Even though this was originally posted almost 2 years ago, in honor of some friends recently being informed their Google Ads campaigns are being switched to use the Data Driven Attribution (DDA) model, yet another push into the AI/ML approach to PPC, I thought I’d dust this one off for a Throwback Thursday.
So, let’s not go back to the beginning of (PPC) time – but it is quite fascinating, in case you want a little trip down Wikipedia memory lane – https://en.wikipedia.org/wiki/Yahoo!_Native
Hindsight being 20/20, you have to have known that automation/AI/ML was always going to be the way of paid search – especially for Google. Think about it – in the beginning, manually creating keyword lists, organizing keywords into campaigns and ad groups and writing ads – although a lot of work, was exciting, okay – maybe not exciting – but rewarding as everything was so trackable. And, if you did it right, it was profitable. So justifiable. And ad budgets EXPLODED.
And technology started to creep in. In 2006, AdWords Editor helped marketers upload massive numbers of campaigns with just a few clicks.
Around this time automated bid managers started to pop up. I became a product manager for several such products (one based on last click analytics and another based on multi touch attribution). Back then, it was more just access to the API to automate your bids based on very simple or incredibly complex bid rules. Good times.
So as automation entered the picture, so did the mentality of ‘set it and forget it’. Having lived in the product world and understanding the negative consequences of the ‘set it and forget it’ mentality (maybe click fraud, maybe just bad ads), I went back to the agency world. I advocated for a human paying attention to campaigns.
But then, jump to October, 2011 – Dynamic Search Ads (DSA). Google starts using AI to make better ad spending decisions. Using its organic crawler, Google spiders your site to find more relevant keywords, categories and landing pages combinations.
Another quantum leap to 2018 – Responsive ads. Google Ads mixes and matches headlines and description lines. So now you are on the hook for up to 15 headlines and 4 descriptions for a single responsive search ad. But Google would ultimately figure out and serve up the best combinations of the headlines and descriptions.
Cynical me will not share the results of some such combos. But now, fast forward to today – or last week. Google has introduced three more ‘enhancements’ (as it does each year right before the retail holiday ad spending season):
Conversion goals/categories – allows Google to orient spend towards account or campaign based categories – moving from the granular keyword-based approach of days way long gone by – to general, broad based goals. Although this is good because it increases inventory (spend), it does eliminate much of the precision and target ad spend that used to be the appeal of paid search.
Tag Assistant – Ugh. Sorry. As someone who spends a lot of time in GTM – I feel that the marketers who write the ads and craft the strategy should not have to be bogged down by the minutiae of tag implementation. There are many nuances and details owned by the implementation specialists that should not be factored in to campaign management.
Explanations – To me, much like ‘Insights’ and other AI built into GA4. Machine learning reasons and excuses for why ad performance is faltering.
Sorry if this seems to be a bit negative. But as someone who has lived through the ‘glory days’ of paid search – from creating and crafting well thought out ads and campaigns – to someone who has advocated against the ‘set it and forget it’ mentality – to someone who is feeling like it is about to be totally out of the hands of the creatives and marketers and all left to the machines and AI that power Google, I am just a little sad.