Regional Roi Bias Options

[arlen]

I'll start w/ the disclaimer that I've not been terribly aggressive nor terribly successful with my PPC advertising over the past year (just focused on other things). I'm trying to develop a strategy and learn what I can through testing over the winter.

During my off season I note a distinct bias for completed orders, i.e. warmer states order more / more frequently during the winter (makes total sense), and ROI is significantly higher. This bias is concentrated in a handful of states, though CA and FL have a notable lead. The bias exists in-season as well, but is much more prominent during the winter months.

Other than limiting my AW ad displays to US and Canada, I don't currently use any regional settings, nor do I really know how they work in any detail. However, I'm thinking since I can see such a distinct pattern, I should be leveraging this to my advantage.

One approach would be to develop a separate set of campaigns specifically targeted to CA, possibly another for FL, and be more aggressive on these campaigns than I am on my non-regional ones. Though I'm not completely clear on the how at the moment I'm sure I'll figure it out, but conceptually this makes sense to me, though I'm not located in either. Would you agree this is a reasonable response and are there issues with targeting a region other than your own?

Are there any other approaches I should be considering to leverage this regional bias?

I should probably think this thru a little more before asking, but I'd appreciate any input from successful PPC users.

[Alan Perkins]

It makes perfect sense, arlen, but it is a lot of extra work.

ROI is significantly higher

How do you know this? You know sales are higher to warmer states, but do you know how much you spent on clicks from warmer states vs. clicks from colder states? Colder states may generate far fewer clicks, but the ROI could be better (even though the sales are lower).

[arlen]

I'll have to do a real analysis of my stats Alan, but in reviewing my logs (and ip addresses) my impression has been that a higher percentage of CA visitors actually place orders than those from other states ... like the rest are more likely to be window shopping. I'm still having a hard time putting all of this info into a form I can base good decisions on, I'm far from a statistician. Also, my sales are limited to a few a day, so it's hard to make real conclusions in a short time span. And I may be wrong, it may just be that I have a higher number of CA visitors and CA orders, but the ROI isn't as different as I think it is. I'm probably jumping to conclusions because my most successful campaigns have largely been dominated by orders from CA. For Christmas gifts, that probably makes sense due to the weather and may not translate to the rest of the year, though CA orders are always higher than other states.

... but it is a lot of extra work.

That's good to know. Before I waste a lot of time, I'd better be sure I'm working from solid stats and not just perceived patterns. Thanks

[Jbrookins]

The nice thing about regional/local campaigns is that it splits that traffic away from your national campaigns.

If you think you're doing really well in California, it would probably be worth your time to launch a California campaign. Not only can you react to that region's performance without impacting the rest of your campaigns, but if it doesn't WORK, it will answer a question without a huge time investment on your part.

The question in my mind is "What will take the larger investment? In-depth log file analysis or building a test campaign for one state?"

[arlen]

The question in my mind is "What will take the larger investment? In-depth log file analysis or building a test campaign for one state?"

lol, if you knew how overwhelmed I am with stats and exactly what conclusions to draw from them, I think the answer would be relatively clear. On the other hand, this is an area I need to get a handle on before my high season arrives, so it's probably a worthwhile endeavor.

[Jbrookins]

lol, if you knew how overwhelmed I am with stats and exactly what conclusions to draw from them, I think the answer would be relatively clear. On the other hand, this is an area I need to get a handle on before my high season arrives, so it's probably a worthwhile endeavor.

I'm a huge fan of testing. Test your theory. Don't think that will hurt you, and if it doesn't work, you can always just shut it down.

[jehochman]

You can use a tool like Google Analytics to check your overall conversion rates for different states. You can sort from highest to lowest conversion rate. If you see some states are much better than others, you can use the Adwords Editor to quickly copy your existing campaigns to make new campaigns that are geographically restricted to your hot zones. You then turn off your national campaign, and run regionalized campaigns instead. You could do that for a few weeks and see how it affects business.

Be careful not to slice your markets too finely or you'll have such low volume on each that it will be hard to spot trends and take an overly long time to optimize your campaigns. Unfortunately, there's no easy to follow rule about how to do this. It takes experience. You have to dive in and try it.

[arlen]

Thanks Jonathan,
I haven't delved into Google Analytics (as I hadn't been using my current stats to best advantage yet), now I have a reason to.

Good advice on not splitting markets up too much. My volumes are still low enough it's hard to be certain I'm seeing what I think I'm seeing. I've got a couple years of stats to work with, but so many changes have been made, I hesitate to look back too far.

<edit>The theory is worth testing ... but I think I'll wait till after Christmas to do so.</edit>

[jehochman]

You may not need Google Analytics. Whatever you have should be able to handle geo location.

[redsonia!]

arlen:

Just to let you know: I seem to have a situation similar to yours. I also get my largest number of orders (by far!) from California and Florida. I hadn't thought of doing campaigns to target those states though. So that gives me something to think about.