We’ve all been there in our accounts—keyword isn’t performing well, and its at position four on average, lets throw money at it until it stops being such a pain. SMART! Then dust off your hands, and go back to the Bieber vid you were secretly watching (we know you do, we can see your queries). But then tomorrow comes around, and hey hey, the average position is way up! Yes! But then looking at the other metrics…and unlike your projections, they are actually worse! High CPL, high spend, low conversions…but why? Well, as I have found through some work with a particularly fickly-performing client, some keywords just perform optimally in a small range of positions…a “Sweet Spot.” Let me explain.

Justin Bieber
We know what you're watching.

To start, it is pretty logical to assume that with majority of keywords, if you bid higher, your metrics will improve. Yes, that is a simplistic analysis, but in general, hey, that’s a pretty solid prediction. So, while bidding on high value keywords, this is what I came to expect. I improved a keyword’s average position that had been performing well to around top position, expecting to reap the rewards. However, I found that even though it was spending more, it performed better in terms of CPL and conversions at a lower average position.

I spent long hours trying to find out why certain keywords performed better at lower positions. There were many variables to consider. I checked the competition in terms of my ad space, I looked at the quality score to see if maybe it had dipped, I looked at search queries., I felt I had exhausted all my options. And then I looked at a wider range of data, and segmented by month in order to see what had changed. Simply put, the only thing that I could see changing month to month was the average position. And then I came to a realization.

Light Bulb Idea
Eureka!

When it comes down to it, very few times will we be able to, as internet marketers, narrow down our problems to one variable. And oftentimes, there are too many variables at play to concretely act upon one. So, all we can do is obey the numbers. If a keyword performs best at position 3.5-3.7, and despite all logic and mathematical projection does terrible at position 2, then keep it at position 3.5-3.7. This, is what I have deemed, the “Sweet Spot.” For infinite combinations of reasons all playing together, certain keywords just perform better at a range of positions. And here is how you figure that out.

For as many keywords as you want (I typically do a top 50 list, otherwise this can be very taxing on both Excel and your brain), look at the data for a 6 month to year long spread. Segment by month in your spreadsheet (it will be HUGE), and identify which 2-3 months your keyword performed “ideally.” And then, take the average of those average positions, accounting for a small spread, typically .2 in each direction. And voila, you have your optimal bidding position, or “Sweet Spot.” Target this position, and keep in mind that the bid may change month over month, but the average position should stay stagnant.

To clarify, this is for problematic keywords—i.e. the ones that seem to defy logical best practices. This is a great way to add another metric, another way, to explain poor performance, and then act upon it. So, gather up your problem keywords, check out the data for a wide range of time, and see if their position fluctuates. And if so, find out where they best perform.