I've been tracking my conversions - and know which keywords work. So far I've been keeping my bids low (under .10) - but I know there's more traffic I can get by spending a bit more.
The question is - how much more do I spend per click (if any) - and how much traffic will I get?
I haven't seen this discussed too much around these parts - so I decided to do a little experiment. I setup an excel spreadsheet, and ran a handful of keywords through google's traffic estimator. Assuming 1st place gets 100 hits/day - here's what I came up with (this was pretty consistent across multiple keywords):
1st place - 100 hits / day
2nd place - 77 hits / day
3rd place - 60 hits / day
4th place - 40 hits / day
5th place - 30 hits / day
6th place - 21 hits / day
7th place - 16 hits / day
8th place - 12 hits / day
9th place - 9 hits / day
I've found Google's traffic estimator to be pretty flakey for predicting traffic on specific keywords - but I do suspect it's based on generally valid data.
Initial conclusion - if you're in 7th place, there's a lot of traffic you're missing. If you can bid your way up to 3rd and still keep a good ROI - you're probably wise to do so.
However, if you're already running close to break even - and you have to double your bid to make it from 3rd to 2nd - you're probably wasting your money.
Being #1 may yield the most traffic - but it's not by a huge amount.
Thoughts?
When the nr 1 spot is at the top "premium" position the volume typically soars. However this is very inconsistent. I think what you're doing makes sense, watch you ROI as you make your way up....
The Google traffic estimator is, in my opinion, utter rubbish. The only way to see what works is to try it real time. And one other thing I can guarantee - if you bid higher you will definitely spend more money.
Another example, I set up some aff links selling very specific toys for Xmas - Google's traffic estimator said that a bid of 4p would generate me no traffic. For over 75 keywords, I'm now at the very top on the "Sponsored Link" band!
My own recommendation is to do some maths (or math, as you say in the USA) as follows..
Let number of vistors needed to make a sale = x
Let average margin on sale = y
Max CPC must equal less than y divided by x - how much less is up to you, but it needs to pay your bills.
I'd also recommend looking at Overture and maybe some other PPC's before increasing your bids on G. Overture, for me, averages 25% cheaper than G for the same keywords.
I agree that the traffic estimator does a bad job of actually estimating traffic (it's gotta be virtually impossible to estimate CTR on a given ad for a given keyword). But - my bet is that google probably got the easy data points right - like the traffic fall of from 1st to 2nd to 3rd to....
I did a little more math on the stats I collected. Seems like the fall of is about 75% per rank. For example 2nd place is 75% of 1st place, 3rd place is 75% of 2nd place, etc (I'm sure there's a more elegant way to express this).
Assuming the rank / traffic fall-off is about right and Google comes up with a remotely correct estimate of your placement for a given keyword for a given bid, it should be possible to combine this with conversion data in a spreadsheet to try to answer the question:
How much should I bid on a given keyword to make the most money?
If you are listed 9th and someone clicks then the chances are that they REALLY liked your ad. If you are listed 1st then they might just be casual browsers; they may be less likely to continue through to check-out, and if they do check-out, they might buy fewer/cheaper items.
As you say; the objective is not more traffic at any price, but a better ROI. Measurement is key :) Don't assume ROI ratios will stay the same.
James
While its possible that someone clicking on a low-placed ad just really liked yours, it also quite likely that they are doing a lot of comparison shopping.
Ninth place ads are not going to catch much of the "buy the first acceptable match that I find" crowd.
It would be kind of neat if Google passed the ad position in the referer string.
Our conversion ratio is about the same no matter what place we're in.
In the top postions you get the "Ninth place ads are not going to catch much of the "buy the first acceptable match that I find" crowd." as CernyM put it, although, you also get a lot of window shoppers in that position, so it averages out the conversion rate.
I think those programs use the old max cpc < profit/convertion as their essence.
Yes this info is managed by GoToast et al but at a price. My campaigns (and budgets) are relatively small, but complex, and I'd like to gather/analyse the info myself.
James
