Using Adwords

[Brian Turner]

I'm experimenting with Google AdWords, and thought I'd detail the sort of things I'm experimenting with, and report on what seems to work and what doesn't.

In accordance with the Google AdWords Terms of Use, I would be able to detail keywords used, but I can describe methods.

Ads Targeting

I've already decided that my target customer is UK businesses. So I've selected UK as the only region.

I appreciate that geo-targeting is pretty rough - I think I've seen it suggested that targeting region by IP has about 80% coverage - but the aim is to keep costs down at first while exploring a specific target market.


Campaign Organisation

I've got a couple of different campaigns at present - one for my business site based on keywords on search only (no content match) - plus a non-profit site I run, which has a small budget for a keywords campaign and sitematch, both as separate campaigns.

I'm following this arrangement simply because I run both sites and wish to promote both - however, I've heard it suggested that increased CTR affects the entire account.

It'll be interesting if this means good clicks on the non-profit site will boost the commercial keyword rankings for my business site.


Ad Groups by Keywords

The business site has three ad groups, each targeting a single keyword root.

This will hopefully allow me to track:

1. How well each specific keyword delivers visitors
2. How well each keyword group converts to sales

Of course, I've also written the adcopy to reflect the keyword targeting, and also tried to include a couple of additional keywords for highlighting in the ad on search.


Broad Match, Phrase Match, Exact Match

In each Adgroup I've created a keyword list - all two to three keywords based around the delivery of a service, rather than the service itself.

I've also set up every keyword as both broad match, phrase match, and exact match, which is done like this:

keyword 1
"keyword 1"
[keyword 1]

This is so I can track whether specific match targeting of any specific keyword works better in some areas better than others.

Of course, some are likely to face deletion if they have low clickthrough rates, but I feel a need to start comprehensive, and see which keywords and targeting formats actually work well for clickthroughs.


Campaign budget and CPC

At the moment I've determined that a monthly allocation of a few hundred pounds will allow a basic test-run, to see what works.

As I'm simply exploring the Adwords system for the first time, and need to find out what works, I figure a few hundred pounds is high enough to provide at least basic data, but not so high as to get burned if I accidently snag high-cost keywords that delivers little.


Tracking

At the moment I've not implmented proper tracking for conversions.

However, my plans at present are:

1. To implement tracking stats software, to track user behaviour on the site.

I'm not sure what I'll use yet, but I experimented with a stats package a few years ago which was pretty brilliant, but above my requirements at the time - think it was Webtrends.

2. To track conversions, I'm planning to set up landing pages accessible only via Adwords clicks (not accessible to search engines or human users on the main site).

This will allow me to create focused adcopy, and make it easier to track user behaviour from clicks.

To track conversions, my plan at present is to set up e-mail addresses customised for each landing page, so I can see which pages are generating the most e-mail enquiries directly.


Tips

Tips that come to mind already - when targeting UK users, have .co.uk in the domain display URL - but when targeting international customers, have the .com form for the domain in the display.


Anyway, see how things progress.


If anyone has their own tips and suggestions, feel free to post them on this thread.

[Brian Turner]

Well, I'm informed that CTR only applies to keywords, not the whole account, so that seems to be that issue settled...

As for the campaign itself - Britecorp itself has spent around £100 without generating a single lead so far.

Certainly I'm not using optimised landing pages, which is bound to affect conversion, and the keywords I'm targeting may be wrong.

Either way, though, it tells me that customer acquistions via Adwords alone is not - so far - going to be as cheap and easy as I may have originally considered.

On the plus side, I've been advertising on the content network for branding purposes - I've found it's a relatively cheap way to get 10,000's of impressions, which I'm using for branding purposes.

While it's hard to calculate ROI on branding, it's definitely a strategy I'm happy to pursue so far.

In that regard, I've found sitematch to be a less effective way of branding - more expensive, and more limiting on the branding exposure. I guess the trade-off is simply that you can be much more focused on a particular customer audience in that way, and something I'll experiment with again.

[Robert]

Hi Brian!

How is your Adword going now?

Robert

[Brian Turner]

I actually stopped - found I'd spent a few hundred pounds without getting a sign-up.

As an experiment in using Adwords I'm glad I ran it - preparing to set something up again very soon - but I already knew my websites failed to sell the company services quickly (ie, Seth Godin's "banana") so when I run it again I'll make sure I have something properly prepared on all sides.

[rtroxel]

Don't forget that TIMING is very important in an AdWords campaign. I spent many $$$ on a campaign during the autumn of 2004, and didn't make a cent. However, had I waited until Christmas season (in the US, it's from the end of November to Dec 25), I would have at least been even.