i've recently 'launched' my first e-commerce site, and i can say that it's been a slightly humbling experience. i'm the type that researches thoroughly before i act (in business) and in this case, i was certain i'd done all of the requisite homework.
i used adWords and Overture estimation and keyword tools, and read forums here, all in an attempt to estimate my potential: roi, conversions, ctr, volume and so on. i felt that i came upon a great set of keywords...until i deployed!
i've set my adwords acnt up like some here have suggested, about 10 adGroups with 1-2 terms in each, 4-5 ads in each term to measure effect on CTR. all terms are exact or phrase matching, and i dont have many negatives. now...i know i could use more phrases...but here's the issue...my volume of impressions is SO LOW that its lower than the CLICKS that google estimates i'll be getting per day.
example:
keyword: [widget], google est.clicks/day=540. actual clicks=3. impressions=200.
keyword: "black widget", google est.clicks/day=140. actual clicks=0. impressions=70.
whats going on? i've heard you guys mention several times that it takes a cpl days to get on 'content network' and i just started monday...is that the issue? are the impressions really low on google alone? i know its not the campaign budget, i have that set at a cpl hundred a day...more than google suggests. Also, all of my terms inside the adgroups are strong or moderate (even though have low ctr, they have a small amnt of clicks) so have not been disabled. my overall ctr is %.6, kind of poor, but hey, i just started on mon.
when i go to google, i do see my ads there for most of the phrases...including the 'bigger' ones that they estimate i'll be getting 500 clicks a day on so i dont think its that they arent showing....so im confused. i mean, i can see google being 20 or 30% off, but they are like %10,000 percent off here. anyone know the deal?
pk
First off - it sounds like you did a good job of setting up your accounts :) - early organization can save you a lot of time later on.
Yes, the estimator can be that off.
Odds are your ads are not shown on partner search sites yet - are your ads showing on search.aol.com?
for example, is this the way i should be setting these up, or does this cause (hopefully) a horrible bug that reduces your impressions by a factor of 10,000? :)
adgroup 1:
[widget]
"widget"
adgroup 2:
[widgets]
"widgets"
adgroup 3:
[blue widget]
"blue widget"
adgroup 4:
[blue widgets]
"blue widgets"
that's the general form i've used. i'm praying that it's incorrect - i'd rather be wrong and have more traffic on the way than be right and slow.
Basically, with all of my terms, adWords estimated i'd be getting ~1k clicks a day, which i thought was probably way too positive, and toned it down in my estimates to 600, but instead i'm getting about 40.
so how much do the partner sites kick in for traffic? and before you guys say "it depends", please give me an estimated range from your experience :)
pk
I have to say it: 'it depends' :). I've seen from as low as 10% to as high as 40%+ increase in search volume with partner sites.
You might want to check out inventory.overture.com and see what type of numbers they have to compare against Gs estimates.
clearly you opted out of broad match, but when you did your estimates did you use exact and phrase matches?
run a check to see the broad match competition for your keywords. a good way to do this is to run a search for your roots while adding a meaningless qualifier like: sasdfds
this will show you those that are broad matching your roots, which can have a significant impact on your impression levels, depending on how many there are.
my guess is that your problem is related to your daily budget which affect the display rate algo or broad match competition which the estimator has significant trouble taking into consideration.
my daily budget is set to 2x the google 'suggested budget'...so i dont think thats the issue. i don't think that the 10-40% potential gain from partner sites will put me anywhere near the expected clicks.
part of my distance from the expected clicks is of course due to my inexperience in improving my CTR and writing ad copy. my campaign is averaging 0.6 CTR or so....pretty sad.
just an estimate on how far i am.
expected clicks from google/day: 1164
actual clicks from campagin today: 48
total impressions from campaign today: 4,570
so...unless google wants me to get a CTR of 25%, i dont know how i'd possibly get anywhere near my expected. have you guys seen the numbers this far off?
i'd assume that even for their estimates, they give a CTR of what...2.5% or even 5% to make it look good? but that i'd still be missing about 10x the amnt of traffic with that CTR. hope i can find that missing 10x traffic out there somewhere :)
speaking of that, i think i just found some of it. when i type in "green widget" to google, i dont get my ad in the results...even though i have "widget" as a search term. shouldn't it show as part of the phrase? or have i really misunderstood the "phrase match" concept.
thanks again,
pk
'green widget' doesnt show my ad in search results and i have a "widget" keyword which i thought should phrase match.
go in, and increase your daily budget by 3-4 times and run that same search again, i think you will see what i mean about the display algo
{KeyWord:xyz}
this will give you an immediate boost, although i say that reluctantly as i am not a fan of the service from a users standpoint. from an adserving standpoint, it simply put, works.
good luck.
looking on overture, they estimate that i'd get about 800 clicks per day using all of the terms that really apply to my widgets, and no unrelated terms (though they say that is with a 5% CTR which they claim you might get in the top 4 positions).
so i guess the amnt of traffic is really out there somewhere...but i'm just not really getting it for some reason.
thanks for the continued advice. i can see how this could be considered an art.
pk
i am not a fan of the service from a users standpoint.
Kind of off topic, but I agree with this.
There are a lot of sites that use dynamic insertion, but either thier product is not what I'm looking for, or their landing page doesn't tell me anything.
