AI can assist marketers in many ways, from qualifying raw leads to predicting customer behaviour. Farhad Divecha, AccuraCast, UK Managing Director looks at practical PPC campaign considerations that an AI can assist with.
What is Artificial Intelligence? Put simply, AI – artificial intelligence – is the theory and development of computer systems that can perform tasks that would normally be performed by a human. These tasks can range from visual perception, language translation, to decision-making processes. Many consumers are already familiar with AI, through products such as self-driving cars, voice recognition software, or song suggestions on your playlist. AI is also extensively implemented throughout business operations, as corporations hope to become more efficient and to make more informed strategic decisions.
From a marketing perspective, AI can assist companies by spotting missed opportunities within a campaign: it can help qualify raw leads intelligently before being passed on to the sales department; can better predict customer behaviour; can ultimately provide a more seamless and personalised user journey towards converting.
While the scope for AI in digital marketing is immense, this article focuses on five practical PPC campaign considerations that an AI can assist with.
Attribution: which channels do my most profitable visitors come from?
Why do I need to know this? Marketing teams need insights to make smart decisions. if they want to invest in the right areas and scale their campaigns effectively. “Last-click” channels don’t tell the full customer story. Brands need to know the specifics – like the sources that may be contribute to income from earlier in the funnel. Perhaps it was from that blog they were featured on, was it that search campaign they spent thousands on? This helps brands decide where to allocate their marketing budget and also gives them a better understanding of how to engage with their consumers.
Where can I find this information? If you’re currently using Google Analytics, you can run an Assisted Conversions report, or look at the Model Comparison Tool under Attribution. Provided that you have enabled E-commerce tracking in your setup Goals. Other analytics packages will have similar information available, it’s best to do a quick search for a breakdown on your specific package.
Why is it important for AI to do this? Programs such as Google Analytics will help up to a certain point, but once you have to do this at scale – for hundreds or even thousands of product lines – then it can become a time-consuming task and impossible to gain any real insight. As a result, most marketers end up running such studies only once or twice a year, if at all, which means you could spend months targeting and bidding at sub-optimal levels.
Qualification: who are my most valuable customers?
Why do I need to know this? Know thy customer is an important maxim for all businesses. But for those interested in generating leads rather than online sales, how can they tell who their most valuable customers will be, after or even before, an initial contact? By being able to better qualify leads before picking up the phone, businesses can allocate more resources to the higher prospects and avoid wasting time on time wasters.
Where can I find this information? A majority of the time you would have to rely on your own order books and some intuition, as regular analytics will only ever provide quantitative data.
Why is it important for AI to do this? An AI can combine your first party data with several external sources to build a more holistic picture of any new prospect. This historical and external data can then be used to give a new prospect a lead score before getting qualified by a human, thereby allowing you to build a better picture of how valuable they could potentially be and how they might be better sold to.
Seasonality: when are my peak sales periods?
Why do I need to know this? As a business, you will likely know the peak periods that bring you the most revenue; the periods that need more of a push. And you want to target and plug all gaps to ensure you’re picked over your competitors! Once again, this helps brands allocate their budget accordingly to gain the most effective results and drive sales.
Where can I find this information? You can find this information in your order history in the backend of your analytics package. If you’re working on a leads-based business or offline business, then you need to find out when the customer journey began and when they made the purchase as they might be different timings from online.
Why is it important for AI to do this? If you have a small product range, then you shouldn’t need an AI, but the difficulty starts when negotiating a complex range of products or services. If you were looking for more detailed information such as the time of day, or day of the week when people buy and research your products, you would require assistance from AI regardless of your product line size. Also, if you’re working across a wide range of products, an AI can complete this task in a fast and effective way purely due to the amount of information that needs analysing.
Granularity: should my bidding strategies change by device, location and time?
Why do I need to know this? Understanding optimal customer behaviour, based on device, location, and time of research and purchase will help define your bid strategy for each of these facets. The budget assigned to various devices should change, based on how your consumers purchase. The more insight that is gained the better a brand can target its audience at an optimised level. Businesses should not limit their potential by sticking to a single, simple bidding strategy.
Where can I find this information? Through your analytics and the order / lead data itself.
Why is it important for AI to do this? You will be able to individually identify what times are best or ineffective for conversions/lead generation. You will be able to figure out when customers are more likely to buy versus when they’re idly browsing. When you start having to do this at set scale across several Ad Groups and across a number of product types then it becomes a humanly impossible task to do regularly and that’s when an AI steps in. Tracking cross-device conversions and cross-platform conversions often adds insurmountable complexity for humans, especially when offline conversions and CRM data need to be pulled in. The amount of work is practically not feasible for any human to do regularly, necessitating the use of an AI-based system that can scale up much more rapidly.
Value: which keywords do I absolutely need to bid on and which keywords can I stop wasting money on?
Why do I need to know this? Understanding which keywords to bid on can be tough – particularly the short tail ones where it’s difficult to identify the intent behind the search query, i.e. are the users just looking for information? Are they a competitor? Do they want to buy? Bidding on the correct keywords alone is never a guarantee for sales. Bidding on the wrong keywords, though can result in advertisers spending unnecessary amounts of money and getting less than optimal traffic/leads.
Where can I find this information? Within Google Ads, the Search Terms report will give you a better sense of what types of searches are bringing the highest value conversions. You can then see which keywords delivered conversions and which keywords took up a lot of clicks and impressions but delivered either no conversions or a disproportionately lower number of conversions. This allows you to very easily identify which keywords you should bid on and which keywords you want to stop wasting money on. If you have the time, you can take this a step further and start analysing other factors associated with each well-performing keyword, such as time of day when searchers on that keyword are more likely to buy. Google Ads allows advertisers to view segmentation reports for such information.
Why is it important for AI to do this? If you want to do this rapidly, at scale, then you need an AI to identify poorly performing keywords and act on them according to the guidelines you have set. A machine can do this in near real-time versus humans who would have to regularly scheduled tasks and come back and revisit this exercise every few weeks or months depending on how often or how big the scale of the campaigns is.
AI can be an extremely useful tool for businesses, both in B2B and B2C sections. Each company is unique, though, and must examine their current situations and see if AI can actually make a positive difference in their campaign – chances are it can!
Farhad Divecha
UK Managing Director
AccuraCast