What are robots good for, anyway?
In the past few years, topics like automation and machine learning have become buzzwords.
Some welcome it while others fear it. So, when did they begin, what do they actually mean, how do they apply to digital marketing, and what does the future hold?
Digital Automation & Machine Learning: A History
Automation has been happening for years and it’s only accelerating. It’s no longer just the automation of mundane tasks like email newsletters, but access to valuable insights without having to lift a finger.
Let’s start right at the beginning. The concept of automation can be traced all the way back to Ancient Greece, when wind-powered machines were used for farming irrigation (Automatic Irrigation). Now, we expect these processes and consider them very simple.
Digital automation began with the advent of computing and quickly evolved from basic principles to the advanced artificial intelligence we regularly use today. It broke into the marketing realm when automated email marketing became popular during the 90’s and 2000’s. Soon enough, this led marketers to automate repetitive tasks, such as adding contact lists or creating landing pages, using data related to their customers or conversions.
On platforms like Google Ads, automation and machine learning have been moving from experimental new features to the only features available. For example, Responsive Search Ads (ads that dynamically pair headlines & descriptions) started out in 2018 as experimental features, and last month Google announced that soon, they’ll be the only option available.
This is also the case with social media platforms like Facebook and Instagram. They design your unique home feed based on your usage data, including what accounts or hashtags you follow, content you interact with, posts you spend longer on and what’s trending at the time.
Although Facebook Ads offers manual options for most campaign settings, automatic processes are nearly always available. We barely bat an eye at how simple automatic budget splitting, A/B testing, placements, formats and creative have made campaign creation, and marketers usually owe Facebook’s systems credit for their success.
What Is Automation?
Automation refers to processes where human input is minimal or non-existent. For example, if someone sends you an email or submits a form on your website, they may receive an automatic thank you for enquiring email before a personalised response. Essentially, automation takes simple operations and makes them happen without your input.
Automation saves you and your business valuable time. As a result, it improves productivity, helps minimise human errors, gathers valuable data, maintains standardised processes and ensures your employees are focusing on their skill sets rather than elementary tasks.
Automation In Digital Marketing
In digital marketing, automation promotes efficiency, great client communications, and personalised marketing strategies. Here are the most commonly automated functions of small businesses:
Let’s take a closer look at automated email marketing, customer relationship management (CRM) software and social media management.
Email Marketing
Email Marketing is the oldest form of digital marketing, yet still the most widely used. Frequent advances in automation now allow businesses to personalise the messages they’re sending to contacts on a large-scale, allowing them to nurture existing and potential customers without a significant amount of hands-on work.
With free and basic software such as MailChimp, you can schedule the date and time of your EDMs so they automatically send to pre-selected segmented contact lists. This means you can be in a Monday morning meeting or enjoying a cocktail on the beach without having to worry about hitting send – everything is already taken care of.
Softwares such as Vision 6 and Klaviyo support advanced EDM automation and are often more costly as a result. Depending on how subscribers interact with your content, automatic segmentation and workflows kick into place to guide them through highly personalised campaigns more likely to resonate and drive conversions and sales. For example, this may include SMS marketing, abandoned cart or service related emails.
Customer Relationship Management Software
Customer Relationship Management softwares (CRM) automates the process of tracking and recording information about your existing and potential customers. Although advanced CRMs like HubSpot may not be necessary for smaller businesses, having a CRM is crucial for any business looking to effectively manage and analyse customer relationships.
Information that CRMs generally automatically record include customer names, contact information, purchase or appointment history, emails and subscriptions. This data can be collected across multiple touchpoints, such as form submissions, social media and emails without much human help.
CRMs are also useful for setting reminders, internal communications, flagging important notes and tracking the effectiveness of your marketing campaigns. The best part is, this data is all accessible in one neat location. An effective CRM can be the key to setting your business up for success.
Organic Social Media
Long gone are the days of manually posting your social media content and diving through Instagram accounts to find the perfect influencer for your business. Now, tools like Hootsuite, Sprout Social and Tribe support scheduling, performance measurement and network management.
Scheduling automatic posts is a significant time-saver for any business owner or social media manager. Similar to EDM software, scheduling tools allow you to time-block post creation and sit back to enjoy dinner while your content automatically goes live at your pre-selected time. As well as taking the mental load off, these tools help improve your engagement due to more frequent and consistent uploads.
Influencer database tools also use automation to save you time and resources. Where marketers used to trawl through hashtags and accounts to find influencers that aligned with their brief and budget, databases now automatically pull together comprehensive lists of influencers suited to your preferences. They also keep track of your personal lists, who you work with, their details and communication between both parties.
What Is Machine Learning?
Machine learning is a subset of artificial intelligence (AI) that analyses data and algorithms to identify patterns and make decisions without human input. By running the same action over and over, machines automatically learn what the most effective outcomes are and pivot their systems to optimise whatever process they’re running. They become more intelligent over time.
Machine learning promotes efficient data handling, identification of trends and patterns, automation, and continuous improvement.
Machine Learning In Digital Marketing
In digital marketing, machine learning is used to understand what’s most effective for every potential and existing customer and, as a result, to provide personalised marketing that drives leads or sales. This is especially obvious in Facebook and Google’s advertising platforms.
Facebook Ads
According to Facebook,
“Over time, as more people view an ad, share feedback on it or click through to make a purchase on an advertiser’s website, our models get better at predicting the estimated action rate and ad quality. Since billions of people use our apps and engage with ads each day, our system gets lots of information to help improve its calculations, furthering our ultimate goal of maximizing value for both people and businesses.”
