1. Is it better to start with manual bidding or Maximise Clicks?
PPC managers often launch campaigns on Maximise Clicks, with a view to shifting strategy once a reasonable volume conversion data has come in.
When there is enough evidence to guide decisions about where bids should be raised or lowered, we can then either shift to manual or – more often – move to conversion-based smart bidding strategy and let the algorithm make use of those data.
There’s good logic to this sequence… but the shortcomings of Maximise Clicks will show up, even in this limited role as the warm-up act.
As all clicks are not equally valuable – the pursuit of maximum (and undifferentiated) clicks, quickly runs against what’s really valuable to us.
The trouble with maximising clicks – regardless of where those clicks come from – is that crappy clicks are often cheaper clicks. (That’s true across multiple dimensions, as discussed re the limits of CPC as a metric in an earlier post.)
Under Maximise Clicks, the algorithm would have every incentive to go after those cheaper, crappier clicks preferentially, and positively avoid the more expensive clicks that are of higher value to you.
Aren’t those differences in click value what you use Maximise Clicks to reveal, before switching to a more suitable strategy to make use of them?
Yes… but even before conversion data accumulates, there are always grounds to favour some clicks over others… There’s:
- Your education assumptions about which keywords/search terms either deserve or need higher bids
- Engagement data such as bounce rate
- The first signals from early conversions – before we have enough of them to let the conversion-based bidding strategies loose
Maximise Clicks will not make any attempt at all to differentiate your clicks, except on the counterproductive grounds of cost alone.
That gives manual bidding a decisive edge in most cases.
2. Should you bid on brand terms?
The most serious objection to brand bidding is empty spend.
More precisely, it’s the waste of paying for clicks – and conversions – that were already heading your way regardless.
This is a strong point. A large proportion of paid brand traffic – especially on pure brand terms – will be coming from a big bite taken out of organic traffic.
On the other side of the scales:
1. Paid brand activity will bring at least some genuinely incremental traffic. How much is open to debate (and often debated) but ‘some’ is a safe bet. And that incremental traffic will of course be high-value traffic.
2. Brand ads allow you to control and update the messaging shown to those search for you, far better than you can with your organic results.
3. Ads can show higher in the results pages than the highest organic results. This is particularly important when competitors are showing ads against your brand terms. When they are, your own ads are often your best defence, ensuring that your results are what users find first when looking for you.
4. Those brand clicks will be cheap. High relevance, strong Quality Score and low competition on your own brand terms should ensure low CPCs. This is more damage limitation than a real benefit, but worth remembering as you weigh up the other pros and cons.
Patrick Gilbert (in Join or Die) draws the useful analogy between brand bidding decisions and ‘Pot Odds’ in poker… where the player weighs up the cost of participating in a hand along with both the chances of success and – crucially – the potential winnings.
While the objection of ‘traffic cannibalisation’ slashes the chances of ‘winning’, the low CPC lowers the risk, while the nature of a brand search multiplies the upside.
(so… yes, you probably should)
3. Which are the most important interface columns?
The number of columns available in the Google Ads interface has expanded rapidly over the last few years (while the depth and precision of the actual performance data available hasn’t… but that’s another story).
The few columns I’d prioritise as the most essential are:
Inputs
- Cost
- Impressions
- Clicks
and as a function of these metrics: CPC and CTR
Outputs
(success metrics):
- Conversions
- Conversion Value if you have it
(and if you don’t have, try to bring it into play)
On the relationship between input and output:
- CPA (cost/conv)
- Conversion Rate
- ROAS (conv. value/cost)
These are all key metrics whether you’re looking at campaigns, ad groups, keywords, search terms, ads, audiences etc…
Opportunity
- Impression Share Lost to Budget*
- Impression Share Lost to Rank
*only available at the campaign level
Status
There are several that may be needed, e.g. Max CPC at keyword and ad group level, with manual bidding / Target ROAS if using the tROAS strategy…
But these depend on the specific setup)
Honorable Mention
The other ones that you could argue for as key columns are Quality Score and its components, in the keywords report.
I wouldn’t elevate Quality Score metrics to ‘essential’ status for reasons I’ve gone into before.
3. Why do higher CTR ads often have lower conversion rates?
Search ads play two roles that are often in tension:
- Generate more clicks
- Generate the right clicks
Ad text that appeals more widely will of course bring in more clicks, but will usually raise the proportion of clicks from users that we have no realistic chance of converting.
Showing the price in an ad is a classic example.
Revealing the price upfront should immediately repel a segment of users for whom your product, at your price, does not appeal… (CTR down)
On the other hand, a higher proportion of those still interested enough to click, will be potentially convertible… (CVR up).
In another example, NextAfter – specialists in marketing for non-profits – have published an edifying case study, showing exactly this CTR-CVR inversion.
On testing the inclusion in headlines of a ‘strong CTA’ encouraging users to make a donation, ads with the CTA showed a dramatic decrease in CTR, but an even more substantial improvement in conversion rate.
Though immediate cost effectiveness sits on one shoulder, and Quality Score on the other… for simply maximising conversions, the two metrics carry precisely equal weight.
