The Power of Social Media

…oh, and why you want to be ready for Scott Hanselman love

This afternoon I was sitting in a meeting discussing a potential upcoming marketing campaign to target developers.  It’s a great concept, and I am looking forward to seeing it come to fruition.  However, during the course of the discussion, it became clear that ad buys would be the primary driver of traffic.

My chief concern was that the developer audience uses ad blocker software at a higher rate than the general population.  We don’t have data on it, and no one was sure how to get it.  I asked the community at Hacker News, and was pointed to this posting about the downloads of AdBlock Plus.  That was a good start, but to really get an answer about which we could feel happy I came up with this hack to get quick and dirty numbers.

This is the part where I learned the following formula:

Scott Hanselman Influence + Survey Monkey + Basic Account = Fail

To get started, I created a survey up at Survey Monkey.  Unfortunately, the basic account has a cap at 100 responses for any survey.  I pinged Scott once I had the bit.ly link set up, asking for help from his tweeple.  I set up the bit.ly link so that I could track clicks on the survey, versus just knowing how many people took it.  Within, and this is not a joke, a few minutes, I had hit my cap.  Whoops.  Within the time it took me to get a credit card into the system, I lost about 150 to 200 clicks.  I would never have known about the survey being closed had @BrianGorbett been on the spot to point it out.

If you happen to work at Survey Monkey, here’s a potential feature.  Don’t close the survey.  Continue to collect the data, but only show me the first 100 responses.  Holding the data captive is a far more likely upgrade path than me coming in after the fact and deciding to upgrade.  I will likely be happy with 100 responses, and probably wouldn’t have known about the survey closing to so many others (especially if Twitter streams come into play).  I have no idea how many of their surveys hit the 100 response limit, and what their convert ratio is on those basic users to upgrade to premium, but my need was now, not tomorrow.  I don’t know when I will need another survey, and had Brian not pointed it out, I would never have known, and wouldn’t have upgraded my account.  That would have been $19.99 lost dollars you didn’t get.

The survey is still open, but as of this writing, I had 991 clicks, about 12 retweets, and 584 survey takers.  Awesome!  It was a pretty simple survey asking what type of developer you see yourself as, and whether or not you use ad blocking software.  I asked people to categorize themselves as “web app developer”, “enterprise developer”, “different kind of developer”, or “not a developer.”  The result is that 55% of the time, people are using AdBlockers.

  Use AdBlock?
No Yes Grand Total
Total 44.86% 55.14% 100.00%

More surprisingly was that it almost didn’t matter how you classified yourself, the percentages were pretty consistent.

  Use AdBlock?    
Type of Dev? No Yes Grand Total
A different kind of developer 40.35% 59.65% 100.00%
A web app developer 46.03% 53.97% 100.00%
An enterprise app developer 46.28% 53.72% 100.00%
Not a developer 43.33% 56.67% 100.00%
Grand Total 44.86% 55.14% 100.00%

Looking at a breakdown by self-identified type, there are quite a few “enterprise” developers out there.  More than I would have thought given the audience of the tweets.

  Use AdBlock?    
Type of Dev? No Yes Grand Total
A different kind of developer 17.56% 21.12% 19.52%
A web app developer 44.27% 42.24% 43.15%
An enterprise app developer 33.21% 31.37% 32.19%
Not a developer 4.96% 5.28% 5.14%
Grand Total 100.00% 100.00% 100.00%

So there you go.  Updated: Total time from survey monkey account creation to having this data was a couple of hours.

If you are thinking about using ads to drive traffic to a marketing campaign, you will want to consider what percentage of your audience is likely to never see the ads, and therefore how your results will be skewed.  I was surprised that for techies (since Twitter drove most of the clicks) were as likely to to have ad blocker turned activated regardless of whether they identified themselves as a dev or not.

  • ML

    Gee… why doesn’t SurveyMonkey host an unlimited number of surveys with unlimited responses for free and hope people will pay to see them?

    Kind of a lame, but typical Microsoft response when you have unlimited resources.

    Real world web developers don’t look at business models this way.

  • Brandon Watson

    I’m really sorry you think that this is “typical” Microsoft. I spent the last 9 years in the resource constrained environments of startups, and go out of my way to continue to operate that way.

    My suggestion for Survey Monkey was based on a thesis which is easily testable. For a certain period of time, they could continue to collect responses, and they present to the survey customer the option of paying for access to the results beyond 100. Perhaps they have already tried this and it didn’t convert very well. What I do know is that they didn’t email me to let me know that I was getting more than 100 responses. There was no upsell. That, to me, is a wasted opportunity.

    The processing power to handle additional surveys is marginal, and the data that is being generated is, or might as well be, free, since it’s nothing but text, and small bits of text at that.

    Your comment that real world developers don’t look at the problem this way is a bit disappointing. I would hope that all developers would look for ways to continue to improve on their monetization strategies so that more apps could be developed, and more knowledge could be shared with other developers. Freemium is a great concept, but identifying upsell points is the hard part.

  • ML

    No email alert at 100? Well yeah, there’s room for improvement there. I would expect that much.

    But pain is a good conversion motivator 🙂 Does it make sense for you to upgrade to their Annual Pro offering now?

    If not, why?

  • Brandon Watson

    Pain is a great conversion motivator. Without an email alert, I didn’t even know I had a problem. I upgraded to the monthly account, for which they have overage charges of $0.05 per reply. For now, that’s all I need, since I don’t have a recurring need for surveys, but who knows. It was so terribly effective on this go round, I may put together some more.

    I can’t think of a reason not to offer this on basic accounts as a way to promote upgrading. Perhaps they tried that already, and it simply didn’t convert well. Good question though.

  • I’m just surprised that they actually charge that much for those surveys. Wow.

  • Well, I don’t know about Survey Monkey, but that’s really good advice for us at FluidSurveys. If you have any other suggestions like that, let us know and expect them to implemented shortly!

  • I did the survey, and selected no Ad Block _software_ because (at most) I use a HOSTS file that blocks ad domains. Since your question was about ad block _software_, I answered NO. Feel free to switch me to YES if you feel that’s more accurate.

  • Brandon Watson

    Aydin, I hadn’t heard about your product, but will certainly check it out. At a minimum, you should know what your user funnel looks like and optimize to get to a sale. I gave a pitch last weekend about product management for hackers at the BizSpark sponsored Startup Weekend. I talk about the user funnel a bit more in there:

    http://www.thefailingpoint.com/2009/08/gettingstarted/product-management-for-hackers/

    Upsell where you can, and test ideas to optimize for conversions.

  • ML

    I will say you’re great at follow-up Brandon. Great talking with you on the phone today.

    Good luck with this campaign.

  • qcregs

    Hi,
    You can also look on http://www.questchain.com

    We certainly work that way