After launching a new social media campaign, Facebook Ads enters a learning phase as its systems test and learn from results, or lack of. At Springboard, we use A/B testing at the campaign level to split test different audiences, creative and ad copy in social media campaigns. This ensures the budget is being distributed to the most effective ads and maximising your results.
This is a form of unsupervised machine learning. Unsupervised machine learning finds hidden patterns in large amounts of data without human intervention or labels. They look at your customers’ behaviour and automatically sort them into groups based on relevant characteristics, consequently helping to locate and target high-value customers.
Google uses machine learning to optimise where your campaigns show and deliver the helpful and frictionless experience consumers expect from brands across many ad types. Firstly, it’s responsible for Responsive Search Ads. This ad type requires advertisers to provide 15 different headline options and four description lines, then Google’s machines test different combinations to learn which copy performs best. By adapting your ad’s content to more closely match potential customers’ search terms, these ads may improve your campaign’s performance.
Secondly, Google’s machine learning helps you get the most from your Smart Shopping campaigns. Aptly named, Smart Shopping uses machine learning to create and show a variety of ads across different networks, including YouTube, Gmail, Google Search and Google Display. In doing so, it tests and learns which products, placements and audiences are the most effective for your business while taking granular considerations into account, such as seasonal demand and price changes. Although it doesn’t allow for much manual management, it helps you hit your campaign goals through automatic bidding for individual products or keywords and certainly makes the process easier.
Finally, machine learning is used to drive foot traffic in Local Campaigns. According to Google, mobile searches for ‘near me’ have more than tripled in the past two years and people still make the majority of their purchases in physical shops. As the first of Google’s ad options to exclusively focus on foot traffic, Local Campaigns takes advantage of this by using machine learning to automatically optimise your ads across properties to drive shop visits.
The Danger Of Leaving AI To Its Own Devices…
While it may seem like automation and machine learning are putting digital marketers out of business, you cannot “set and forget” automated digital marketing systems. As with most technology, there can be disadvantages when they’re not managed properly. Humans should always be involved to prevent errors and ensure client needs are being met.
Data quality, cost and usage should all be considered.
Data Quality
Machines don’t have emotions or understand the world like humans do. No matter how much more efficient machines and automated systems are, it’s impossible to replace the human connection which is so important for creating unique, meaningful and emotional campaigns which resonate with customers.
Furthermore, if your automation relies on average data, it’s likely you won’t achieve great results. For example, Google Ads cannot generate a retargeting audience if you haven’t had more than 100 people engage with your business online. AI only performs the tasks it’s required to do and lacks the ability to think out of the box. If automating your efforts requires more maintenance than the time it would have taken manually, consider whether it’s worth it.
Cost
Tech giants spend millions of dollars on implementing and updating new AI – it cost Apple over $200 million to acquire SIRI, and over $26 million for Amazon to acquire Alexa. Although small businesses don’t use AI to this degree, expenses increase as systems get more complex and modern.
Let’s use email marketing as an example. MailChimp’s Essentials package costs $9/month and offers A/B split testing and multi-step journeys. The Premium package costs $299/month and offers social post scheduling, multivariate testing, pre-built journeys, branching points and multiple starting points. The Premium package obviously supports a more personalised marketing experience, but it comes at a cost. Of course, the level of automation and machine learning your business needs is highly personal, so it’s important to find the right automation solution for you.
Data Usage
You don’t need to automate everything. Automation and machine learning should be focused on tasks that are time consuming and will provide genuine value to your business. It should also not be considered ‘outsourcing.’ Using data will take you so far, but it’s imperative to involve humans during decision making processes.
Use the data your systems collect for good (we’ve all heard of the Cambridge Analytica scandal!). Just as machines learn from algorithms, humans can too. Analyse your data with a fine tooth comb and use them to your advantage – how can you improve a system, ad, campaign or client relationship? At Springboard, we consult with our clients to analyse and explain data related to their SEO, SMM or PPC campaigns to establish a strategy moving forward and set them up for success.
How We Use Automation At Springboard
Springboard Digital was founded on the principle that high-quality digital marketing doesn’t have to cost millions. We aim to be a long-term digital marketing partner for the small businesses of Australia, or larger organisations with less complex digital marketing needs.
The key to succeeding with our vision lies in automation.
As we’ve established by now, we can’t simply automate everything.
Our experienced marketing professionals know what signs and red flags to look for in campaigns, how to interpret results and understand the nuances of campaign structure. We value this and simply cannot replace it with robots.
In saying that, large portions of digital marketing work are simple, tedious tasks which are completely automatable.
For example, many digital marketers will conduct manual budget checks multiple times per month to ensure Google isn’t overspending or underspending the assigned campaign budget.
Thanks to our tech wizards, we now have an automated reporting system that retrieves spend and budget data from Google Ads daily and compiles it in a spreadsheet for our manual review. It calculates how much of the budget has been spent, how much remains, and recommends new daily budgets where necessary. Automation has made this process efficient and simple while achieving the same result.
We also use a number of third-party automation tools to optimise accounts, such as Opteo. They suggest new keywords, scan industry-level keywords for opportunities, and suggest when we should refresh ad copy and creative.
From an internal marketing perspective, we use Hootsuite to schedule organic social media posts, and Hubspot to manage our client relationships.
Although humans are still required to initiate or monitor many of these actions, automation allows us to work more efficiently so we can focus on delivering the best results for our clients.
When the robots do their thing, we have more time to do human things.
If you want to experience the power of automation in digital marketing, reach out to the Springboard Team today!