I’ve been quite busy since last time, testing three new product ideas without committing to any of them before they were proven. Two bombed, but one of them has been successful enough that I’m moving into the next stage. I built a more robust Adwords campaign to test multiple variables and teach myself how to use this resource more effectively. Along the way, I found a few articles online (especially this one) that were particularly helpful, in addition to the guides within Google.
Here’s what I did to move into the second stage of Adwords testing:
1. I categorized keywords and split them into groups. For example, since I am working on an apparel product, I have a group for the material, one for the method of manufacture, one for the unique design, etc. I also made a group called “general” for all keywords that don’t fit into one of the other categories.
2. I created a unique ad for each group. This way, I could try to tailor each ad to the user’s search.
I saw some improvement in my statistics after implementing these changes, but it still wasn’t ideal. After a few days, I did the following:
3. I added more keywords in each category. Specifically, I added exact match versions of most of my existing keywords.
4. I created more ads in my top three categories and set the campaign to serve all ads equally to test the different copy. This was the most fun part, because I could be creative and try lots of different combinations in hopes of hitting a winner.
After making these last two improvements, I did see some interesting trends, most notably in the different ads. I only have a few days worth of data right now, so I can’t say anything definitive, but one of my ads is showing a clickthrough rate over 2%. I can probably improve this further by weeding out all of the underperforming keywords, but I’m not ready to count anything out just yet.
Overall, I’m relatively pleased considering that my landing page design is mediocre at best (I don’t actually have a product to sell yet) and I’m very new to Adwords. Still, I need to continue improving my stats and lower my costs. I think I’ll see some positive changes when I remove the bad keywords and again when I flesh out the landing page. I’ll also try some different ad copy to see if I can beat my current numbers.
This is just starting to get exciting. Hopefully I’ll have a prototype to show soon.