What are smart bidding strategies in Google Ads?
Any bid strategy which incorporates website behaviour, is fed data by a “conversion” and can be directly linked to a business objective is going to be a smart bid strategy. In fact, the only truly Smart Bidding Strategies that can be used to deliver genuine business growth in the long-term are tROAS and tCPA.
target CPA (£)
target ROAS (%)
Maximise Conversions (volume)*
Maximise Conversion value (£)*
eCPC
“*Max Conv | Value bid strategy should not be thought of a smart bid strategy as it has a number of caveats and should only be used in special circumstances”
Anything which doesn’t seem to incorporate intrinsic business value as part of the bid is definitely not a smart bid strategy! These include:
Maximise Clicks (website visits)
Target Impressions Share
Manual CPC
CPM (Display Only)
“Target Impr. Share and Max clicks are “automated” but not Smart bid strategies ”
The Primary “smart bidding” strategies are tROAS and tCPA. “Maximise Conversion Value” and “Maximise Conversions” are secondary smart bid strategies as although they are steered by website conversions they should only be used as a stepping stone to tCPA or tROAS and have a few other caveats.
Performance Max Campaigns (PMax) can only be used with Smart Bid strategies (inc “Maximise” variants)
Target ROAS (return on Ad spend) is a primary bid strategy for ecommerce sites active on Shopping campaigns & Performance Max Campaigns (PMax)
Target CPA bidding is a primary bid strategy for lead generation business models
eCPC is also a “smart bidding” strategy but Google has down-weighted this as a preference in favour of tCPA and tROAS
“Maximise Conversion Value” bid strategy always seems to stick out as it is actually only available in search campaigns (and now PMax) but not “standard” shopping campaigns. This is a useful bid strategy for ecommerce sites active on Search campaigns
Target Impression share on Search campaigns is a position based bid strategy based on the competitiveness of the auction so while it is automated, it is steered by auction dynamics and not any “activity” based metric from the website such as a conversion.
What’s so smart about smart bidding anyway?
Smart bidding features changed the Performance Marketing landscape. It uses real-time signals based on the likelihood to convert and is uniquely tailored to:
your business
your website visitors and customers
your budgets
your marketing objectives
how well your websites converts users
In other words, it makes a split second decision from the impression of the users actual search (not the keyword in the account), incorporates historical data about that users intent (if available) and combines it with statistical data as to the value this search/user is worth to your business. It is truly end-to-end and as such it requires website conversion data and some volume of data to test and learn with.
When is it a good idea to test Broad Match?
Thinking of enabling Broad Match because it’s Recommended in your account? Broad Match only becomes “Smart Match” when a Campaign :
has maxed out its current potential, i.e.
dominates in Market Share,
is not Limited by Budget
and is hitting business goals based on tROAS or tCPA.
When these requirements are met it’s possible that “smart match” can go beyond specifics of queries and find new incremental conversions as it has high quality conversion data in terms of volume and recency.
Broad Match becomes Smart Match when used in conjunction with Smart Bid Strategies, tROAS or tCPA on Campaigns that have no additional headroom in their markets.
Broad Match Super Powers Are Not Available to other Match Types
Which Smart Bid Strategy to Use and When to Use it
With all the automation that is happening, actually defining the bid strategy and applying its targets at the the right time and place is a primary component in improving performance and delivering incremental results. It is a huge lever that should have an even bigger warning sign about the effect changing this campaign level setting can have. Not having a full understanding of the business goals, allocated and future budgets and historical performance can lead to “user error” about how the strategy should be adjusted, which will send campaign performance down the toilet!
“Defining and setting the bid strategy allows good or bad data into the system to learn from. Strive to only let in the good stuff.”
Minimum Data Requirements When Using CPA or ROAS Bid Strategies
1. 20 Conversions in 30 Days per Campaign
This is a minimum threshold (or whatever the latest number is as this gets less as Google gets better) of good data that the smart bid strategy needs. Google says so. It can vary and it might be more or less depending on the business operation. What’s important is that it is a consideration based on volume and recency. For Performance Max Campaigns, it may even be prudent to only sub-divide Asset / Listing groups if each one can support at least 30 conversions per month.
2. Never be “limited by budget”
Not having access to as many data points as possible will cause the bid strategy to under perform and should be avoided. If a campaign does have a “warning” about being limited by budget then either:
increase the budget
pool resources with other campaigns in a Portfolio if there is budget headroom in the other campaigns to do so
change the bid strategy to “Max Conv | Value”
If this happened recently and conversion volume is not an issue then increasing the tROAS is also an option as long as by doing so the campaign becomes no longer “limited by budget”.
