Bidding on your keywords, much like account structure or landscape painting, lends itself to a certain amount of personal interpretation. Figuring out your starting bids, how to adjust them over time, and projecting the expected impact involves a certain level of guesswork. While there are tools available now to help you set your starting bids, knowing how much you can afford to pay and what it might do for your performance takes a little bit of math… and some help from the Bid Landscape tool.
The Bid Landscape: What is it?
Simply put, the Bid Landscape shows you how different estimated bids would have impacted your performance over the past seven days. At each bid level, you’ll see the estimated amount of impressions, clicks, cost, and top-of-page impressions, giving you an idea of whether or not the traffic increase is worth the jump in average CPC. Here’s a shot of it in action for a competitor keyword over a seven day span (12/26/14 – 1/1/15):
Their current bid price is set at $.65, with an average CPC of $.55. At their current bid, they’re limited to an average position of about 4.91 -- pretty solidly on the side of the page, with very few mainline impressions to be found. While some of these suggestions are fairly aggressive, there are a significant number of additional impressions out there to be had for this particular keyword -- impressions that will translate to both clicks and conversions.
You may not have realized this, but the estimated click totals do account for an increased CTR from moving to the Top 3, which I’ve laid out here in Excel:
By merely increasing our bid from $0.65 to $0.91, our CTR is estimated to double, and our click volume is seeing an estimated five-fold increase. Meanwhile, our Average CPC is expected to rise from $0.55 to only $0.69, so while we’re seeing a larger increase in cost, our overall efficiency is still well within our tolerance at a 25% increase.
Taking this tool to its logical conclusion, you can also project out expected conversion gains/losses, as I’ve done here:
In the above example, I’ve calculated our estimated average CPC, as well as our estimated conversion metrics based off of two different models: one assumes a linear conversion rate (which is probably unlikely), and the other assumes a falling conversion rate due to the introduction of more traffic, both from being higher on the page, and from being more competitive in more auctions.
That falling conversion rate is a worst-case scenario (and also not a scientific calculation on my part at all, just a gut instinct based off of experience), but it can be useful to temper your expectations. The higher bid will bring in more traffic, and depending on the match types involved, you may need some negative keyword additions to make sure that traffic is qualified.
How Much Should You Pay?
So we know that there’s traffic available, and we have a reasonable idea of what that traffic might cost us… so given these choices, how much should we actually be paying? The simplest answer to that question lies in two metrics: what’s your ideal Cost per Conversion for that keyword, and what’s your Conversion Rate?
Here’s a two-part formula that I’ve used for years. While it can be done in one step, I prefer the two-part version for the ease of explanation.
Simply put, know how many clicks it takes for you to get a single conversion. If you have a Conversion Rate of 1.00%, then it takes 100 site visitors for you to get that one conversion.
Based on your business goals, you probably have a Cost per Conversion in mind. If we have a $100 goal Cost/Conversion, and it takes 100 clicks to get that conversion, then it follows that the most we can afford to pay based on our goals is a $1.00 CPC.
For our above advertiser, they have a 1.06% Conversion Rate and a $75.00 Cost per Conversion goal. Based on that math, it takes 94 visitors for a conversion to happen and a $75.00 Cost per Conversion goal, that gives us a break-even CPC of $0.79. Based on what we learned from the Bid Landscape tool, we can then reasonably assume that the $0.91 bid is a safe bid for us (given the estimated average CPC of $0.69), and we likely have some overhead to be even more aggressive. Obviously, these numbers will change as more performance data rolls in, but at least now we have an idea of where to begin and what we can hope to gain by making this change.
What about you? Find any actionable insights coming from the Bid Landscape, or something you’d like to see from that tool? Any tips on using the break-even CPC in your day to day work? If so, feel free to let us know in the comments or ping us on Twitter.