Hacker News new | past | comments | ask | show | jobs | submit login
Thank HN: for helping me get traction with DuckDuckGo and Traction book – AMA
183 points by epi0Bauqu on Oct 7, 2015 | hide | past | favorite | 80 comments
(This is my old username, which I thought fit well with this post. You may know me now as yegg.)

I launched DuckDuckGo on HN (then Startup News) seven years ago: https://news.ycombinator.com/item?id=315142 and the idea of Traction book five years ago: https://news.ycombinator.com/item?id=2098068. If I hadn't gotten encouragement and excitement in these threads (which I did!) then I might have quit each shortly after. So thank you for that!

In 2009-2010, when I was struggling to get traction for DuckDuckGo, I started doing a series of interviews on my blog with successful founders about how they got traction in an effort to uncover a structured process for doing so. Naturally, I interviewed a lot of HN greats like patio11 (Patrick McKenzie), justin (Justin Kan), garry (Garry Tan), kn0thing (Alexis Ohanian) and other startup icons like Eric Ries, Jimmy Wales, etc. Last year I put a bunch of these early interviews on YouTube if you want to check them out: https://www.youtube.com/user/tractionbook

If I hadn't been part of the HN community I probably wouldn't have done these interviews or the series of blog posts that led to writing Traction, which in turn led to getting traction for DuckDuckGo.

A lot of people on HN have used DuckDuckGo and read the book and gave excellent feedback on both, a lot of which we have acted on to make these things what they are today today. The proximate cause for thinking about this was the second edition of the book came out yesterday, and I'm in a reflective mood.

I'm extremely grateful for being part of this community. Ask me anything, and I'll try to help where I can.

How much traffic does the Tor hidden service for your search engine get? Would you characterize the usage as significant, or is it mainly kept up as a public service for a small number of people who use it?

It is not significant, though we'd love it to be.

In my case, the only reason I don't use it is because I cannot add it to my browser; `opensearch.xml` statically points to https://duckduckgo.com/

I love DDG's bangs and instant answers. Do you have any statistics on their usage? Like counts of how many times a bang or an instant answer was used?

And will Dax play a bigger role in the future, in branding to give more personality? It wasn't easy finding the name of the duck I see daily. (https://duckduckgo.com/?q=name+of+the+duckduckgo+duck)

Thanks! We are working on something that could tell the platform (duck.co) how much each instant answer is being used. Now each answer has a page (at http://duck.co/ia). Bangs are a bit different but we're working on open sourcing that too.

Dax, yes! We're going through internal branding strategy right now and your comment is very welcome in that context.

Instant answers are brilliant, and the community contributions really flesh it out. The text even changes color with the theme. I'm wondering if something like this could be ported to JavaScript and made available in NPM or Bower.

Thanks for your answers, and for DDG! Waiting to see more of Dax and other improvements. The DuckDuckHack documentation looks better now so I will try making an instant answer for Dax's name.


Congrats for your search engine. I admire the work you did. The only reason I don't use DDG is that I'm Greek and the results for Greek keywords are many orders of magnitude off-mark compared to Google. Why is that? Any hope to improve results in the future?


I'm not sure if we improved for Greek in particular yet but this generally (non English) has been a focus recently so it should really be improving. I'd be curious if you notice any improvement and if the region toggle does anything for you.

I'm using it all morning (UTC +2 here) and I must say that it has improved drastically from the last time I've used the service! Cool :-)

Does Bing output good results for Greek queries? DDG had drawn from Bing in the past.

No, both Bing and Yahoo! lag far behind Google. DDG seems to be catching up though.

Gabriel, do you plan to implement time-bound search filters on DDG? I'm probably not alone in finding it very usefull and would gladly avoid jumping back and forth between DDG and Google. Anyway, kudos for all you've done.

Yes! People have been asking for this for literally years (and I want it too). For various reasons it has been difficult to do, but now we are very close. So look for it soon.

