HN Explorer

Ask HN: I tested an idea with AdWords, are the result good enough to work on it?

10 points by gws ↗ HN
I wanted for a while to experiment with testing ideas with Google AdWords so I finally did it going a bit meta: I tested the idea of a service that let you test ideas (landing page at http://unbouncepages.com/test-your-idea/). I followed the textbook: made a quick page describing the problem more than the solution, set-up a Google AdWords campaign, leaned back for three days and watch the results (I did do some tweaking of the AdWords campaign after the first day but it didn't really change anything).

So here are the results: in three days I got 245 visitors (all from ads click-through) and 27 conversions (people that left their email) for an 11% conversion rate. I spent $127 for a CPC (cost per click) of $0.64 (in particular $0.94 for search terms and $0.57 on the display network, that is the ads on third party web sites). The CTR (click through ratio, how many click per ads diplayed) was 0.07% (0.62% for search terms, 0,06% for the display network). Interestingly, ~75% of the clicks came from the display network, unfortunately I was an idiot and cannot track conversion from display vs. search (hey, it was my first AdWords campaign!).

So the big question for you guys is: are these numbers good enough to say, yeah, I'll go work on this? Or more in general, what are good numbers for an experiment like this? Thank you folks!

7 comments

[ 2.7 ms ] story [ 33.5 ms ] thread
tocomment
I'd say go for it. It sounds like a cool idea.

Maybe you could market on forums like the four hour workweek?

bkanber
What you haven't yet told us is: what's the lifetime value of a customer of your service?

Your stats above show an acquisition cost of $4.75 per user. That feels a little high, but you also have to consider the "landing page vs real product" problem. You may have gotten a higher conversion rate if you had a real product launched. You also don't have any experience yet advertising your product, so we can assume that the user acquisition cost will go down because of that as well.

So considering all that: what's the lifetime value of a customer? $10 or $100? If it's closer to $100, then even an acquisition cost of $5 would be great, but if you end up spending $5 to get a $10 user then either a) your idea isn't that compelling, or b) your marketing is suboptimal (edit: or c-- your product or industry demands that type of return, which probably isn't the case here).

So while I don't have "the answer" for you, you should start thinking about lifetime value of a customer and what you expect your user acquisition cost to become 6 months after launch. Weigh those two against each other and you'll have a good idea of whether this is worth pursuing or not. Best of luck!

mapster
Ok. So it cost you $127 to get 27 emails. Before you ask HN (is this good enough), you should ask these 27 people if they will in fact buy. Then you will have a sense if there is a market. Now it is pure conjecture. Be bold and ask. If you get mixed responses, ask them if you could interview them. Then pivot as necessary and iterate. Only your customers can tell you if the idea is good enough. You want their 'yes, I will buy" and not Adword stats to tell you if you should build or not. Best of luck
djsamson
This is a good preliminary test, but more customer development needs to be done. You need to have problem and then solution interviews. Don't just show them your solution and ask "Will this help?" Of course it will to some extent.

You want to get a really good understanding why these people have this problem and how they've tried going on without your product. Ask them how much they would be willing to pay to get their problem solved. Then ask if your solution is helpful and show them a demo.

Steve Blank is the Godfather of customer development. I highly suggest the Startup Owner's Manual. But if you don't have time to read that all right now, just realize you need some in-depth interviews (preferably in person interviews) and then use the Google Adwords landing page test as a mass scale test.

Good luck.