The Future of Online Advertising and Machine Learning in 2019
My name is Nacho Rodés, I’m the CEO and founder of Spaceboost. This is the story of how we created Spaceboost and how it ended up coinciding with two global market trends to become a template for the future of online advertising and machine learning in 2019. Ready?
As so many SaaS companies, Spaceboost began life naturally, in response to a specific problem that we had as advertisers. While working in an Online Travel Agency, Pablo Callejón (Spaceboost CMO and founder) and I created a simple tool to make our day-to-day working lives more efficient. Specifically, we were frustrated with the limitations of manually filling in, updating, and attempting to apply intelligent modifications to the advertising investment we were managing… all on Excel. That frustration would lead us to stumble into the future of online advertising.
Machine Learning in Advertising
To save ourselves from a slow and painstaking circle of manual trial-and-error, and basing our theory on Philip Kotler’s four P’s of marketing (Product, Price, Placement, Promotion), we created a (very simple) algorithm: a bidding optimizer that would allow us to stabilize the Price (ROI) of our marketing investment regardless of the amount of Products and Placements (keywords) that we were working with.
We didn’t know it at the time, but with that algorithm we already had our Minimum Viable Product. Later on, working at a Barcelona-based airline, we were able to process this at scale thanks to the help of Spaceboost’s third founder, Albert Morcillo, now our CIO.
A few years down the line, here we are, scaling Spaceboost internationally. Since we created that first algorithm, two global market trends have converged to create the perfect storm of machine learning, customer behaviour, and automation technology.
In this article I’m going to explain how this happened, and lay out the role I predict AI ad automation technology has to play in 2019: democratising business and ushering in an era of more efficient data-led advertising.
Trend #1 – How Search Changed the Game
Everyone knows that Search is important. So much so that it has become a truism to comment on its central role in the way the world works. However, the effect the rise of Search (and I include Shopping in this) is having on online business – from ecommerce to lead generation to travel to agencies – is so transformative that it can’t be overstated just how comprehensively it will shape the way we work.
The customer journey as we understood it is no longer recognizable. From research to purchase, Search has empowered the user as never before, and consumers are no longer beholden to the large brands they would traditionally have turned to in high value moments. Customer loyalty has been usurped by usability and convenience: 9 times out of 10, the business with the ads that can deliver the most finely attuned response to the customer’s search query will be the business that gets the sale.
What does this mean for business? Before Search became the definitive tool to find information, it was the biggest companies with the biggest ad budgets who would monopolise their verticals by the sheer force of their scattergun ad buying power. Now, no matter your size, if you are at the top of the SERP with the right message at the right moment, you win.
Machine Learning + Online advertising = The democratization of business
So how to get there? Pay Per Click Search advertising is the future of online advertising. Not only is it the future, it is the present. For me it is the inevitable big winner from the development of Search and we are already seeing the impact of this trend.*
Why? PPC advertising breaks the competition down into millions of winner-takes-all micro-markets where relevance is king**. Search by search, click by click, the advertisers that are able to most effectively respond to a searcher’s needs can outperform brands with less effectively created and delivered ads. The beauty of PPC is that it is not necessarily expensive to outperform another advertiser for a high-value click. More than paying top dollar, it is about timing, preparation, and organization.
The key advantage of working with PPC: it’s not a black box. It is measurable, it is efficient, it can be optimized, and – most importantly – it means you can identify the business results that your advertising budget is achieving.
The only block in the road? The problem we came up against when we made that algorithm: the limits of manually creating, updating, and carrying out ad campaigns. Which leads us to Trend #2…
*It seems to me that Google (and Bing) will increasingly prioritize SEM over SEO results. It is something that we are already seeing in extremely competitive verticals like flights and hotels. The exact way in which this plays out will depend on other factors that only Google can know, but which will vary depending on the particular market, device, time of day, search query, etc…
** The latest survey data shows that, especially on keywords with high commercial intent, searchers don’t care about clicking on a sponsored search result if it is better catered to their needs than an organic search result.
