Search Term Match Type
Here’s a useful dimension that doesn’t get its fair share of airtime… One that gives you a unique perspective on the value of the different match types in your search campaigns.
Search Term Match Type relates not to the match type of the keyword, but to the match-typing rules by which the search term matched the keyword. It segments performance – from impression to conversion – by each ‘band’ of those rules.
So if you have only broad match keywords, Search Term Match Type will split them into those three bands:
• searches that would have matched the keyword by exact match rules
• those that fell outside exact match criteria, but would have matched if your keyword had been on phrase match
• and finally, impressions that were only matchable by the extra capacity of broad match
You can find this metric as a keyword-level segmentation, or in the Report Editor (where there’s a quite a bit of gold to mine if you don’t spend much time there).
There is another, similarly named dimesion in Report Editor – ‘keyword match type’ – that’s the more intuitive one, and simply splits activity by match type at the keyword level.
But the key benefit of search term match type is the more granular insight it gives you into how and why keywords are performing as they are.
If your broad match keywords are performing well – it may be due to their success with search terms that would still have been triggered, had the keyword been on exact or phrase match (the inner rings in the diagram above) while the incremental contribution of its broader matches may in fact be dragging performance down…
This metric reveals that otherwise-hidden dynamic.
• Create this report via insights> report editor – add search terms match type, cost, conversions (or conv. value) and cost/conv. (or conv. value/cost)
Seasonality Bid Adjustments (in both directions)
Google describes Seasonality Adjustments as an ‘an advanced tool that can be used to inform Smart Bidding of expected changes in conversion rates for future events like promotions or sales’.
But they’re more than that. They are a control that gives advertisers a rare lever of influence over the behaviour of Smart Bidding strategies.
What seasonality adjustments let you change is the ‘expected conversion rate’ (‘expected’ by the smart bidding algorithms, that is).
As a result, the algorithms will become proportionately more or less bullish – affecting how high they bid, in which auctions.
This feature offers a significant element of control with smart bidding (even when on a Maximise strategy without a target).
SBAs are not new – and they’re not obscure – but they are (in my estimation) under-appreciated… and I’ve seen these use cases in particular fly under several radars:
• You don’t have to wait for a promo or any other major external event to make use of SBAs. If you ever want to increase (or decrease) bidding aggression across your account, SBAs are a neat way to do that (and don’t forget that they can also be applied to a single campaign or any subset of your campaigns.)
• Downwards adjustments (when you simply want to tune up efficiency in the short term, across multiple campaigns). An agency I work with routinely applies negative adjustments after promos, to compensate for the often dramatically suppressed conversion rates experienced post-sale.
• When you have various ROAS or CPA targets at the ad group level, and you want to keep those ad group differences in place while making an overall adjustment to the campaign’s bidding levels, rather than shift all of your ad-group level targets (which there’s no way to do in bulk 🙄) – keep them as they are and use a seasonality adjustment to push overall bidding up or down while retaining all those ad-group level gradations.
Location Exclusions
Since bid adjustments (generally) don’t work with smart bidding, variables like audience, location and device have most potency as a lever when you identify underperforming segments and exclude them.
Location is one of easier and more reliable variables for finding those patterns.
The relevant optimisation practice is to:
• Check the Matched Locations report (in the UK, the most useful level of matched location to view tends to be ‘county’ … it has a good level of granularity while breaking down UK targeting comprehensively).
• Take a long enough date range to gather a meaningful data sample on as a good number of locations
• Identify under-performers
• Exclude them from targeting (or separate them for differential treatment)
The Conversions-by-conversion-time Metric
In some ways it makes perfect sense: the ‘value’ of a conversion as optimisation currency rests squarely on what it tells us about the clicks that led to it. Allocating the conversion clearly to those clicks clarifies that relationship.
In other ways it’s unintuitive… and can lead to underreporting. (The conversion occurring on 5th December, allocated to clicks that happened in November, never made it into the monthly report that you sent on 04/12).
More importantly, it’s easy to forget.
When doing short term analysis – ‘What did our Google Ads activity produce and cost yesterday?’ – [Conversions or conversion value by conversion time] is by far the better metric than standard conversions / value… and the difference between the two can be quite substantial.
It also makes for quite a useful quick check if you’re ever wondering whether a couple of conversion-less days point to a tracking issue.
Check for conversions-by-conversion-time – and if any of these show up in the last few days, then conversion tracking was working when they did.
Date-based Custom Columns
Most of our (day to day) analysis in the Google Ads interface is best done with a ‘middling’ date range – say – 30 days.
We need sufficient data to see real patterns, and to contextualise the very short-term changes we can see on the charts.
But we also often need to see ‘the latest’: What did each campaign spend/achieve yesterday?
And so, date range switching between ‘30-days’ and ‘Yesterday’ is a key contributor to repetitive strain injury among us PPCers.
That’s where these simple custom columns: Yesterday Cost, and Yesterday Conversions come in:
For Yesterday Cost, head to:
Columns > modify columns > new custom column > enter ‘cost’ > filter by > date range > yesterday
And slide it over to the left, so it appears right next to the daily budget – giving you a quick view of whether (or how nearly) each campaign met its budget yesterday.
Similar columns for ‘average daily spend’ over 7 days (or whatever time frame you want to keep in place) or average CPA/ROAS, are easy to set up, and can seriously streamline your top-level analysis.
