Digital is constantly evolving, so it can be difficult to bring clarity to team leadership, direction and planning for success within PPC. Here’s a short guide to help PPC leaders focus on the workstreams most likely to boost performance and shift the needle, embracing digital change effectively.
Don’t sweat the small stuff!
As we continue on the journey through automation and AI is at the forefront of everything we do, it’s become clearer than ever that we shouldn’t be worrying about the small things that we may have cared about in past years.
As an example, the loss of a full view of search terms has gently nudged us into a world where search queries aren’t at the forefront of our success or diagnostic Performance Indicators, and we aren’t paranoid about what queries we’re matching either, because we trust in the machine.
The insights we now get from Performance Max (Pmax) are driving us to think about how our customer is interacting with our brands, focusing on what signals we can give to the machine to send it in the right direction, again, because we trust in the machine.
Why not think about focusing on the bigger, more impactful things, like:
- Taking stock of business KPIs over a longer period of time (don’t worry about daily performance shifts if you don’t hit your KPIs on a daily basis – machine learning is set up to achieve your KPI over time; worry about the inputs that will deliver performance).
- Leaning into artificial intelligence (AI) by recognising what inputs we can give the machine, and what customer insights drive those inputs. For example, challenging our teams to become creative experts.
- Measuring the success of PPC performance effectively and understanding the role each of your channels has for business performance.
- Lifting your data offering to feed digital strategies (we’ll talk about value-based bidding later).
- Engaging in a test and learn mindset! We’ll only get the best from what AI has to offer us if we test test test.
Of course, there are always detailed insights we need to help us answer the big questions. At Croud, we’re continually finding innovative ways of pulling the data needed to do this into one place, for example, our proprietary tech Whitebox.
Automation isn’t a mantra anymore; it’s a way of life
We’ve all shifted as practitioners from being focussed on how much we can tweak and edit in order to bid for the right person at the right time, to leaning into automation and all of its glory.
Understanding the inputs we can give the machine to achieve the desired outcome, and refining them, should be one of our core focuses. When thinking about how to prioritise our team’s time, and strategic workstreams for the future, using the inputs/outputs model is a good starting point.
Whilst we’re on the topic of automation, we can’t ignore one of our industry’s current hottest topics – AI in Search.
This is nothing new; it’s been around for a long time and has been what’s driving machine learning and our use of automation for a while, through innovations such as Large Language Models. Google recently announced the use of Search Generative Experience (SGE), which enables users to engage in conversational based queries, with AI showing snapshots of information in response to searches – another shift in the way users search and Google reacts.
While Google assures us that paid ads will run alongside SGE and have dedicated slots on the search engine results page, it’s likely that as Google tests the relationship between traditional ads and SGE, we’ll need to adapt to complement SGE too.
We’re adapting to the use of AI tools (and everyone needs to get comfortable with this) by leveraging it for creative builds at scale, categorisations, forecasting, using bidding solutions with AI at its core (Pmax) and more…
Data is pivotal for the future of paid search
Data feeding the machine is nothing new, but we’re in a transitional moment where it’s more important than ever to get it right – where the opportunity in the quality of an advertiser’s data, and in their ability to activate it, has become ‘make or break’.
When planning what the future of your PPC activation looks like, why not consider:
- Leaning into value-based bidding. Advertisers (regardless of vertical) should be thinking about the value (revenue, profit or margin) each conversion and each customer drives. Improving your bottom line KPI is key to success. Then, focus on how that value can be passed back to Google Ads or Search Ads 360 (SA360) to enable return on ad spend bidding and ultimately, the machine finding users who have a value to you. You can go a layer deeper by bidding with new customers in mind here too.
- Whether you’re using your first-party data effectively. The most obvious form here is via Customer Match lists. In a world where privacy is at the forefront of everyone’s mind, and the loss of third-party cookies is on the horizon, Customer Match is as important as ever. Not only does it enable additional strategies and avenues for interacting with your customers (via app, email, display, social), but it enables the bidding algorithms to optimise with this in mind. You could take it to the next level by incorporating your data into custom bidding models (using your data to drive propensity modelling). The industry is also seeing the growing significance of first-party data used in conversion attribution. Hashed personally identifiable information offers an approach to user identity that doesn’t rely on cookies, which are vulnerable to browser technology like Apple’s ITP. Google’s Enhanced Conversions is one way that first-party data can improve the quality of the conversion data that powers smart bidding.
- Thinking privacy-first. Our industry – and society more widely – has woken up to problems with personal data and privacy. Growing attention to privacy has been fundamentally very disruptive. Data protection regulation (e.g. the EU GDPR) and attitudes to privacy have had dramatic implications for the industry’s reliance on user-level tracking. Google Consent Mode is a unique solution that allows Google Ads, SA360 and Google Analytics (GA) to benefit from modelled metrics that use the collection of cookieless pings from websites to restore (most of) the reporting for non-consenting users. In measurement, traditionally we relied on user-level tracking and digital attribution. Instead, we’re now augmenting this tactical data with Causal Inference studies for reporting incrementality (see also below); and we’ve seen innovation from Google in-platform with conversion lift testing.
- Creative is a data point that feeds the machine! This means investing in creatives and robust testing methodologies across channels, bringing the wealth of knowledge your teams have to enable products within PPC such as Pmax and YouTube. Advertisers among the most mature in their Analytics practice will be pre-testing creative: using machine learning solutions to predict its performance even before it goes live.
Testing and measurement
We can feed the machine better, but we can also measure better. The classic ‘How successful is my marketing spend?’ and ‘How successful was that test?’ questions aren’t going away, and are now more of a cross-channel challenge to think about seriously.
As PPC advertisers we’re often faced with proving the value of activity or testing in an ever-changing environment where stability is hard to come by. A few things to consider:
- Aiming for the stars is always good; but starting with Search is a good idea. Aligning your Paid and Organic KPIs helps give you a holistic view of Search performance, giving you a platform for search engine marketing strategies. Having a unified view of how channels interact can identify otherwise missed opportunities.
- Before testing begins, are measurement and agreed success KPIs outlined? Considering how the testing environment might impact tests is key (consider learning periods, media activities happening at the same time etc.)
- Proving value with Causal Inference studies can help eradicate the age old question of whether demand shifts are the cause of positive performance.
- If you aren’t already on the road to GA4, there’s still time! By now if you haven’t opted out, GA will have already built a default GA4 property for you as Universal Analytics was deprecated in July.
- Marketing mix modelling and econometrics should be prioritised if you are to ever truly understand the role each channel plays in the user journey. Sometimes we can neglect the halo effect our activity has across other channels, and vice versa, resulting in losing customers who would have converted, by removing a channel from the mix. A true holistic view can help guide where budget allocation is truly going to drive value.
Summary
As PPC changes, PPC leaders need to adapt too. We’re no longer focussed on the small changes we see day to day because we’ve got trust in the machine to deliver performance. Our constant focus now is how we can improve the data input to the machine, to shape and improve the output. AI has received an influx of attention following the launches of ChatGPT and Bard, and we should embrace it – working smarter, not harder. Finally, we’re thinking more about how we measure the success of changes we’re making, and the role our channel plays for business performance.
