- Adwords
- 2009-09-27: Marketing by numbers: I'm running some pay per click ads through Google Adwords. If you're not familiar with that, if you look on the Google search results page, usually the top three results, and the right hand column of results, are paid ads.
- Businesses pay each time you click on one of those ads. How much they pay depends on demand for those spots and on a formula that tries to ensure the ad is relevant to the search query.
- Remember Google is nothing if it doesn't provide relevant search results. So Google doesn't want people advertising weight loss products when we're searching for a new car, for instance. It manages that by pricing those ads out of profitability.
- One of the factors Google takes into account when setting the price we pay for the position we want is the click through rate (CTR) which is the number of clicks the ad received divided by the number of times it was displayed, expressed as a percentage. Ad appearances, displays, are called 'impressions' by Google. So if your ad appears 100 times and 1 person clicks on it, you have a 1% CTR.
- Clearly, what you say in your ad affects the CTR. If your ad says "our product's really boring, I don't know why I'm advertising it" no-one will click, whereas if it says "you won't believe how much weight I've lost / how well behaved my dog is / what my house is worth" maybe they will.
- So, rule number one in marketing is "everything is a test". Adwords encourages you to create alternative ads to run alongside each other. So we can write two ads and see which one has the best CTR. When we decide that we can ditch the worst ad, and create a new variation based on the better ad. That way, we gradually improve.
- Well, I just checked an ad I've been running for a week. I've been slowly improving the ad for a while and all's been well, but this week my new ad got a CTR of 3.27% against last week's winner which only got 0.71%. So last week's ad tweak provided a 4.6 times improvement. Wow.
- Testing. You never know what will work, but the rewards are there when you finally stumble upon it.
- And .. I've gone through maybe seven different ad variations to get that. For me that underlines yet again that while clearly it's good to make a good first stab at a website and its text, it's not really what you the client and I the consultant decide would be great on your page. It's for the users to decide. All we have to do is give people the choice, and listen.
Next: Kenneth Williams -
Previous: Our kettle lasted .. -
Back to blog list
Tweet this blog 
Share 
Bookmark