Yes, yes, artificial intelligence is threatening millions of jobs, everyone will be replaced by robots, machines will do every job better and faster, yadda yadda yadda. We know all that. We've been going on and on about it for years.
But so far, it ain't happening. Not on the ground. Not with the companies that we work with. Apart from a handful of digital leaders perhaps, the rest of the business world has precious little to do with AI for now. Chat bots are still crappy and most AI solutions available for marketing are nothing but rudimentary automation tools.
So, when will it happen, and where will it happen first? Let's discuss!
First way of the dodo: PPC
Whereas branding, creative, and strategy are - for the moment - safe from AI, there are some aspects of marketing that can already be executed better by AI. The first industry to bite the dust will be PPC and its cousins. There is no use for humans fidgeting with ad formats, bid adjustment, and link optimization. You are trying to beat an algorithm that gets smarter by the day.
At the moment, Google itself is offering AI-based suggestions how to improve your campaign, Very often the suggestions are useless. We run a campaign for branding and marketing consulting and Google keeps telling me to add the keyword "movie theater" for whatever reason. But it is improving.
The next level of AI involvement will be automated adjustments to bids, alerts for quality scores, and budget optimization. This will have to be done by a third-party provider with access to your PPC data. You wouldn't trust Google or Facebook to "optimize your budget".
Soon you will see Google and Facebook offer fully automated solutions, so don't bother with your PPC certificate. Optimizing digital campaigns is the first marketing activity to be done completely by AI. There is also an incentive for Google to offer data-driven creative solutions, but we expect more trial-and-error before these really make an impact.
Death to A/B testing!
We love A/B testing, but it's the next thing on the dodo list. Companies which base their business on A/B testing services will ultimately go under as sophisticated AI will do the testing for you. IBM Watson’s Cognitive Technology for marketing can already suggest what to test and how to test it. Phrasee is a tool which analyzes customer reactions and responses and then, through machine learning, suggests how to adapt your language to better engage with your audience. Many more are on the horizon.
Goodbye SEO companies!
At this point, SEO companies are still offering values, because there are still things that have to be done manually, by programmers and coders. But as we move forward, most of what an SEO company does can be done by AI. Amongst the thousands of SEO consultants out there, only those that offer extra value or develop their own analytics tools and sell subscriptions will survive. We are more inclined to think that Google will come out with a service that automatically optimizes your pages. AMP hasn't lived up to the promise but it is growing. Wordpress is another candidate to offer fully-optimized web presence that continuously optimizes for SEO.
Social media posting
Social media posting will also be a victim of AI--to a point. At my company, we decided on MeetEdgar to manage our own and clients' posts. We can expect this kind of AI to get better over time, not just posting at random intervals, but posting and re-posting content that actually makes the most impact, taking under consideration reader reactions, shares, and likes, etc. This has freed up time for our marketing team, who can now concentrate on the creative process.
Who's next?
In general, we see the areas most likely to be impacted by AI as follows:
- Repetitive tasks such as posting and reposting
- Data analysis tasks, with the results, then used to improve creative
- Optimization based on machine learning
- Processes following strict rules ("if this then that")
- Interface to business intelligence systems
- Comparing investments and ROI balancing
- Analysis and reporting, although the current output of analytics tools is far too technical for management to accept
What AI will not replace (for now)
Marketers should not be threatened by this development. The tasks we listed above should be taken over by a computer. They are repetitive, data-driven, and uncreative. By turning them over to machines, marketing teams will have more time to focus on the creative process, strategy, and all other tasks humans excel at.
What AI certainly can't replace is experience and intuition. Most AI systems are blissfully unaware of the difference between countries and cultures, and will blindly make suggestions based on their creator's inherent bias. So if like we, you are working in multi-cultural marketing, bridging gaps in understanding and helping companies to internationalize, then your job is safe.
We also expect a backlash against machines, in various degrees. The more automated the digital space becomes, the more users will see through the fact that they are being shown ads and fed content through algorithms. There will be a movement for "humanity" and marketing and a craze for authenticity. For the foreseeable future, marketers will grapple with AI, and employ machines to make their lives easier. Any talk of AI replacing marketers in large numbers is surely premature.