I'm getting to the point that when I see a dynamic insertion ad, I ignore it for something that stands out.
i've played around quite a bit with the budget, do you think i may have bugged the algo? setting it to
$800 (which was 8x above the suggested) didnt have any effect on getting the ads to show...
and i know the broad search terms are working for my competition...they show the same ads for all of the broad matching terms, so i'd assume they are all broad matching. should i scrap it and start a new campaign?
pk
There are 3 categories of matches on that page.
In most cases "more specific keywords" show for broad match.
For "expanded broad match", those words generally, but not always trigger a query (hence the Google wording 'may' trigger your ad).
Then there is 'additional words to consider', these are related terms which will definately not trigger your ad, but give you a good starting point of more keywords to consider.
If the expanded and more specific suggestions are multiple word queries, it's suggested you add them as keywords so you can have more control over tracking your keywords and knowing which will preform. Thus not relying on broad match todo all the work for you.
The other possibility (and the most likely culprit) is that your keyword has been lowered from broad to phrase/exact match by Google. [webmasterworld.com...] Post 19 has the data about the phenomena.
the only problem im having right now is this, if i use the dynamic title insertion with the exact match keywords, i get some really boring ads:
widget
buy your widgets here
guaranteed to work
www.widgeter.com
so im worried that dynamic titling wont convert well until it becomes broad matched. I dont see any of my competition using dynamic titling for what i assume to be broad matched terms. But i am pretty sure that they are using broad matched terms, so i guess if i write good enough static ad copy, i can switch it to dynamic later?
and to clarify, the reason that i want to use broad match is because i'm trying to maximize traffic...no matter how many phrase/exact keywords i can think of, even 150(i know you pro's could think of more though) doesnt come close to 15% of the estimated clicks/day that google gives for the broad match. i'd prefer to use the targeted keywords, and started off doing so, but getting 50 clicks/day isn't going to work for me when something else tells me i can get 1000.
thanks again for the help,
pk
so i guess if i write good enough static ad copy, i can switch it to dynamic later?
You can make new ads at anytime, so you can also right now make both static and dynamic ads and compare the CTR/Conversions against each other.
I suggest that if you are going to advertise on 'boring' one word keywords, you put them in their own adgroup (and possibly their own campaign to control spending), as they will require different ads and a lot more negative keywords than your 2-4 keyword combos.
adgroup 1:
[widget]
"widget"
Since "widget" is a phrase match with a single word, it will act as the same as a broad match. Since it's a broad match and single word, the estimation of images (or clicks) will be high as you mentioned above.
However, since your ad is new, the "widget" as a broad match (or phrase match as you think) will behavior as an exact match. That is a reason why
yep, estimates are for the actual keywords i have, no broad matches...though as i just identified above, i'm having a prob with my phrase matches:
'green widget' doesnt show my ad in search results and i have a "widget" keyword which i thought should phrase match.
Since it is an exact match, you will get less volume than the estimation.
My recommendations:
1. Think it over. Do you want to use a single word for a broad match? In a couple of days, "widget" will turn to broad match.
2. If not. Change your keyword
3. Reduce your budget.
Does a two word phrase "blue widget" always work as a phrase match or does it also start off as an exact match while it builds CTR?
from my limited experience, yes, "widget" does act as a broad match. my campaign has a ctr of 1.8-3.0 over the last 3 days, so i think the "broad" has kicked in.
now i have the little problem of attempting to target that traffic a bit better, which is difficult with broad kw's.
the reason it occurs to me to do this is because my conversion rate is pretty poor.
i made 498 kw's and the broads account for %25 of the clicks...so that isn't going to be able to get too targeted, but...the other %75 need to be placed in 7 or 8 different groups. the thing that sucks, obviously, is that inorder to do this, i have to do it manually, and lose the decent CTR.
oh well, i guess it will come back. thats the beauty of traffic.
I always assumed a phrase match would work as a phrase match even if it only had one word.
A single-word phrase is an exact match.
Let's consider the following example:
From AdSense:
Phrase Match - If you enter your keyword in quotation marks, as in "tennis shoes," your ad will appear when a user searches on the phrase tennis shoes, in this order, and possibly with other terms in the query. For example, your ad will appear for the query red tennis shoes but not for shoes for tennis.
If "tennis shoes" is a single word "tennisshoes", the ad will appear for all queries which contain tennisshoes such as red tennisshoes, tenisshoes red, for tennisshoes, tennisshoes for, etc. Which means that it's an exact match.
Does a two word phrase "blue widget" always work as a phrase match or does it also start off as an exact match while it builds CTR?
A two-word phrase always works as a phrase match.
Note that "phrase" means more than one.
These are the matching and syntax options:
[this is an exact match]
"this is a phrase match"
this is a broad match
An exact match means that search is inputted exactly in the search string as it is in your keywords list.
For example:
If your AdGroup contained the keywords:
[widget]
"blue widget"
If the searches inputs: widget
then your [widget] ad will show.
If the searcher inputs: red widget
then your ad will not show as you don't bid on that keyword.
If the searcher inputted: crazy blue widget
then your "blue widget" ad will show.
Now, if you bid on the keyword:
widget (this is broad - no "" or [])
and my account was broadmatching this keyword,
the ad could be shown for these searches:
blue widget
widget
widgets
pink fuzzy widgets
You can broadmatch a one word keyword, you can exact match a two word keyword, and many more combinations. Google gives you a lot of control of how your keywords are displayed if you take advantage of all the matching and negative keyword options.
