Duck Duck Go Traffic & Sponsorship | Hacker News Reader

Duck Duck Go Traffic & Sponsorship

gabrielweinberg.com

89 pointsugh36 comments

That's a $4.23 CPM if my math is right, based on $5,000/1182204 qualified page views in April. That's awfully high.

It's subject to change based on demand. However, I think there is something unique and compelling about being on every page for a period of time, which warrants higher rates. Also, the current discounted rate is $2.96CPM.

The search engine is private, so it's hard to tell demographics. But I'm pretty sure the user base is generally highly technical and highly educated. This also warrants higher rates.

And I could make an argument that your userbase is more likely to ignore/block ads (although my guess is that you're serving the ad itself, so AdBlock wouldn't work).

I've had a fair amount of experience experimenting with all sorts of online marketing techniques, and it's just an awfully costly experiment to shell out $1,500 for a week. If you had a Reddit-style minimum of $20, then the CPM wouldn't be a complete non-starter for an experiment, but that's a very, very high bar.

Of course, you have a pretty rabid fanbase, so early advertisers are probably going to get a higher-than-average CTR because users are going to want to support you with their clicks, but that's just going to end up artificially depressing advertisers' conversion rates, which DDG won't have any insight into.

So, I guess all I'm saying is 1) consider lowering your rate and 2) consider dramatically lowering your minimum.

Remember, kids, AdWords are a gateway drug to unprofitable user acquisition.

Points taken. I also do a lot of online advertising experiments, and so I know exactly where you are coming from.

Really it's more that I just don't want to spend a lot of time on this right now, as I know the reddit team has. Ideally I'd love to syndicate either the deck network, reddit, or someone similar, though no good options seem available to me atm.

Out of curiosity, why are you reinventing the wheel on search advertising? Is there something you find fundamentally flawed with AdWords-type ads?

Seems like monetizing a search engine is shooting fish in a barrel, getting it to grow would be the business challenge.

Isn't this almost what DDG is avoiding? That would just violate the less spam and ads point. Unless you kept it to say one ad per search term or something.

Ads on search are fine with me if the results are (consistently) better than google consider me switched.

Someone ought to do an extensive (say a month long) side by side review of google, bing and ddg and see how good it really is in practice.

If I could spare the time I'd do it.

Great point-- I imagine this is quicker for him that an adwords clone, which might require some clickfraud protection, self-serve flow, payment processing, etc. AFAIK, the OP is just one person... I recall reading a study about who clicks on ads, and the OP's demographic is decidedly not it. You can play the "pay per click" game if your audience clicks. Otherwise, he might have better luck with the "pay per influence" stuff that The Deck does.

To the OP: a question regarding your stats page. Would you be willing to post more stats? Specifically, I'd love to know how many uniques visit more than 10x per month. In other words, how many people are using you as a primary search engine in their life?

You're right--I'm focused on improving search results right now.

As for stats, I don't have those as I don't store IPs (http://duckduckgo.com/privacy.html & http://ye.gg/ddgp). I do know that before the switch (to no IPs) I had about ~500K page views and ~150K uniques, so scaling that up I'd estimate 360K uniques (worldwide, i.e. compete.com is just US). I still don't know how many are 10x/month, but obviously much lower.

What do you think is appropriate?

Try to get in to CPC right away, CPM is not very useful unless you are targeting a specific niche.

$0.05 to $0.10 should get you some traction, evaluate at the end of the first period with your advertiser if they got value for their money. It should be a win-win if you expect it to become a steady money supply otherwise you'll just be hopping from advertiser to advertiser. Ideally you want your first advertiser to be jumping up and down to be allowed to sign up for all ads for the rest of the year based on that first month :)

Granted it's a narrower niche, StackOverflow has a CPM of about $10.

CORRECTION: $5 CPM (general site), $10 CPM (tags). So Gabriel's pricing is on the high-end of the scale, but not awfully high in my opinion.

It can also be targeted down to the tag though which makes it worth much more, you usually see people advertising on Ruby, .NET, SQL, etc. and not on the main site.

This is just a stop-gap measure to get some income before you implement context-sensitive ads, right? Most advertisers won't get any value from most searches.

I think the traffic is still too low to offer a unique context-sensitive ads approach. So this is an experiment for now. I believe this exclusive approach offers a unique branding experience, but we'll see :)

If you don't want to run AdSense ads, Microsoft's AdCenter allows (I believe) more customization with look and feel: https://advertising.microsoft.com/publisher

Only that, monetizing a search engine with other search engines' ad products would be a bit funny.

well he is, in part, driving his search engine with other search engines' indexes so it would be entirely fitting! :)

What does that mean exactly. Are you talking about the no click stuff, or do you mean that DDG isn't crawling the web and indexing it, but using results from other search engines to build its own?

> or do you mean that DDG isn't crawling the web and indexing it,

He does his own indexing - but also uses (IIRC) Yahoo BOSS and Bing API's.

I wasn't familiar with Yahoo BOSS.

http://developer.yahoo.com/search/boss/

So Yahoo is letting people use its search technology free of charge? Wow.

Edit: it looks like Gabriel answers my question here:

http://www.reddit.com/r/IAmA/comments/bbqw7/i_am_the_founder...

Note that Yahoo BOSS has an unclear future pending the final Microsoft/Yahoo search integration.

If the current CPM comes in too high for most advertisers (given your current traffic), why not try auctioning off that exclusive ad space?

An auction would have the benefit of helping you figure out what kind of CPM potential advertisers are ready to pay for.

I was just trying to keep it simple. I considered an auction, and might move to it eventually though.

I've always liked the Daring Fireball approach to advertising (not sure if he should get credit for it, but he was the first one I saw doing it), and I really hope it works out well for you, Gabriel.

I love DDG.

He should get credit! I'll update the post to reflect such.

Daring Fireball is a member of the invite-only advertising network called The Deck. To be a member is to be in an elite group of influencers.

http://decknetwork.net/

I'm not referring to his sidebar. I'm referring to his weekly sponsorship by a single company.

Sorry knee-jerk reply. Daring Fireball has multiple lucrative revenue streams.

I wonder what kind of metrics DDG will offer - I think it would be interesting to know what the user was searching for when there is a click through on the sole sponsor banner.

Incidentally, I emailed them and not surprisingly didn't hear back :).

Here's the traffic link: http://duckduckgo.com/traffic.html

Hrm, you can't trust aws stats. In my opinion they are frequently over-reporting traffic. I recommend using a JavaScript based logger such as Google Analytics or Site Catalyst if you want better numbers.

I'm not using it straight up, but just for display purposes. I manually wrote a script to ferret out the correct page views.

Even if not for the traffic page but for your own statistics. I too highly recommend using Google Analytics or installing Urchin.

good luck

Reply on news.ycombinator.com