How To De-Risk Your New Idea Using Google Adwords - Part 1 of 2 | Hacker News Reader

How To De-Risk Your New Idea Using Google Adwords - Part 1 of 2

blog.stevebarsh.com

16 points_pius9 comments

I have seen, over and over again, this idea of making an AdWords account go to a placeholder page to test out ideas that don't quite exist yet. It has routinely been suggested by people who I would otherwise consider quite intelligent and well-informed about marketing on the Internet.

Have these people used AdWords in the last 3 years?

Google is not neutral at all about what content you put up behind the click. If your page is content-free, their automated algorithms which calculate Quality Score, a key metric which determines how much you pay, are going to penalize you. If you use the landing page as a squeeze page (i.e. "sign up for our mailing list for when we release!", the most frequent suggestion) your QS will drop so low you will probably terminate your test quickly, after you start getting charged 10 to 100 times the going rate for clicks.

AdWords will also manually ban an account they catch doing this. It stands to reason for Google: they want people who click on ads to get what they expected, not hear "Psyche! That positioning was just a marketing test. Well, that will teach you to click on AdWords ads in the future, won't it?"

What would your suggestion be?

That was my immediate impression also, without the data you provide to back it up.

I guess an alternative solution would be to build some content, just release early and use this as a way to test it (because of the metrics AdWords provides) before investing additional resource in the project.

Which basically means that unless you can create a viable product with minimal input, this tip won't help.

I think you may be underestimating the amount of work, savvy, and experience it takes to get "content" to work with AdWords in a cost-efficient manner and extract meaningful business intelligence from your metrics.

Sure, it will give you rapid feedback on your ability to compete with other AdWords users. Here, I'll give you rapid feedback, too, without seeing your campaign: your campaign is terrible, compared to every other actively maintained AdWords account. That's harsh but its about as accurate as the feedback you'd get from analyzing your AdWords stats, and it has the bargain of not costing several hundred or thousand dollars to figure out.

Note that it is not exactly easy to read the AdWords tea leaves without both AdWords experience and experience in the particular market you're trying to read. If I posted my AdWords stats for the first six months, would you know how to go through the virtual squid ink of metrics to figure out whether or not those results were good? Off the top of my head, I'm pretty sure I had a 1.8% CTR at one point. Is that good, bad, or irrelevant? Here's my answer: depends or irrelevant. That is my answer for almost all AdWords metrics with the exception of CPA, and if you are actually selling something already you're pretty much outside the scope of this discussion, aren't you.

I had a profitable business for a year and change, using AdWords the whole while, before I figured out how to use AdWords in a manner that effectively advanced the business goals relative to the time I spent on it.

You're right about the content, and even more right about properly understanding AdWords metrics and finding the right information.

Let's say you even have AdWords experience - if your new idea is in a different realm to your experience, you still won't be able to interpret data meaningfully; if your new idea is in the same realm, then you should be meeting a known need not relying on AdWords to tell you whether your idea is shite or not.

Maybe I'm missing the point here, but it sounds like he's suggesting using AdWords to do some quick A|B tests to solidify a position and confirm assumptions. Sure, the low-QS page will perform poorly in absolute terms, and it will be difficult to extract meaningful raw information; but it could still be helpful to compare the relative performance of a few nearly identical ads that point to the same content.

The problem with low-QS is not just that it will perform poorly in absolute terms. In addition to various monkeying with your positioning they will do, which will distort the heck out of your assumptions you're confirming, each click will cost a punitive amount of money. My highest QS ads go for 3 to 4 cents a click. Your low QS ads, for the same keyword, will cost $1, $5, or $10. (And the sky is the limit on that if you are in a niche more money-rich than educational bingo cards. Basically, it is priced to punish based on whatever the prevailing CPC is in your niche -- if it normally cost $8.20~9.80, it now costs $100 even, etc.)

Distorted the assumptions is bad, but paying a buck a click for market research is not, especially if it saves you $1000s in dev costs.

(I still don't think I'd feel right doing it, though).

OT: How long does it take for your Google ad to start getting impressions when you submit it? Mine are embargoed for several days, and I cannot figure out why (nor will they tell me).

This idea was suggested by the four hour workweek (tim ferriss), that's probably why it's getting a lot of play.

I was thinking of trying it out for myself sometime soon, but I see from the other comments on this thread that it's probably not as much of a slam dunk idea as it sounds.

Reply on news.ycombinator.com