Trend #2: Machine Learning Hits the Mainstream
As a point of reference for the state of digital advertising, you can do worse than look at what Google is doing. And almost everything Google added to its advertising program last year was Machine Learning-driven. Here’s a quick look at some of the most recent features by way of illustration:
- Smart Campaigns
- Local Campaigns
- Responsive Search Ads
- Smart Bidding Strategies
- Ads Added by Adwords
- The new, AI-powered interpretation of ‘Exact Match’
- I could go on…
All of these changes are based on integrating Machine Learning into the process of advertising with Google Ads. Google have even announced this year that they are going to begin optimizing certain Google Ads campaigns behind the scenes – unless you opt out. This AI-integrating trend will only increase in 2019, from insights and reactions to voice search keyword trends to increased testing and optimization recommendations for your text ad creatives. Automation will soon be the new normal.
Online Advertising in 2019: Machine Learning for All
Why? Because Machine Learning and PPC advertising is a great combination. The key to a well-functioning PPC campaign is that the ads are the most specific, up to date, and relevant for the keyword that triggers them. But as any PPC account manager will tell you, it is hard work keeping on top of a campaign of any size. To name a few variables, to be successful you need to live and breathe:
- Which keywords you are spending your budget on
- Which keywords are performing best
- Which keywords you need to pause (immediately)
- Which negative keywords you need to add (immediately)
- Which audiences are performing and which to ignore
- Which copies are most effective
- Which locations are performing best
- Which devices to prioritize
- Which channels to prioritize
- Landing pages are highly coordinated with each ad
- And much more…
Not to mention updating all your creatives whenever you update your product catalogue, implementing data-based changes to bidding strategies, reacting to and preparing for high-value seasonal moments, and fire-fighting when something goes wrong.
It can be done, but it’s a lot of work, and above all it is difficult to identify and react to actionable insights that will improve your performance. All of which leads us to the next section…
The role of tech partners (like Spaceboost): prioritize your own ROI
Google is interested in increasing Machine Learning in your Google Ads campaigns because more people dedicating more money on Google Ads means more clicks and impressions and more revenue for Google, as well as better quality results for its users. Simple, right? But there is an essential distinction to be made here.
For all Google, Bing, and Amazon’s power to connect your brand with exactly the people you want to reach, their business models are set up to prioritize revenue for them – that is, clicks and impressions on their channel. Understandable. But this model does not necessarily translate to increasing the most important KPI for an advertiser: the ROI (return on investment) of your marketing budget.1
Spaceboost’s value is based on two simple tasks (that humans can’t do alone):
- Automate the creation of comprehensive, ultra-segmented ad campaigns that cover an extremely long tail of customers. (It makes better performing PPC campaigns).
- Use machine learning to drive visibility for the ads which drive results for your business, not just clicks, impressions, and revenue for the channel you advertise on. (It prioritises results for the advertiser).
Simply put, tools like Spaceboost are the final piece of the puzzle. With our Adtech platform you can take your business information – from product information to landing page urls, to how much you want to invest – and convert it all into a well-oiled, highly valuable, and extremely quick-to-learn integrated solution that works to make ROI the focus of your campaigns.
That first algorithm we made? That’s now The Spaceboost Optimizer – a series of (much more developed) algorithms that carry out bidding adjustments in real time to optimize performance within the budget that the advertiser defines.
This is how we empower marketers of all sizes to leverage the power of Search to drive results for their business, and more effectively manage their time and resources to scale intelligently.
Conclusion: Automation-empowered Advertising
Automation tools are not going to revolutionize your business on their own, they have to be used and guided intelligently by a human marketer. But what that marketer can do is significantly more effective and valuable than what that marketer would be able to achieve manually.
I prefer to think of it as automation-empowered marketing. This is the future: it means better ads for customers, higher quality service for users of Google, Bing, and Amazon, and more efficient and profitable data-led marketing investment for advertisers.
Whether or not your business is ready – and as ever some traditional advertisers (laggards) will fight this shift every step of the way – the future is automated. In fact, the change has already happened, the global trend of customer behaviour has already cemented it. Now it is up to advertisers to react.