| 7:10 pm on Sep 27, 2004 (gmt 0)|
|It seems like a common problem of losing keywords due to low CTR. I am wondering what is important to maintain a LONG term adwords campaign? |
Judging by the number of threads revolving around this topic, it would seem like a common problem. But really, IMO, it isn't that common at all, at least amongst more experienced advertisers.
I'd say that, by far, the majority of AdWords advertisers seldom lose a word to low CTR - at least after the first few weeks.
What are those advertisers doing right? They are probably being very specific and very targeted.
To be more specific, they are most likely doing some or all of the following things:
* Avoiding very general keywords
* Avoiding single-word keywords
* Avoiding 'What If' keywords. (Example: What if someone searching for the address of Carnegie Hall (your keyword), might decide that what they really want are tickets for a rock concert in Seattle? (what you actually have on offer).
Instead, they are probably:
* Using multi-word keywords that directly relate to their products or services
* Writing excellent ads that are about the exact same thing as the keywords that cause the ads to show
* Using many tightly targeted Ad Groups to allow for the above
* Using negative keywords to prevent their ad from showing to folks looking for something else
I hope this helps a bit - and I hope others jump in with their very informed $0.02 worth. ;)
In my opinion though, it is all about being targeted, rather than general.
| 8:56 pm on Sep 27, 2004 (gmt 0)|
I'm no expert, but having run several campaigns recently, adwords works in situations where typical consumers are likely to go searching for a product/service, otherwise it can be difficult.
I've got a possible problem with a recent campaign - the ads are fine, click through rate ok, but at the end of the day the site doesn't really have what the average consumer is looking for ( it's a trade-only site serving larger companies ). I haven't given up exactly, because it probably only has to garner a new customer once every few months to pay off, but it's difficult to measure effectiveness. Maybe adwords isn't the best for this company.
By contrast, another site which advertises a basic local service that ordinary consumers are likely to be looking for is clearly working very well. The spend is very low (about £10 so far!), but it has already produced many conversions, sometimes even without a click (telephone number is in the ad!).
| 9:43 pm on Sep 27, 2004 (gmt 0)|
List keyword with 'all time' setting.
On keywords with 100 impressions or more give them the axe. 0.5 CTR or lower give them the trash can.
| 10:35 pm on Sep 27, 2004 (gmt 0)|
AdWordsAdvisier - you say
"Instead, they are probably:
* Using multi-word keywords that directly relate to their products or services"
I agree and am trying to implement on this on my campaign which I started using broad keywords. But what I am finding is that my targeted keyword phrases are being assigned a very low initial CTR by Google and due to the relatively low number of impressions (5 per day is typical) Google is not moving over to the real CTR which in most cases is over 20%. The result is that I am losing position (2 down to 8 is typical) by adding a targeted phrase to an adgroup with an existing more general phrase. How many impressions does a new keyword phrase need before Google will rank it using its real CTR rather than a notional one. Also what CTR does Google assign to a new keyword phrase. I have heard it is the average for existing bidders on the same phrase. But what if there are no existing bidders? This may sound hypothetical but I operate in a very niche market and I honestly believe (and teh Adwords estimator tool supports my view) that there are no other bidders on some of my keyword phrases. Any help appreciated.
| 4:05 am on Sep 28, 2004 (gmt 0)|
I've asked for the specific CTR assigned to new ads and got the 'booga booga! very complicated' answer. So don't hold your breath. For keywords that are in danger of gettting axed, there are only three situations that I can think of:
- keyword too competitive. I can't afford to be competitive, so I end up on page 900. If that's the case, I let the keyword die
- keyword is not targetted enough. That's obvious to me because I'm in a niche market and sometimes try 'nearby' market keywords. If my results are poor for advertising on 'basketballs' on the term 'nike basketball shoes', I give it the axe.
- if it seems targetted, and it's not too competitive, I don't let the keyword die - even if it's small traffic. Keep trying different creatives, different and more targetted ads. Copy what your competitors ads are doing etc. 500 decent performing low traffic keywords performing well can make a nice little campaign.
| 7:59 am on Sep 28, 2004 (gmt 0)|
Why not bid high for a short time if there are many competitors. Just to boost your CTR. Then after this short period decrease to 0.05 or whatever is worth.
| 5:33 pm on Sep 28, 2004 (gmt 0)|
You can bid as much as you want, but if your title isn't targeted to the user's search query, your CTR will be crap. There are to many people out there taking the time and effort to write descriptive ads for keyword combinations rather than having 1 ad for 1000s of keywords.
| 11:30 pm on Sep 28, 2004 (gmt 0)|
|You can bid as much as you want, but if your title isn't targeted to the user's search query, your CTR will be crap. |
No, no, no!
It's your ad that counts; the message that it sends to the people searching.
Your ad is either relevant to them or it is not.
The headline is the emphasis to the body copy. If your body copy is... to quote.. "crap", then your headline will make no difference. People do not look at the headline and blindly click through.
Would you click on your own ads solely based on the title? I doubt it
The keyword in adwords is relevancy!
BTW: I've just copyrighted the phrase "The keyword is relevancy..."..;-)
| 11:50 pm on Sep 28, 2004 (gmt 0)|
Disagree. I have changed body copy in my ads to "Is anyone really reading this?" and the CTR was nearly the same. The title is what matters. Period.
| 2:00 am on Sep 29, 2004 (gmt 0)|
Sorry AWA, but I must split some hairs. In some cases one can break some of the general rules you suggest.
|* Avoiding very general keywords |
I've got some campaigns doing great -- really, really great -- on very general keywords. Other, however, flail at it. It's hard to know before you test them. Test them. If they fail, cut them quickly. A lot depends on how strong your offer is, how weak your competition is, and how good your keyword research is.
|* Avoiding single-word keywords |
*Relevant* single-word keywords can do great. Some can be really huge. The key is be *relevant*.
| 7:52 am on Sep 29, 2004 (gmt 0)|
Read my old thread about using targeted ad.
| 9:12 am on Sep 29, 2004 (gmt 0)|
|I have changed body copy in my ads to "Is anyone really reading this?" and the CTR was nearly the same. The title is what matters. Period. |
That could be interpreted as a call to action - I'm surprised CTR didn't actually go up!
| 4:52 pm on Sep 29, 2004 (gmt 0)|
|Sorry AWA, but I must split some hairs. In some cases one can break some of the general rules you suggest. |
No apologies required, cline, especially since I quite agree with you! ;)
I guess I should have qualified my 'rules' as being a way to get off to a good start. Then, once a solid foundation has been established, it's not a bad idea at all to see what creatively breaking the 'rules' can accomplish.
At least if careful monitoring is part of the plan!
|The key is be *relevant*. |
There you go. AdWords summarized in five words.
(I just had a hilarious mental picture of a 100 page book, called "AdWords Success", authored by cline. On the first page appear the five words above. The other 99 pages are blank, and provide space for notes by the reader - as they experiment and learn. Hmmmm.)
| 8:02 pm on Sep 29, 2004 (gmt 0)|
|Why not bid high for a short time if there are many competitors. Just to boost your CTR. Then after this short period decrease to 0.05 or whatever is worth. |
Bidding high does not increase your CTR - a targeted advertisement does.
I'd advise against this approach, or at least make sure you cap your losses at a number you can live with.