There are plenty of automated bidding solutions out there. Use them. But don’t think that they know everything. There are plenty of situations where automated bidding doesn’t have enough info to act on. New product builds, funneling traffic to the right keywords, and highly seasonal campaigns are all areas where automated bidding can fail you if you need quick moving results—considering most automated bidding solutions run best with a few weeks of data.
In these cases, it helps to know what you’re doing for managing bids manually. In the interest of time (and so you’ll still have something to pay us for), I’ll just be focusing on one specific metric you can take advantage of.
It’s a big one though and it’s especially important for new keywords. How do you decide when you have enough data to push a new keyword’s bid up or down? How do you decide when you have enough data to pause a new keyword?
The key point here is that you don’t just need to know if you should push a keyword bid up or down but you need to know when. Anyone can look at 6 ad groups that have 10 conversions each and a profitable ROI to say you should push up bids there. It’s a lot more difficult to make decisions on 35 ad groups you just launched 3 days ago that have a small amount of spend and 0 conversions.
An easy way to figure this out is to determine your inverse conversion rate. The formula for conversion rate is:
So finding out the inversion conversion rate will let you understand how many clicks on average it takes for your keywords to get a conversion.
Next Steps:
- Find a segment in your account that seems comparable to your new keywords/ad groups
- If it’s a new NonBranded campaign, I usually look at overall performance metrics from NonBrand in the last month to get a benchmark.
- Here’s a typical data set:
- 205 Conversions
- 8489 Clicks
- Conversion Rate = 2.41% (or .0241 in numeric terms)
- Now you, take the inverse of conversion rate and get this:
So you have 41.5. This is the number of clicks on average that it takes for 1 conversion to happen. I call this the click threshold. Because it lets me know if a keyword has enough data for me to push it up or down (or in some cases pause it)
If a keyword has 20 clicks and no conversions, I won’t push the bid down. Because I know my average click threshold to get a conversion is 41 clicks. So there’s not enough data to make a decision on this keyword yet.
Now you can use this data however you’d like. Your NonBranded conversion rate will be different than your Branded conversion rate so your click thresholds should vary between these two segments. Sometimes it makes sense to determine the click threshold for different categories within NonBranded campaigns if your keyword performance changes drastically from one category to the next.
The overall goal is to get your keywords enough traffic to make the most informed assessment of their performance before changing their bids. Don’t give up on keywords or ad groups before you have enough data to do so.