How do we use PPC to target humans?
Most audience personas begin and end with three things. Age, gender and location. Maybe some interests will be thrown in, but I think it’s important to dig so much deeper than this because plenty of us will be the same age and gender and have lots of fascinating differences between us.
What else do your customers like? How do they like to interact with their favourite brands? Why do they buy from you and not your competitors? How do they behave on your website? And importantly, how do the people who don’t buy behave on your website?
I like to think of it like this; if you went to a pub with your best mate, I’d bet you’d know what drink to order them. That’s how well you need to know your target audience.
The second point on how to target humans is by using the tools you have available. My best recommendation for this is the cheapest tool out there and it’s Google Analytics. Don’t be scared of going overboard with goals to track user behaviour. They’re limited to 20 per view so people tend to stick to just a few core ones, but you can have 25 views per property so my recommendation would be to make loads and really understand how the right users behave on your site and also how the bad ones do.
As well as GA you also need to make sure you’re looking after your CRM. Working agency side I was never given access to a clients CRM, and a lot of clients I speak to hate their own CRM platform for various reasons – so it’s no wonder they’re neglected. But it still doesn’t make any sense. Your customer data powers activity across search, social, email, display and not to mention all the lookalike audiences these can power as well. If you’re putting garbage data into these things you are going to get garbage data out.
The reason you know what drink to order your friends is because of all the data you’ve saved in terms of memories. You can’t do this for all of your customers but your CRM can.
Next – what do we do with all of this data?
First and foremost I would say you could probably be utilising RLSA better. Of course, they’re great for bid adjustments on campaigns but they get really powerful when you can use them to communicate with a person differently depending on where they are in their purchase journey with you.
There’s nothing worse than being a loyal customer and searching to see a new customer discount. RLSA is how you can avoid this. The same person can be searching the exact same keyword but depending on whether they’re a new customer, basket abandoner or loyal customer, they’ll react better to different messaging.
If you don’t want to create multiple campaigns for different lists then you could also try the IF function ads, which I think are one of the most underused features in Google Ads. They essentially work to an ‘if this – then that’ model and can show a different message depending on an audience list or device. They’re a cool way to get your ads speaking to the human behind the keyword a bit better.
When talking about audience lists we also have to mention exclusions. If you’re a subscription brand or online service do you really need to be bidding on brand when your customers search just to log in? I’d argue probably not, but if you do, why not use that ad space to cross-sell or just remind them why they are a customer.
What if you don’t have any or enough audience data of your own yet? No problem. Google, Bing and Facebook have plenty of their own for you to use. Age and gender are at the top of most audience personas but it’s another one I don’t think is used enough to send bespoke ad messaging.
The type of message that will grab the attention of a 17 year old searching for their first car vs what will be of importance to their mum or dad searching will be very different. With age data you can offer the teenager a message about the colours and music system and the parents a message about the safety features and warranty.
Google in-market and affinity audiences can also be pretty good these days. Use GA to understand the most important ones for your business and set up a campaign pairing them with really broad keywords. We’ve had great results doing this, especially when you have enough data to switch to target CPA or ROAS bidding.
As basic a thing as search query reports are, I think they’re an important element in targeting humans better, if you spend time really trying to understand the intent behind a search. A query may be relevant but does that person really want to see an ad at that point? If so, what kind of ad do they want to see? What stage of the awareness journey do you think they’re at? This will provide a better experience for the user and save you lots of money.
On the topic of money – household income targeting is available in the USA and is a really important demographic option to have. Consider someone searching for a ‘luxury white shirt’. The exact same keyword made by two people on different salaries can mean really different things; one may be expecting to pay £50 and the other £250.
And lastly, where someone makes a search also makes a difference. Are they nearby a store? Are they at an airport? Did you know you can also change mobile targeting to people on WiFi or not? This is a big indicator that someone is out of home or work, which may impact behaviour especially if you are a B2B product.
Three last points to summarise: