February 28, 2022

If I had £1 for every time we said “keyword” in a day…well I’d have enough to show your ad to your entire target audience, in the top position of Google Search results.

PPC came of age around the keyword because of its unique targeting capabilities. You can reach a specific individual, asking a specific question, at exactly the time they want to learn the answer. If timing is everything, the keyword is the mic drop of marketing.

What’s almost as important? The amount we’re willing to pay when someone clicks the ad. 

Attached to every keyword is a monetary bid. So, for every keyword, there’s a handful of other advertisers wanting to show their ads. We enter an auction, and you guessed it – highest bidder wins. To get your ad in prime positions, we must pay top dollar. 

Google ads (or what was Google Adwords) initially was particularly great for small businesses because it meant that the small businesses and big guys had equal advertising opportunities: competitive prices, targeted ads ranked based on quality, relevance, and you could decide how much/how little to spend. Unlike paying over the odds for a giant billboard in town for people that may/may not be interested and just so much glanced at. 

The keyword has evolved over time. From pure auction-based exact match bidding, we’re now able to use machine learning and deep insights to understand behaviours and consumers in the modern world. 

Google knows more about us than we think. We work with Google to get your ad in front of the right person, at exactly the right time. 

Let’s say person A is in the market for a mattress and has spent weeks investigating numerous brands online. Researched local mattress shops, and clicked on specific shopping ads to view individual products that they’re interested in. Google has all this data. And the next time they go to Google ready to purchase, they simply type “mattress”, and Google knows exactly which shopping ads through dynamic remarketing or local stores that they should show. 

Person B on the other hand is just beginning their shopping journey, so when they type “mattress” into Google, Google changes the results to a higher funnel more informational SERP (Search Engine Results Page), same keyword but dramatically different results.

Google will also use your location to determine a context and deliver Google Ads results. For example, if you’re searching for car breakdown services when you’re stuck on the side of the M1 50 miles from your house (don’t feel bad, we’ve all been there) then you’re likely to be shown ads for emergency breakdown services. Whereas if you’re searching from your usual home IP address, it’s likely that you’ll be shown ads for breakdown cover (for when your ¼ tank of petrol doesn’t quite see you to the end of the journey). 

There’s so many different targeting features available. Keywords are great, but they’re just a lever. The Google Ads interface has lots of other buttons, and little twisty knobs, and dials that we can use to try and figure out how to get the right person using the right language.

Here’s how the humble keyword has evolved:

Keywords started as exact match.

Back to the mattress keyword analogy. We’d bid on [mattress], you’d search “mattress” and our ad was shown. If you typed “best mattress”, no ad. Or “mstttres”, for those with chunky thumbs, no ad. Only the exact keyword and relevant search would trigger the keyword. So, keyword research would be done on misspellings, Google auto complete and commonly used terms. Yes, it’s as boring as it sounds. 

Then came ‘close variant’. This is where Google automatically discovers the close variants around the keyword in question. If someone is after the best, or local, or is crap at spelling – then Google will trigger the ad. 

Of course, with this comes the irrelevant searches. Like “mattress image”. Cue the negative keyword list. Robust negative keyword lists are built to exclude ads from triggering on irrelevant searches. 

What next? Intent. 

Here is where we must trust Google. Google’s claims that they’re able to better tell that intent from a variety of data points unavailable to the poor, helpless human advertisers – like the recent search history of that consumer. Because of that, Google also allow match types to appear for words where Google believes the intent mirrors the user intent, not necessarily the keyword intent. 

To continue with our mattress example, if you bid on the exact match keyword, “best mattress for back sleepers.” Someone searching “best mattress” could be shown your ad for that keyword, or even just mattress. Since Google may be aware from their previous browsing history or searches of that consumer, or really all those other data points that they have, that that person is in fact a back sleeper and that that query and that keyword would match.

It’s really clever. We now also trust Google to use this information to optimise and adjust bids. Using different data and attribution models to make proactive bidding to suit audience intent, the audience and your business objectives. 

It’s amazing to see how the keyword has evolved over time. From being a manual process, to using Google’s intel to really maximise results. I’m excited to see what next…with the new data driven attribution models, measurement tools like GA4 and audience segments, I believe that we are headed for even more automation – particularly with third party cookies being scrapped and Google opting for a group profiling system instead.