One way we (Zapier) got early users in the door was by trolling forums. We would manually search the internet for people complaining about integrations that didn't exist (eg. http://webapps.stackexchange.com/questions/13007/how-can-i-a...). We would follow up directly, build the integration, and instantly have a new customer. Clearly this doesn't scale [or does it?] but it put us on the right track early when we weren't sure what people would respond to.
This is one of the first genuinely original ideas that I have seen in a long time. You can create a site, use AdSense to drive traffic to it to see if your idea has merit.
This one little additional step lets you actually get into conversation with people visiting your site before you launch is very much lean startup.
Now as a developer building the product I can hear the objection you won't be able to get the product built if you keep getting interrupted. But you may be building the wrong thing. So this makes a lot of sense to me.
> You can create a site, use AdSense to drive traffic to it to see if your idea has merit
I started a number of sites like that years ago in college, with nothing but a static site built with a cheap template in a day, and some AdWords credits. If people actually signed up to learn more, or ordered a product, or whatever, then I'd spend the time building a real business around it.
The first time I saw that written down was in Four Hour Workweek several years ago. Now millions of people have been told to do the same.
We tried to build functionality as fast as users were requesting it. It seems like the AdSense approach will turn a lot of folks off once they realize it's only a landing page. Clearly valuable in determining if an idea backed by marketing can get traction.
But, how do you convince those AdSense users to share their true problems (or do you just target a really specific keyword niche) before they bounce?
It seems like the AdSense approach will turn a lot of folks off once they realize it's only a landing page.
I'm not a big fan of AdSense smoke testing, but if you're going to do it, you can fairly trivially raise perceived value above "landing page." Offer an immediate incentive for giving their email address.
For example, if you're going to offer software to automate X but haven't built it yet, offer to give people their own personal copy of your guide to X-ing (written by an X-ing expert). Just put in the email address we should send it to. P.S. Want to hear about our thoughts on X every week or two? Make with the clicky-clicky.
I know it's not the best of 'business books' but "The 4 Hour Work Week" has a great section on doing just this - I found it really useful in thinking about the very start of a MVP.
Where did they find twenty-five thousand people who consented to participate in their market research? I'm having trouble convincing myself they aren't spammers.
If you were talking to a non-technical marketing person, and he asked you "Hey, can you tell us how many of our customers we haven't talked to in 3 months", and you said "4,386 out of 12,381" seven minutes later, and he accused you of lying to him because it is inconceivable that you could have counted to 4,386 in seven minutes, how would you feel?
You and I both know that computers count very quickly indeed, that SQL joins are possible, and that seven minutes was probably 6:30 of looking up an API call or DB structure, 30 seconds of typing, and 100 miliseconds of waiting for the query to execute.
Here's something that you may not be intimately aware of: people who are competent with Internet-style marketing can use "computers count to big numbers really quickly" with permission-based marketing approaches, like e.g. setting in motion an SEO strategy which passively attracts almost arbitrarily high numbers of visitors. Bingo Card Creator -- pretty freaking niche -- has hundreds of thousands of people sign up every year, and involves no spamming.
Yeah, that was more confrontational than I wanted to be, sorry guys. If thousands in the niche exist and are reachable in a permission-based way, kudos. That's still kind of mind-blowing. I guess I was primed to expect bad behavior after seeing a reference to cold-calling.
Not even that - the article only claims "25,000 messages". Multiple messages in the same conversation, as well as multiple conversations with the same person. The number of _people_ may have only been in the high hundreds…
The percentage isn't that great, but we usually aren't the ones initiating conversation. The quality is definitely what drives us to continue using it.
Olark has a free trial. When we first put it on our site we were pretty hesitant too, but after the trial we were sold.
