Connect with us

Indago Digital

Machine learning and the future of marketing


Digital marketing enables us to track every dollar; every click, every open, every ‘sign up now’ and accumulate lots of data in the process.

As a result, marketers have become laser-focused on the numbers and sift through opportunities to optimise and scale at speed. It’s instilled a meercat-like propensity in us to scan the landscape for the next ‘wunder-app’ or platform innovation, but some might argue that’s been at the expense of ad and product quality. Could machine learning right the ship?

TALKIN’ ‘BOUT A REVOLUTION

Big tech platforms have made continuous headway when it comes to delivering results using machine learning. As they’ve evolved and improved their capabilities, they’ve coaxed advertisers into use automated bidding and campaign types. It’s been more of a stick than the carrot approach, with manual options being removed in favour of automated ones (i.e., see you later Expanded Text Ads; hello Responsive Search Ads as the only creative option) with little or no alternative. I daresay the reason is that often we have to be pushed before we’ll take the leap. We’ve become so accustomed to being in the driver’s seat and analysing and tweaking every facet of our ad campaigns that relinquishing control is nigh on impossible.
I want to suggest a positive outcome that could arise from transitioning some of our ad campaigns to automated options though.
By handing some control over our ad management to machine learning, we could free ourselves from fixating on volatile quick win and rapid scale strategies that ‘chase attention’ (that’s a Gary Vee-ism) and welcome a return to more sustainable long-term growth strategies that consider the full customer journey, from acquisition to retention.

MACHINE TO THE MAX

The latest machine learning development comes out of the Googleplex; it’s called Performance Max and it’s Google’s most automated campaign type yet. Google has been announcing new bidding options at breakneck speed, but Performance Max goes a step further by combining automated bidding with cross-channel optimisation.
Based on a conversion goal, Performance Max leverages Google’s entire inventory from a single campaign. Once you’ve chosen your budget, bidding strategy, landing pages and creative elements, your Performance Max campaign sets off in pursuit of potential customers across YouTube, Display, Search, Maps, Discover, Gmail and all the rest of Google’s channels. It combines Google’s understanding of user intent, behaviour and context (based on 3.8m searches a minute!) with budget optimisation to serve your most engaging ads to the right people at the right time – capturing new conversion opportunities as a result.
Google reports that advertisers who use Performance Max campaigns in their account see an average increase of 13% total incremental conversions at a similar cost per action.
All this is great, right?
The answer to that question comes back to your marketing ethos.

GOOGLE ATE MY HOMEWORK

As it stands now, marketers who like to have control over every detail of how their ads are being served will find Performance Max campaigns challenging. You can see which creative combinations have performed best, but you can’t ascertain how many impressions or clicks it received, or glean any meaningful insights on audiences, channels or keywords. It’s not a set and forget – as with all automated strategies, manual oversight interplaying with machine learning is the sweet spot, but marketers who are used to a lot of visibility and the ability to tweak and test will find it hard to justify underperformance when it’s simply a case of, “Google didn’t hit the target”. There is also no way to set keywords and you can’t add negative keywords either.

MACHINE LEARNING: THE GREAT MARKETING LEVELLER?

As automation and machine learning continue to develop and outpace human intelligence, the playing field will start to level out and the advertising opportunities available to marketers will very likely become equal. Facebook advertising is on the same trajectory. They already offer automated placements across their channels, and they’ve started to remove target audiences now, too. As they dwindle down the options, they’re encouraging advertisers to use broader targeting options and rely on their machine learning to find their audiences.
It won’t be long until machine learning-powered paid media channels are showing products that have the same purpose to the same people at the same time they’re ready to buy them.
Marketers who are comfortable with automated ad management processes and making periodic tweaks will be the ultimate winners of the automation age, but only if they use the time it affords them to start playing the long game.

THE UPSHOT: FOCUS ON THE FUNDAMENTALS

When your campaigns, audiences and funnels are all built by the same machine as those of your competitors, the only differentiators you’ll have to play with will be better ads, better landing pages, better offers and better products. My hope is that this shift will bring about a return to growth marketing in place of growth hacking.
While growth hacking wants high acquisition and rapid growth, growth marketing is a long-term strategy that uses innovative, data-driven techniques to nurture the brand-customer relationship. Instead of clambering for snippets of attention across umpteen channels and optimising them all like a DJ at a drum and bass festival to get the most leads at the lowest cost, growth marketing is concerned with developing market leading products and services, building loyalty, enhancing customer lifetime value, producing exceptional content that clearly communicates benefits, and doubling down on-site optimisations to increase conversions and achieve enduring market dominance.
While we wait to see how the evolution of machine learning unfolds, one thing is certain – automation is the future of advertising. Platforms are investing heavily, and the pace of change is swift.
The best way to realise the opportunities for your business and future-proof your career is to be at the forefront or, at the very least, to start running some tests. The rewards are likely be worth it, with improved marketing performance, increased productivity, and the ability to shift the balance to more tactical, customer-centric and sustainable ways to grow your business.

Advertisement

Luke Ashmore-Delaney, head of performance marketing at Indago Digital

The post Machine learning and the future of marketing appeared first on Mumbrella.



Source link