My process is to always align the key metrics with the business KPIs. For example if the client is ecommerce, then conversion value/cost (ie. ROAS) is the most important column, followed by conversion value. Whereas, if the client is more lead-gen focussed (their goal is to get form fills or phone calls), then cost/conversion is most important, followed closely by conversion rate and total conversions. CPC is important for both of these examples because it directly affects the results, but the cost per action you’re aiming to complete is more important. For example, if one keyword has a high CPC which you choose to not include, it may also have a really high conversion rate which actually performs better in terms of ROAS than a lower costing keyword, you could be missing out on a valuable opportunity.
The other metrics are important to decide where you allocate budgets. For example, if one keyword has a conversion value/cost of 5 (ie. a 5x ROAS), we know it’s performing well, but it might have an impression share of 95%, in this case there isn’t much point in increasing budgets for this keyword because we’re already appearing for most of the searches, but we might want to look at ad copy to improve the CTR. If the CTR increases, we can then put more budget into the keyword because we’re driving more clicks.
On the other hand, if we have a keyword that has an excellent CTR, a really low CPC and thousands of clicks, but a non-existent conversion rate, you need to really think, is this actually doing anything for me? It may be a good upper funnel keyword that’s giving you good brand exposure, but using other mediums such as display would probably do that for cheaper. This the kind of keyword that may look good on the surface to an amateur, but actually isn’t doing much for the account except costing money. This is the kind of keyword I’d pause.