Strategy for Inherited Adwords Account

Want to optimise without losing all the traffic

[HelenDev]

I have inherited our company Adwords account and have been tasked with looking after and improving it over time.

It has several campaigns with a few ad groups in each, and quite long lists of keywords in each of these, all broad match.

I have been feverishly swotting up on all the Adwords tutorials and have come to realise that great big unmanageable lists of broad match keywords is not quite the way to go!

The first things I have done is to delete any kws which had no page impressions ever, and delete very broad kws with little relevance and a high bounce rate.

The next step I would like to take is to move over to phrase and exact match rather than all broad matching.

For example I have a list that looks something like this:

blue widget
blue widgets
some other widgets
more widgets

Would I be right in thinking that with broad match, having the first two is unnecessary anyway, as it automatically pluralises things? Should I delete one of them, and if so, which?

Alternatively, should I simply add

[blue widget]
[blue widgets]

to the list? Would these override the others anyway, therefore not costing any more, and, over time, showing whether the broad match versions actually get relevant traffic anyway?

Any advice would be appreciated. I want to optimise our campaigns but I would like to try and avoid throwing the baby out with the bathwater, as it were :)

[Sharpseo]

I've had bad experiences trying to keep existing campaigns. Lots of times it's better to start fresh.

If groups have killer quality scores and are well set-up, they might be worth keeping. But it sounds like the acct is kind of a mess.

If you decide to start fresh, make sure you do it right. Separate content/search ad grps, carefully segregated kw-based grps, etc.

[HelenDev]

Thanks for the reply Sharpseo.

Unfortunately starting from scratch is not an option. The campaigns may be a bit messy with regards to keywords, but the ad groups are set up logically, and we will be using separate ones for the content network (not using it yet).

I would just like to tidy the keywords a little, and improve bounce and conversion rates for what we have. If you were faced with the simple list I posted

blue widget
blue widgets
some other widgets
more widgets

what steps would you take to improve it without losing the good part of the traffic we already have?

I am specifically interested in keyword matching options, rather than splitting the ad groups down further at this stage.

[SEMblahblah]

keep the broad matched...just add exact and phrase to the current list.
look at the conversion rates, cost/conversion and more importantly, ROI.
for starters, set a max cost/conversion limit you can hit which would ensure that you don't run into losses (money earned/conversion - cost/conversion).
for keywords with good conv rate, increase bids (keeping in mind the cost/conv limit) to gain higher positions to drive more traffic.
Similarly, for keywords with poor conv rate, cut down bids to make sure that your cost/conv is under control. this would ensure that whatever traffic you get (however little that is) is not affecting profits.

[HelenDev]

Thanks SEMblahblah, those sound like good things to try :)

Should I also worry about poorly performing keywords? In my session today on the Adwords learning centre I came across this

A poor performing keyword can affect the Quality Score of an entire ad group or campaign.

Does anyone know what they mean by this and does it mean that you should get rid of all the less well performing kws - e.g. those with no or few clicks, for the good of your ad group as a whole?

My manager doesn't object to kws with no clicks as he sees it as free exposure/awareness advertising - is this correct or could it be a problem?

[tomasvdb]

it is a problem for google as to them, it's a waste of advertising space. No clicks means no revenue for them. You can severely damage your standing in Google adwords if you have a lot of irrelevant keywords and the day that you want to set up an ROI based campaign, you're going to struggle to get your min CPCs down.

if i were you, i would pull a complete KW report (including the ones that never got an impression) and start weeding out the worst ones.

btw, use adwords editor, it is much more suited for cleaning up accounts than the web interface.

[HelenDev]

Thanks that's very interesting. I think I need to be more ruthless with deleting keywords!

My manager's current strategy for a new ad is to come up with a huge long list of different possible kw combinations, add them all, wait a month and see which have worked. The info I have read on the web seems to suggest doing the opposite - start small and build up slowly.

[T_Media]

Hi Helen,

I would say your most important priority right now is to get some keyword tracking setup.

statcounter.com is a good place to start as its free and easy to setup.

Your broad matching keywords are generally bad to have, but don't delete them just yet.

Get keyword tracking setup and you'll be able to see what exact search terms are generating clicks.

Because you have broad matched keywords you'll get a lot of terms you may not have heard of before, but you'll also get a lot of searches that you know will never convert.

The goal is to go through your referrer logs everyday and look for new search terms. Add the desirable keywords as either phrase or exact matches and add the ones you know will never convert as negatives.

You will decrease the amount of junk visitors you recieve and thus pay less, leaving your daily budget open to more of the kind of visitors that convert.

Eventually, after a few weeks, you'll see far less new search terms appearing, if you're confident you'll not see many more good terms appearing; it's time to end your relationship with the broad matches forever since they have done enough data gathering.

[HelenDev]

Thanks T-Media. I'm already using Google Analytics as my stats package and I can see all paid and unpaid keywords which bring traffic to my pages, so I don't think I need statcounter unless it's doing something particularly special.

Looking at the organic keywords is one of the things I planned to do anyway, and I suspect may be more useful than coming up with random combinations of words which no-one might ever search on.

I was thinking that it's probably not worth paying for terms which we are already high in the SERPS for such as

blue widgets ourtown

or is it?

[netmeg]

I do. For one thing, Google ads show up on other properties than just Google (such as AOL) and for another thing, I don't think it hurts us to have more of a percentage of the real estate of the page to be up at the top of the SERPS *and* the ads. And your SERPS position could also drop at any time, temporarily or permanently.

[HelenDev]

Thanks netmeg, that's an interesting thought. I guess because it's a very targetted kw phrase the quality score should be high and the cost should be relatively low.