Most of the time, Paid Search optimisation tends to be reactive.
With a well-seasoned eye for opportunity, we see where current spend could be doing more work, or where more spend could be put to work, and we know which knobs to turn and levers to pull (at least in theory…) to make that happen.
But there is a limit to this kind of optimisation.
By reacting to account data alone, we can only improve outcomes within certain parameters.
When we step back from the narrow day-to-day optimisation, then we can question, recalibrate, and – ideally – expand those parameters…
There’s nothing definitive about the three levels of action vs tactics vs strategy I set out below. Their boundaries are blurred and there are plenty of other ways to draw them… but they may be a useful way to think about the different, mutually-supporting modes of Paid Search optimisation, and a reminder of the balance we need to strike between them.
Layer 1: Reaction
Here we assess the most visible signals of performance, and make the changes needed to upweight / expand our most profitable activity, and to down-weight or cull the most wasteful spend. Standard optimisation.
Except in rare circumstances, any action that comes under this simple reactive mode is generally ‘good stuff to do’ based on the principle of maximising conversions or revenue while minimising waste.
The danger is simply to spend too long in this mode, and forget to step out of it.
An unbroken, narrow focus on metrics (even the right ones) takes us down furrowed paths.
For one thing, it leads to the problem of ‘local maxima’.
While we may be doing about as well as we can given our current
• Campaign types
• Approach to keyword coverage
• Bid Strategies
• Landing pages
• Offer pricing
etc… we might need to come down from our local peak – the highest point in the immediate area… and explore the wider landscape of possibilities to find something closer to the ‘global maximum’.
For another, it doesn’t tell us much about where to strike the balance between competing aims. We almost always want more conversions, or more revenue… and we almost always (ideally) want to spend less to generate them…
Striking a balance between those two desires is where we earn much of our living… but without guidance, it’s not obvious where that balance should be in any given case.
Improving ROAS from 4 to 6 might look great in isolation… but if you’re having to slash absolute revenue to achieve it, there’s a point beyond which you’re degrading the effectiveness of the whole operation rather than enhancing it.
To keep those balances in check, we need to zoom out a little…
Layer 2: Direction
Here we know in more detail what we’re aiming to achieve, and we’re using that knowledge to strike the balance between more and more efficient…
Is our focus on growth? If so, the emphasis will be on volume over CPA or ROAS, and our preference in day-to-day decisions will be for…
• Aggressive bidding (and bid strategies)
• Upper funnel activity (and possibly an upper-funnel-oriented attribution model)
• Experimentation
• Expansion
• Maximising the good, with greater tolerance for the unproven
Knowing where on the scale of more vs more efficient your aim lies, hugely improves our decision making as paid search marketers, and ties our actions much more closely to our clients’ priorities.
Having actual targets for revenue/leads or for ROAS/CPA is better still…
(NB setting hard targets for both revenue and ROAS – though common (and far better than no targets) – is a little quixotic. Realistically your revenue target itself will set the limit of what’s achievable for ROAS, and vice versa.)
To calibrate your optimisation mode, ask whether you / your client is focussed more on new customer acquisition or profitability… growth or consolidation.. and other such overlapping concerns
Layer 3: Bigger-Picture Strategy
At this level, we zoom out, both in time – to take the longer term view – and in context, to break out of the Google Ads silo and encompass other aspects of the overall business effort.
The wider time lens forces us to consider where we are both seasonally, and in the life span of the account, and the business.
Now we can set a roadmap that respects the changes both in what’s possible, and in what’s most beneficial as an account matures, and a business (and its website) become established.
For example :
First we'll aim for the low hanging fruit (bottom of funnel traffic including brand and core terms) and establish a sustainable CPA...
Then we'll expand cautiously into upper funnel keywords and targeted display...
Then we'll start remarketing campaigns…
Then we’ll replicate successful activity on Bing Ads, and add Discovery and YouTube to the remarketing mix...
For this, we step up again from layer two, and ask whether our targets make sense with regard to where we are now, and where we want to be in a year’s time.
To zoom out in context, we’ll look at where Google Ads sits in the overall marketing effort.
This is relevant both to establish the role we need Google Ads to play…
(Are we covering the upper funnel chiefly with Paid Social, in which case our Google Ads efforts might be best directed to the core, lower-funnel activity?
Is the site not well established in SEO, in which case brand activity becomes a more valuable piece to play.)
… and to establish how PPC might collaborate with other marketing channels and overall business objectives.
For example, if you’re running Facebook Ads as well as Google Ads, look at how you can use audiences and insights gained on one channel, to enhance your targeting – and results – in the other.
Practitioner, Tactician, Strategist
In Google’s Search Summit last year, Nicolas Darveau-Garneau described Google’s advertising system as moving towards being like a self-driving car. You will just input the direction; the algorithms will find the best route to get there.
Raise an eyebrow as we might at the gap between vision and algorithmic performance… This trend will squeeze us out of the layer one… as we’ve seen happening for years, with precision controls being replaced by machine-learning-driven systems.
For now, there’s still often a huge amount of potential to improve on layer one alone. If you’ve carried out many PPC audits, you’ll know you’ll know that basic optimisation practices often just don’t happen as much as they should (and there’s usually room for improvement here in all of our accounts…)
But it makes sense that – to retain our value as the AI tide rises – we increasingly need to occupy the higher ground of strategy, and take a wider, bigger-picture view.
