Google made $175,000,000,000 last year from search advertising. Why is it putting this at risk by making its results dramatically worse?
I spent some of my summer travels reading about the current strategy being demonstrated by Google search, because that’s the kind of exciting guy I am. After all, Google is possibly the most important channel for the marketing success of almost every B2B business. It’s also at the heart of what my own company does. This stuff matters.
The verdict from many of those in the industry isn’t that complementary. Writers such as Cory Doctorow have been trying to make sense of why the search engine has gone all-in on what he describes as “AI slop”, providing answers which often aren’t true but are always confident.
To investigate what he’s complaining about, try searching Google for “what are the advantages of the [made-up, non-existent model number] [type of product]”, such as the “ty5 pneumatic actuator”. You’ll quite likely get an authoritative-sounding response, despite the model not even existing. On investigation, the response will turn out to have very little substance. Indeed, it’s probably rubbish.
What’s worse, in several of the tests I tried, at the end of the Ai answer, which took up much of the results page, there was a link at the end to a general background article that might help. In every case in my limited tests, that article was also rubbish. Agencies are filling up the web with Ai-written vacuousness which seems to appeal more now to Google than articles we’ve spent years refining as good references.
Given that Google made $175,000,000,000 last year from search advertising, why is it surely putting this at risk by making its results dramatically worse? Doctorow has a plausible theory, and I’ll discuss that tomorrow. In the meantime, see the problem for yourself by having a play with the type of search outlined above.