Stay marketing and tech-savvy. Get the latest in martech - subscribe to MarTech Today.
Should PPC agencies be scared of automation?
Paid search professionals might be nervous about losing their jobs to automation, but columnist Anna Shirley makes the case that PPC automation may actually benefit them.
I’m sure many of us working in PPC (or across the wider digital spectrum) have been asked at least a handful of times in recent months, “Should we be worried about automation?”
I’ve heard it from clients, and even prospective job candidates. And I can see why the question is commonly asked; it is a very interesting subject with a lot of possible responses, ranging from, “Yes, we’re all doomed. Let’s head to the hills!” to “No way, machines could never do what I do!”
In this article, I want to look at some of the main arguments against fearing the rise of automation and machine learning to try to dispel some of the confusion and uncertainty around this topic.
I’ve split this into three key areas: Campaign structure, bidding and scripting.
Automation will continue to change the way we manage campaign, ad group and even ad-level structure. Long gone are the days of unmanageable-sized accounts with incredibly granular RLSA vs. non-RLSA structures and thousands of ad groups for individual keywords and their match types.
Google recently announced the ability to use IF functions within ad copy, allowing us to alter the messaging for users depending on the device used or which audience list they fall within. By having ads that dynamically change for each individual searcher, we can ensure a tailored and relevant message appears.
For example, you could show a particular offer code to anyone who has visited your site before but not purchased, while you could show brand-new visitors different messaging that showcases your main USP. For users searching on a mobile device, perhaps a message highlighting that they can purchase in “one tap” will help to engage those looking for a quick buy. You can even use your Customer Match lists here, e.g., showing VIP offers to those with a loyalty card.
With the launch of adaptive Shopping campaigns in DoubleClick Search, you no longer need to monitor performance of each individual product ID, as your campaigns can be sub-divided automatically depending on performance goals (set up in a conversion goal or within a bid strategy). By automating this process, you can allow algorithmic bid strategies to adjust bids based on revenue or conversion metrics in a much more fluid way than managing this manually. You can therefore:
- increase revenue by bidding higher on top-performing products.
- decrease the amount of money wasted on products that perform poorly.
Google recently released an adaptive Geo option within DoubleClick, too, following the same principle as adaptive Shopping campaigns. Once you’ve selected a country or state to target within your campaign settings, you can let DoubleClick Search add in more tightly focused location targets (but remaining within your overall country or state target) to apply automatic performance-based bid adjustments.
For example, if you’ve targeted London within your campaign settings, DoubleClick Search may target particular boroughs like Camden and Westminster with more aggressive bid adjustments if they perform much better than other areas. Areas will be added and removed based on performance, much like what happens within adaptive Shopping campaigns, allowing for a flexible structure that is always looking at recent and relevant data.
If you haven’t tested any algorithmic bid strategies and are still using manual bidding, I would get testing now! Automated bid strategies through AdWords calculate a new bid for each and every auction, based on your performance goals (clicks, impressions or conversions). AdWords also factors in signals that we don’t have access to within the interface or in third-party platforms (e.g., search partner data).
Without automation, it would be impossible for a human to both have access to this data and to make enough changes to replicate the changes made via automation. The machine learning aspect means that over time, the bid strategy performance should continue to improve and update, depending on targets set. By combining bidding strategies with a flexible account structure (using the adaptive campaigns and personalized ads discussed above), you should have the perfect combination of strategy, data and tech at your disposal.
AdWords Scripts are an easy way to use the AdWords API to automate certain processes within accounts. They’re an extremely powerful account management tool and can be used for a wide range of processes, most importantly:
- building and customizing automated reports.
- managing campaigns, ad groups and keywords.
The possibilities here are (almost) endless. At Merkle | Periscopix, we use scripts across all accounts, from basic budget pacing reports all the way up to automated account changes, competitor analysis and regular reporting dashboards. If you find yourself doing the same manual tasks over and over again, take the time to do some online research and see if there is a possible scripted alternative, saving you hours of time and allowing you to focus on higher-impact changes.
For one client within the recruitment sector, each team member spent two hours each week reporting on performance for a number of campaign types. Through AdWords Scripts, we were able to pull the data automatically into Google Sheets each morning so the client could dip in and out when they fancied, freeing up the team’s time to spend making meaningful changes in the account. The client was happy, as they could access up-to-date data whenever they wanted, and the team felt that they were having more of an impact through optimization and testing.
There shouldn’t be anything for us to fear when it comes to PPC and automation. The very concept of automation and machine learning often conjures up all sorts of images of robots replacing humans, but when it comes to the world of PPC, automation leads to more effective and efficient processes, benefiting the advertisers themselves, as well as those of us working in the industry.
Over time, I expect that account management time spent on a lot of manual tasks will decrease with the ever-growing opportunities to automate day-to-day processes like bid adjustments and personalized ad copy. However, more free time shouldn’t mean we have more time to watch cat videos on YouTube but should enable those of us working in PPC to focus on strategic direction.
One thing I do envisage is the need for those managing PPC campaigns or those involved in driving strategy forward to improve their own awareness and knowledge around the subject. If you look at the dramatic shift in PPC from a keyword-based model of five to 10 years ago toward an audience-based approach, we’re now seeing the the shift toward automation as a process that can tie these together. For these reasons, I am confident about the future of machine learning and automation for those working in agencies. It opens up many more opportunities for those working in the industry through the growth of more specialized job roles and the need for an increased knowledge in this area from all, all the way from Account Director to Analyst level. This will obviously benefit our clients, too, with agencies being able to spend more time on the strategy and future growth of their accounts.
Therefore, the only thing of which PPC agencies should be fearful is slipping behind the curve and missing out on the huge opportunities in this area for the development and retention of their staff and their clients.
Opinions expressed in this article are those of the guest author and not necessarily MarTech Today. Staff authors are listed here.