I try not to dwell too much on the ‘good old days’ of precision PPC.
Google Ads is a different animal now, for better and worse… and it’s our job to make the best use of it we can.
But with spookiness in the air, perhaps we might succumb just this once to the ghosts of Google past, and take a moment to reflect on some of those now dead features that once we loved about Paid Search.
This valued metric was laid to rest in October 2019, after years of loyal service.
Average position was excellent for one particular purpose: indicating how much scope there was to ‘push a keyword harder’ by its raising bid.
(I would often filter for keywords with low avg. position and low CPA, to find ‘low hanging fruit’, where a bid hike would almost certainly boost results.)
That function is now best served by Top and Absolute Top Impression Share Lost to Rank…. But not as simply, and not as precisely.
Shortly before Average Position was removed, Marin Software produced this (at-the-time very useful) chart, attempting to map search impression share metrics onto the old average positions.
(chart displayed with thanks to Marin Software)
…And since Impression Share metrics were added as inputs for creating custom columns last year, my dreams of resurrecting ‘Estimated Average Position’ as a metric in the interface have been revived!
Real Ad Rotation
This went away more quietly than most…
Originally we could opt to rotate the ads in a given ad group evenly. A good option among active campaign managers who knew what they were doing and wanted to make their own judgements.
In 2012 – ‘rotate evenly’ became a time-limited feature. Ads would rotate for a while, and then move into auto-optimisation.
Then in September 2017 Google announced that our ad rotation options would be reduced to two – not including ‘rotate evenly’.
But long before that ‘simplification’ announcement, the ‘even rotation’ of ads was not genuinely even… Not even close to even… just a bit more even than the alternatives.
Still, it isn’t worth spilling much ink on this change, given the rotation cauldron of RSAs, and the fate awaiting their predecessors…
Expanded Text Ads
As of the end of June 2022, Expanded Text Ads are no more (no more new ones anyway… and our existing ETAs were turned to stone…) leaving Responsive Search Ads as the only option within standard search campaigns.
Apart from this being yet another example of reduced advertiser control, one of the reasons why PPCers were (or are) unhappy about this is that RSAs still tend to underperform ETAs in practice. (I go through some compelling evidence for this in a recent video).
Search Term Data
Serious old timers like me will remember when GA showed organic search terms (an almost unthinkably valuable data set these days…)
When that data all collapsed into ‘(not provided)’, the commonly-muttered explanation was ‘Google wants to encourage us to use AdWords in order to find out what our users are really searching for’.
The reason given by Google, of course, was privacy.
Similarly when Google announced a huge cut in visible search term data back in September 2020, the reason given was privacy.
(the response to that reason was largely sceptical…)
And yes, we’ve had ambiguous messages about partial resurrection of lost search terms, including official word in September 2021 that Search Term data had increased, backdated to February of the same year.
(the increase was limited in practice, and largely involved zero-click impressions. Not the most useful)
In any case, the long-term trend has been in one direction…
Match Types with healthy boundaries
The precision of match types has been chipped away in increments over the years. In 2010, Exact match opened up to ‘close variants’…
Seemed innocuous enough. But you know how the story goes…
It’s a long-term trend, and one that has accelerated in recent years.
And while Google’s advances in natural language recognition – and general ‘relevance-detection’ – have given Broad Match a much stronger case than it used to have (not to mention Broad Match’s exclusive use of certain auction signals)
…nonetheless, as a tool for advertisers to control the search term-keyword relationship, match types have lost much of their value.
Google’s unspoken response is that that control isn’t what we should want… it isn’t what’s best for us… I won’t wade in on that just now.
Modified Broad Match (BMM)
Then in 2021, the big one.
Modified Broad Match had a truly useful feature that has no replacement: The ability to specify which words within a phrase were essential to its value as a keyword, and which were loose guidelines, expendable, and fair game for substitution in finding alternative variants.
New +lots in central +Kansas
Again, before its final retirement, BMM’s ability to truly ‘pin down’ specific words had been heavily diluted… but its in-practice behaviour was still highly valued.
BMM wasn’t always in our toolkit, but a few years after coming on the scene in 2010, it had become the go-to match type for many PPCers.
There we have it. Google giveth, and Google taketh away.
Gmail Ads & Smart Shopping
Gone but not forgotten.
Spend wasted on traffic from outside your geo targets used to be one of the first things we’d check in an audit, before Google changed the way the locations report worked in late 2020.
Gone now are the joys of simply toggling to ‘user location’ to see a clear list of stats relating to the countries in which your clickers clicked.
Luckily this one was hidden rather than killed off completely…
We now just have to head to the Reports section to find this data, and create a table or chart with ‘user location – country’ (or ‘city’/’region’) as the dimension.
That Whole 'Data Driven' Thing...
Google Ads… AdWords… used to have quite a specific central theme at the heart of its identity.
The ability to act in a data driven way…
There are two sides to that idea: The ability to act, and the ability to analyse.
Skip through the graveyard above and you’ll see that every one of these tombstones acts as a nail in the bigger coffin for this central theme.
Here’s a chunk of the transcript from the first incarnation of my Paid Search course, written in 2018:
“Google Ads gives very clear and detailed insights about what’s going on with your campaigns… You have stats on how many times people see your ads, click on them – and go on to fulfil your goals on site… And you can cross refer all of that with information on where the users were geographically, what device they were using, their gender, age, the time of day, and other variables…
…If you analyse this data carefully, you can determine exactly what’s working and what’s not, and use those insights to optimise your activity.“
I couldn’t introduce Google Ads like that today with a straight face.
The name change of the advertising platform itself – made in 2018 – reflects the shift.
While the ‘AdWords’ picked out the key tools at our disposal for analysis and targeting (keywords)… ‘Google Ads’ broadens and de-specifies the scope.
Last week in a live session with my course members, I was reflecting that a recently-announced change to the RSA combinations report gives the appearance of an ‘FYI’ dashboard, rather than a display of data for genuine analysis or actionable insight.
An hour or two later, I opened
Data Studio what’s this? ‘Data Studio’ had become ‘Looker Studio’.
Which tells the same story…
A move away from the idea of advertiser as agent empowered by data, to advertiser as interested observer… the passive beneficiary of Google’s use of data.
You wouldn’t think it while I’m on a rant like this, but I believe Paid Search has a bright future… and six days out of seven, my focus is on how to get the most out of it in its current – and future – form.
I have a free mini course on the subject if you’re interested: How Google Ads Has Changed
I do think there are very good reasons to be wary of the speed and extremity of Google’s Big Automation Project… and in many areas, good reasons to resist it.
Do not go gently…