There’s a strange thing that sometimes happens with CPA bidding in Google AdWords.
Following a few days of fluctuation, the ad impressions drop to almost nothing.
There doesn’t appear to be any change to the settings and the billing is OK – so what’s going on?
In most cases it appears that Google’s algorithm simply gets nervous – you might even say it panics!
To understand the phenomenon in practice, we have to understand how CPA bidding works in principle.
Once we have a threshold of conversions in our account (generally around 30 in a month – so one a day) we are able to switch our bidding strategy to CPA bidding, essentially telling Google to get as much traffic as it can at whatever price it can provided that the conversions are being generated at a cost no greater than the target CPA.
Google will suggest a CPA, bit we are also able to override this suggestion and fix our desired target – more of this in a moment.
At this point, Google will look to combine a myriad of factors from the search term to the geographic location of the searcher, any demographic information that it might have at its disposal coupled with factors such as the device being used and the time of day, in an attempt to pitch their bid at just the right point to get the conversion within budget.
And it is at this point that the algorithm can begin to show human traits. Because in certain circumstances it simply fails to hit the target CPA. So it makes a few changes to see if it can get back on track… but nothing happens.
It then takes a logical step. I say logical, because it actually makes total sense and no sense at the same time…. it drops the bid. Sometimes significantly. Because the “thinking” is that if the bids are low enough we should be more visitors for the budget and have a better chance of getting a conversion at a lower cost.
But this rarely works. The panic that causes the bids to drop simply exacerbates the situation and the campaign collapses in on itself with bids of one or two cents and zero impressions. Poor Google… it feels worthless and ashamed at having failed to hit the targets.
So what can be done? In the first instance, give Google a fighting chance. Whilst the minimum number of conversions might be 30 a month to initiate CPA bidding, this is actually very low – the more conversions you can give the system to work with, the better.
And be generous with the initial CPA bid. If Google suggests $50, perhaps try $65. The temptation to override the suggested bid expecting Google to somehow find the traffic that will come in at $25 is powerful, but you should resist – you are setting things up for failure.
And finally, keep a close eye on the CPA campaigns. They can spiral out of control quite easily, and you are better to take over the controls sooner rather than later before more damage can be done.
I’m not a great fan of CPA bidding – but I don’t blame Google. I sometimes think I’m just being mean and expecting too much from an algorithm whose feelings can be hurt too easily.
How have you used CPA bidding – has it worked well, or have you struggled to find the bidding sweet spot? Use the comments below to share your experience.