Complexity in search has massively increased over the last five years. Google partly simplified this with Enhanced campaigns in 2013, but accounts have continued to swell despite this, mainly because marketers have found new ways to improve relevancy and personalisation through scale.
The level of this complexity is simply not sustainable and not the way the industry is heading. So what are the ways in which we expect to see large scale campaigns change in the coming years?
Simpler campaign structures
Less complex campaign structures were born out of Google’s enhanced campaign update in 2013, but new innovations and segmentation has led to this swelling once again. Savvy marketers have strived to make the most out of segmentation for demographics, geography, time of day and, in the process, have created duplicate campaigns. This duplication goes against best practice - creating more work and complexity, but is currently the only way to ensure specific copy for differentiated segments.
Just as Google created mobile preferred copy to solve the same issue when segmenting devices, there are currently betas in place to test copy for RLSA and Geographic region which is a trajectory that I expect the search engines to head over the next few years. Provision of specific copy and ad customisers where possible to avoid duplication, will mean that structures can stay at the same scale in terms of campaigns and keywords, whilst still benefitting from bespoke copy.
Feeds are no longer just for shops
Integration of paid search accounts with inventory management systems revolutionised the way large scale retail was managed and optimised. By allowing retailers to automate any changes made on site such as price, product or stock levels, and push them directly to campaigns, a greater level of accuracy and performance was achieved with a significant time saving benefit. These feeds primarily drove the shopping results, but in time they were expanded to be used for standard search bidding too. The ability to push these changes in to placeholders, then to ad customisers in search copy and keywords meant that large structured campaigns could be set up and managed with the same input required for shopping, generating similar results.
This opportunity was taken up by advertisers beyond the retail space with auto, travel and finance - all using feeds to ensure relevancy at scale. Success in these areas means that the proliferation of management by feed in non-retail environments will only pick up steam in the coming years.
Machine learning to take control
The search engines are making a major play in the AI arena, using machine learning to optimise the performance of everything from YouTube to beating world champions at popular board games. Application of the same technology to paid search means automated optimisation can be scaled to take control of many of the manual factors that make management of large scale campaigns time consuming, such as time of day, device and RLSA adjustment. These data points need to be configured in the first instance, but once they are, AI is able to take over management to save time and improve performance. There will be, of course, the need to manually override these decisions for strategic purposes, but by the very nature of machine learning, performance will improve as used.
This naturally plays in to the field of automated bidding where only handing over control of bidding decisions to an algorithm is now required to gain a competitive advantage in the auction. Intra-day bidding and a large element of AI-based automation will increasingly be able to make accurate allowance for external factors, adjusting bids according to factors in near-real time. Proper goal setting and human intervention will still be a must, but the heavy lifting will fall more and more on a machine rather than an exec, allowing experts to spend more time on strategic direction.
A keyword-less future?
With very little fanfare, Google introduced a product called Dynamic Search Ads (DSA) back in 2011. DSA effectively automates the process of keyword selection, ad copy creation and bidding by scraping, and uses information on the advertisers’ site. Many advertisers dismissed the product at initial launch as, by its very nature, it took control of management away from search teams. Due to this, it has been underused at the larger end of paid search campaigns up till now, but Google has been working away to improve its intelligence and performance.
At the recent performance summit, Google hinted towards its future direction where products such as DSA will be at the heart, bringing together all of the features put forward in this article. This future will mean simple campaign structures created and updated through feeds and onsite data. Optimisation of such accounts will take place through setting appropriate business objectives and goals and allowing the AI and bidding platforms to make appropriate decisions to achieve them, whilst the input from users is to steer strategy, creative and marketing direction rather than spending time updating bids, copy and budgets.