Do some US states convert better than others?

Using Adwords to target at the state level.

[lgn1]

We are launching a new campaign in 2012 for the American market, and we want to pick 10 states for testing, balancing to achives 20% of the market by population. It would be far to costly to test on the entire US market.

We are looking at the young urban professional female as our demographic.

Rather than just pick 10 states at random; do some states perform better for ecommerce than others?

I'm not worried about a few tenths of a point; I'm worried about major differences.

So excluding Alaska and Hawaii or obvious reasons, and excluding any seasonal or other influence's (ie selling snowboards in florida); does any body have evidence that geographic unbiased products would convert better in one state rather than the other?

[buckworks]

Yes, there can be significant regional variations in conversions for the same product.

I don't just mean raw numbers, but higher per-capita conversion percentages.

You'd expect to sell more in Wisconsin than Wyoming, simply because there's more people, but if your widgets were twice as popular among Wyoming users, you could afford to bid more in that area.

Regional responses could vary for many reasons, such as economic ups and downs, seasonality, local fashion trends, natural events, previous familiarity with the products, and so on.

Round up whatever facts you can that would help to give you clues. Then make your best guesses and fine-tune as you go.

Watch for conversion sweet spots, and consider targeting those with their own campaigns and more aggressive bids.

Remember to set your demographic preferences to focus on females.

[RhinoFish]

you didn't give us enough information really.

1st question, are you talking about Search or Display advertising? (female demog can be targeted via keywords on Search to some extent, so i'm not sure which you mean)

state breakdowns won't lead you to urban necessarily. pick city targeting instead. or here's an idea, target tablets.

anyhow, your question is an hour long discussion that depends on many things. give us more relevant info, and we can do the same in return.

happy hunting!

[Planet13]

There are many, MANY demographic variations from neighboring cities in the US.

For example, the demographics for the California cities of San Francisco and Oakland are entirely different, even though when you look at them on a map, you just see that they are separated by a bridge across a bay.

In fact, Oakland and Berkeley, which are right next to each other, have significantly different demographics.

On the other hand, San Francisco and San Jose have more similar demographics, even though geographically they are farther apart.

I hope this helps.

[lgn1]

Sorry, I can't give more detail, without specifying what business I am in; and the great thing about this forum, is by keeping our anonymity; discussions can be very open.

I figured by the time I got to the state level, sales per capita would begin to average out.

I suspect that economically depressed states, would fare the worst; however we sell our goods at a modest discount, so this may work in our favor.

With every business being different, the top/bottom 10 state list would probably be different.

I guess the answer is to start a campaign, with what I believe is the best factors, and then modify the campaign as soon as I have a statistcally significant sample size.

Eventually I can start fine tuning to specific urban areas.

Looks like my multivariate and factor analysis course will be put to some pratical use.

[LucidSW]

> I suspect that economically depressed states, would fare the worst

There you go. But would they use computers and have Internet connections? Probably not. But even in those states there are people who would need and use your product. You just don't know so normally, I think it's not a good idea to target only a small percentage of the population.

If you want or have only enough budget to target 20% of the population, get a good sample of the general population and target the states with the top 20% of the population or a number of states that come to that percentage. Or target cities instead of states. Target big and small cities, east/west, north/south. That should give a good sampling.