I’ve written a couple of posts on the Analytics Blog about the eMetrics Summit I’ve been attending in San Francisco over the last few days. Although the conference has a lot of analytics content it is very much about how we manage business success. By deciding what success might look like, you can then take steps to understand what your customers are doing before, during and after they come to your site.
Now pay-per-click on its own serves as a way to get users quickly and efficiently to your site. You bid on relevant keywords, write relevant ads, you get a good click-through-rate and traffic starts coming to your site.
But there’s much analysis you can do before you even kick off the campaign to ensure that you have optimised it before you even pay a penny.
Use the Research Tab in adCenter or the Excel Add-In to do some really thorough keyword research. By thorough I’m not just talking about generating similar KWs or knowing how much traffic you might get. I mean diving a little deeper into the data. Have some goals in mind. You might want to sell more products or you might want to generate more leads. Whatever your goals, make sure you research your keywords with any business limitations in mind. For instance if you only deliver to a certain geographical area, make sure you know what the geographical make-up of that KW or KW group is. You may find that most people looking for that product or service live too far away, so then you have a couple of choices. You can exclude that geographic region when you set up your campaign or you can take steps in order to enable your business to deliver to that location.
Checking out traffic trends over the week can be very helpful too. Some KWs may get more searches at the weekend. Now your business might depend on a real conversation with a real person to close a transaction and your office may not be open at the weekend. In order to limit potential negative branding, you could turn off your campaign at the weekend, or alternatively as I talked about a couple weeks ago, you could get more creative with your ad copy.
Finally, seeing which age and gender bucket your KWs fall under can be very interesting because the report sometimes throws up some surprises which you can latch onto and capitalise on. If you know that a keyword’s demographic data is biased towards women aged 18-24, your ad copy and landing page text would need to be very different than if you discovered the KWs you were about to bid on were being search for predominantly by men aged 65+. What you think your product's target demographic is might be very different from the demographic that actually buys it. A new jewellery design for instance may be targeted at women, but it may be men who are looking to buy it for their wife or girlfriend as a present.
Thinking laterally and doing some in-depth research and analysis before you start building a campaign can help reduce cost and improve initial return on investment. By setting some business goals and giving yourself the best possible chance of achieving success right out of the gate, you're helping to positively build your own brand, as your customers will receive a much better user experience.