“Data Quality and Data Freshness is key”
What Happens if ‘Target’ Bid Strategies are budget constrained, conversion limited or the Target is too optimistic?
If the Machine Learning cannot learn, over time the campaign will slowly die.
Too high a ROAS target that can never be achieved will result in degradation of performance such as less spend, less traffic, less conversions
Not enough conversions? Same
Not enough Budget? Same
Flowchart on how to Extract Maximum Value From ROAS and CPA Bid Strategies
NO
Change Back to Max Conv | Value Bid Strategy
NO
Search Impr. Share > 60 - 70%
YES
Get more conversions by: expanding your keyword coverage with DSA ad group if none have been enabled.
Expand Match Type to include a copy of all exact match as Broad Match
Done all that?
Never be limited by budget:
Increase budget?
Lower ROAS? / Increase CPA?
Lost Imp Share (budget) > 10%
YES
Improve Quality: Most time optimising for performance (within Google Ads and on Website) should be spent in this state and covers majority of effort, discussions, tactics and testing
NO
Hitting Bid Strategy Target
YES
YES
NO
This framework is also relevant to Performance Max except IS can’t be used so you’ll need to be clever to dig deeper than google recommendations or Performance Planner.
“Maximise” bid strategies should be handled with extreme care
A “Maximise” bid strategy is suitable for Campaigns that are limited by budget or newly launched Campaigns in a new marketplace where there is no historical data to take advice from. If there is not enough budget to grow the account then indefinitely using this bid strategy is fine but for everything else it should be used as a stepping stone to tROAS or tCPA. The opposite is also true, in that tROAS and tCPA should not be used if it’s not possible to stay uncapped by budget. The easiest way to know this is that the Daily Budget is always higher than the amount being spent.
“Do not continue to use “maximise” if you are consistently hitting a Search Impression Share (IS) over 70% or are not loosing any IS due to budget (every day of the week).”
As budget is the main limiting factor then use a metric capable of delivering insights about it. In the flowchart below I’ve used Search Lost IS (due to budget) but I might substitute this for Search Impression Share.
If the campaign is not budget constrained, the bid strategy will go out of its way to spend all of the budget it has in the desire to get more traffic that might convert. This could mean a “sensible” CPC of £5.49 might turn into £27.30 per click. The bid strategy is simply doing the best it can with the bad guidance it’s been given.
“Be careful not to turn a ‘Max Conversion’ bid strategy into a ‘Spend all the budget on anything’ bid strategy ”
Flowchart on how to manage “Maximise” bid strategies
Lost Imp Share (budget) < 10% and > 20 Conv in 30 Days?
Reduce budget to ensure it is always ‘limited’ while reviewing other options to pool resources so that minimum conversions are consistently achievable
Lost Imp Share (budget) < 10% and < 30 Conv in 30 Days?
YES
Get more conversions by: increasing budget | expanding keyword coverage | expand targeting | improve relevance
NO
Lost Imp Share (budget) > 20% and < 20 Conv in 30 Days?
NO
YES
YES
Assuming future budgets and conversion volumes remain above minimum thresholds
How Do “Automated” Smart Bidding Strategies Work with “Manual” Bid Adjustments?
As a general rule, bid adjustments are ignored by Smart bidding
Bid adjustments cannot be finessed to any great extent when Smart bid strategies are used. Some options are available but these are reserved for making strategic decisions about binary scenarios, for example, Mobile Devices can either be completely excluded or not. There is little middle ground. Rather than being outraged, this makes complete sense, as with 1000s of signals and insights available to the strategy dumbly pulling on any manual overrides is pure vanity.
“The only significant manual adjustment to Smart bidding is Audiences”
Audience bid adjustments are likely maintained as a feature as they are fundamental to RLSA (Remarketing Lists for Search Ads) campaigns.
As part of any account audit, one of the first major Campaign settings that I look for is that:
Smart bidding is being used across all campaigns
All bid adjustments are removed
It is in no way transparent within Google Ads that previous levers are not being applied. For example, there can be bid adjustments on location of +50% or Friday evenings of -25% which are not in use and are being ignored by the smart bid strategy. This is why I recommend removing them completely to keep things tidy and avoid false interpretations of what is and isn’t impacting performance.