Paid Search

PPC Automation: Who's Flying The Plane?

What on earth do planes have to do with Paid Search? By the time we've landed (the end of this post) I hope it will all make more sense.

The benefits of PPC automation have long been debated. Can you really trust a machine? In this post, I'll discuss two methods that we use to automate processes in our campaigns and demonstrate their merits - scripts, and automated rules. Prepare for take-off.

Scripts

Scripts are fundamentally a document written in JavaScript that can impact a PPC account at scale within a few short moments, saving you time that you can invest elsewhere.

There are several scripts resources available - search for what you need, grab the one you want, and away you go. Alternatively, you could learn JavaScript and write your own.

So how does scripting help? I'll share with you an example. Say you’re a sunglasses retailer, and you are looking for new ways to drive PPC efficiencies over your key seasonal trading period.

Using a script, we recently aligned regional weather patterns to Search interest levels to try and ensure our advertiser was at the forefront of customer’s minds. Typically, this involved boosting bids at scale for high converting keywords to gain a positional advantage on the competition. A higher position equals a higher CTR, and coupled with a strong retail proposition, we had the formula for great performance!

We took this to the next level by looking at historical data in the account and geo-segmenting the top performing cities and targeting those specifically. (Side note for anyone interested in taking this further: you can also dynamically insert the user’s location into your ads for more tailored messaging and a likely higher CTR).

How did it work? Well, when the temperature fell or rose based on criteria we predefined, bids were either boosted or reduced. We always expect a strong correlation between temperature and sales during this period but the script meant we could react better than the competition and really capitalise, the results were great.

Temperature

Not only did we save money that could have potentially been wasted when the temperature fell (who wears sunglasses in the rain anyway?), we managed to appropriate engine spend to customers at exactly the right time, delivering a strong return on investment. 

Without scripts, making large scale changes to a big account based on users’ location and the weather for that day, one-by-one would be a huge undertaking. The time we saved using the script allowed us to focus on other areas of campaign performance, which maximised the success of the campaign over the fiercely competitive seasonal peak period.

Rules

Generally speaking, rules are specific conditions you set to trigger automated changes in campaigns at scale. The difference versus scripting? You don't need to be able to read or write JavaScript.

You can use rules for a number of outcomes, some of the most popular in my experience involve automatically increasing bids to the first page or even to the top of the first page, to name just two. You are also able to pause/enable activity too.

We often use a rule to affect bidding at keyword level on one of our campaigns, in effect to try and maintain a specific position on the landscape. It’s very simple, you define the position you want, specify how much you're prepared to pay, and set whether you want this to happen daily/hourly/weekly, etc. The engine makes the decision based on data in the account, again over a time period you define.

Why use a rule like this? Because across the account there are lots (and I mean lots) of high performing keywords for a variety of products that perform better in different positions on the landscape. Manually managing this process is, above all else, a drain on time, and conversions can be lost when doing this.

As the pilot of a plane, spending time checking in personally on each and every passenger would mean not so much time keeping the plane in the air! While individual optimisation of keywords is essential, a degree of automation based on intelligently chosen parameters helps to ensure we have the time to focus on the areas that will be most impactful for overall performance.

Rules and automations are great; they can scour large accounts quickly and deliver you outcomes and results based on a set of variables and the data already there. However, if there is a sudden spike in activity due to a breaking news story, or any other unpredicted surge of volume – they can’t react like you can. Make sure you set them up correctly and you build in effective regular processes to monitor, maintain and evolve them for the best results. 


comments powered by Disqus