Digital Wizardry Episode VIII

The future hasn’t been written yet. This week Google announced a few of its plans for the future of its flagship Search Engine. As we know, whatever Google does to the First Results Page has a big effect on the paid media. These changes are being trialled in the US, but we can expect a worldwide rollout in the next few months, if everything goes okay, but for once - these changes are in doubt.

What stays the same

• Sponsored - in bold text on every ad section, shopping, search and paid location results.

• Shopping/ PMax ads on Intent searches (eg. Buy, Cheap, Near Me searches)

• Search Ads top and bottom of First Page

What’s New

• Generative Content

What’s Generative content?

The issue that Google has been facing in the last decade has been ‘New Searches’ - search text that they have never seen before. It’s easy to show the results for ‘Where can I buy an office chair?’ - but it’s harder to show the results for ‘I need a blue fabric office chair, with a headrest.” Or “Show me office chairs without casters or arms to purchase near me.”

With greater user-friendly search engines come more friendly styled search queries - and thats too much data for Google to store, process and use in their machine learning. So the solution is to use AI (more machine learning) to break down the queries using past queries/results data and generate a unique result at the time, to guide the searcher to their best results.

We’ve seen this a little in the ‘People also asked’ section - Google admitting that the Organic First Page results are unlikely to yield the desired answer, but the big difference with the new system is that the Content for the clarification section will be generated by AI and by Google, not referenced from existing third party websites and sources. So the question is how accurate will that text be?

The issue with AI is still that it can only respond with the information it has been given, and that information will always have its own bias, style of language and nuances that are easily lost by the algorithm in the ocean of comparative data. If the information used to create the unique content has one or two wrong facts or uses data out of context, then the result will be misleading, wrong and have the opposite effect of what Google is trying to achieve. Niche cases and uses of language will be lost in the query and valuable information overlooked by the AI judging a peer-reviewed article and an extremist blog as equally valid.

While Google has said that the generative content will be transparent, to show the sources, it will not be as prominent as the result - and the more vulnerable members of society or even people in a hurry will not be able to check the quality of the sources.

So what does this mean for marketing and advertising? While we know that Search Ads will still feature prominently on the first page of results, the searchers will be encouraged to use longer tail search strings. This means our keywords will need to be more flexible than ever, pushing Broad Match, Phrase Match and Dynamic, over Exact Match, leading to more PMax style targeting, where AI will try and control both the searchers intent and the advertisers intent and match them up, like an intuitive dating app.

The hardest hit will be the SME businesses. Most of these advertisers are in a niche marketplace, sometimes using the same word or abbreviation to mean completely different things, something that will be lost on the algorithm. Many of these really on the human touch, from both the searcher and the crafted ad copy to get the right customers to the website, and ease the similar, but irrelevant searches away from clicking through.

What will this future look like? Try using the Google Bard experiment to ask about your business. https://bard.google.com (you’ll need to sign up) and ask it about your business name, your website, where you are based. The AI generated content will tell you what your potential customers may soon be seeing.

If the results worry you, then you aren’t alone. I’ll be going through solutions and what you can do about it in a future blog.