accurately than a person could.
Examples of algorithms in action
One of the best-known algorithms is used by Google to produce the most relevant page results for your search terms, which leads businesses to spend significant time and energy ‘optimising’ a brand’s homepage. Elsewhere, Amazon’s algorithm suggests products to accompany a recent purchase, while Netflix automatically suggests other shows you might enjoy watching. Meanwhile, our experience of social media, such as Facebook and Twitter, is totally shaped by algorithms that dictate what we see in our feeds.
Some brands have deployed algorithms in highly creative ways: Snickers changed its price more than 5,000 times in five weeks in Australia by building an algorithm that gauged the prevailing mood on social media. This so-called ‘hungerithm’ automatically adjusted the price of a Snickers bar according to ‘how angry’ social media became; the greater the vitriol, the lower the price. This clever application of their "You're Not You When You're Hungry" brand positioning led to a sales increase of around 67%.
Do algorithms threaten marketing?
While the evidence suggests that algorithms are very effective, their use has attracted criticism and calls for caution. A huge amount of negative media coverage has been directed toward social media giant Facebook for not being able to prevent its algorithm from promoting fake news and extreme content.
Also, there is an ongoing debate within the marketing world about whether algorithms are the enemy of creativity. The argument is that, in the past, marketers and advertising agencies had to rely on research and human insight to come up with a visual or concept that captured the imagination of consumers. As algorithms can analyse and predict past and future responses to creative executions, there is no need to ‘risk’ an untested idea. Some creative practitioners are concerned that, if this trend continues, the art of creativity in marketing will be gone and, with it, their jobs.
Moreover, algorithms could spell a threat to advertising’s function as a catalyst for discovery. For example, retargeting algorithms aren’t yet smart enough to know when a customer has already bought an item or lost interest. This only feeds into a growing culture of advert rejection and could be behind the rise in ad blocking, as technology originally intended to make advertising more effective becomes a nuisance.
The uncertain future of algorithms
If an algorithm’s strength is determined by the data that helps to shape it, the dynamics that helped to bring about GDPR may spell an uncertain future for this technology. The public is becoming wary of the developments in artificial intelligence, and the idea that hyper-personalised content builds a ‘filter bubble’ around users online. Equally, for marketers who use algorithms to purchase and sell advertising space in real time (known as programmatic advertising), the technology has been tarnished by the idea that it encourages ‘herd behaviour’ by over-exploiting the popularity of certain products. Studies show that product and content ‘voting tools’ rapidly begin to focus on the top few products and ignore the rest.
A heuristic approach
Many in the industry, accepting that algorithms aren’t a silver bullet, have realised that heuristic thinking has a role to play. A heuristic method of learning involves discovery and problem-solving, using reasoning and past experience. It basically helps the user come to a ‘good enough’ solution – close enough, but not the full answer. It is thought this approach will give marketers, say, an 80% chance of generating leads, sales and other marketing targets.
In essence, while algorithms are a massive labour-saving device for marketers, it’s worth remembering that algorithms can only do the best with the data they’ve got. Also, smart marketers will test their channels and tactics – using the heuristic method – to figure out their messaging, unique value proposition, what ad copy resonates, and so on, to find a scalable, repeatable process – a humanised algorithm – to secure new customers and keep the business competitive.
To keep up to date with the latest trends in algorithms and SEO marketing, attend our advanced SEO training course.
CIM members can listen to our latest digital marketing webinar on demand now.Back to all
- 267 views