Great! I think it's the last thing missing for me to use DDG exclusively. Cheers.

Are there any plans for a Google trends like analysis tool or any plans on opening up anonymous search statistics?

On a personal note, DDG has been my daily driver for about 2 years now. I love the bang shortcuts (!man and !cpp mostly).

The first few months or so, I ended up following almost every search query with a "!g query", but search results have really, really improved. Now I only have to use Google for local topics and/or very recent events.

What are your future plans for DuckDuckGo? Knowing what you know now, what would you have done differently with DuckDuckGo over the last seven years?

Two good questions!

There is a recent PEW study showing that 40% of people would prefer a no-tracking search experience. And yet a very small percentage of people have ever heard of DuckDuckGo. As a result, we think we have a lot of room to focus on making the product better and growing, and that is really our future plans in a nutshell.

In hindsight, there are always activities you realize now were a waste of time, and it is easy to say you wouldn't do them now, but in the moment it is much harder. So in general, what I've been trying to work on is making sure we're always working on the right projects, and scoping them right, given the information we have in front of us. I would have liked to be better at that earlier, but in general I wouldn't have done that much differently.

So you should ramp up marketing. Or it sounds like you want to improve the quality and rely on word-of-mouth.

As a user, I am always a bit afraid that DuckDuckGo might be gone one day.

How do you keep the DuckDuckGo afloat ?

DuckDuckGo is actually profitable! It is a myth you need to track people to make money in Web search. Most of the money is still made without tracking people by showing you ads based on your keyword, i.e. type in car and get a car ad. These ads are lucrative because people have buying intent. All that tracking is for the rest of the Internet without this search intent, and that's why you're tracked across the Internet with those same ads.

I think it's great what you're doing (DDG is my default search engine). And I'm glad you're profitable so you can stay alive.

But I'd like to think that if you could find a way other than ads, you'd take it in a heartbeat, even if you made less profit, but as long as you could be self-sustaining. Is that true?

I have a very strong opinion about advertising and its effects on the web. I would love to hear your feedback on them:



I'm trying to spin up a movement to shift the web away from its reliance on ads. Let me know if you can be an advisor or resource for that. One day I might thank you in the way you thank all those other founders you listed above. :)

Yup, that's very true but the numbers just don't add up well on search in terms of subscription or donation. We have tried to do much less advertising, making it up on affiliate revenue, but we can't do that much because of our privacy policy. We can only use programs that can be anonymous, which means the companies run them themselves through URL parameters, and that's really just Amazon and eBay.

If you don't mind sharing, how much do you get through those affiliate programs compared to advertising?

Also, would you consider providing some kind of paid model in addition to ads? I use DDG dozens of times per day as my primary search engine; I would happily pay for it. I realize that model doesn't scale to the entire world, as only a fraction of people will pay, but do you have some way for people who like what you're doing to pay you a modest amount, with as little friction as Patreon and similar?

We haven't disclosed it, but advertising is the bulk.

The problem with paid is three-fold. We're a for-profit company so don't want to come off as taking donations like a non-profit. Second, we don't want to collect personal information. So we could do something like bitcoin but that isn't exactly low-friction :). Third, it is promoting that it exists. We have a hard problem getting the word out about anything not on the SERP and we wouldn't want to put it there.

Open to other ideas though.

Regarding personal information, I have no problem providing payment information, as long as that isn't tied to accounts/searches/etc. (A separate service, similar to Patreon, seems ideal.) But I do realize that the number of people willing to pay you may not justify the work needed to set something up.

I do understand your concern about not wanting to appear to take donations, though. I just wish I had some way to pay you for your service that doesn't involve viewing or clicking on ads.


We really suggest people who want to help to just help spreading DuckDuckGo, which is honestly our biggest problem (not money): https://duckduckgo.com/spread

Could it be up to the user? For example, subscribers do not see ads, but everybody else does? Or does that not work for your advertising customers?

It more doesn't work for our privacy policy since we don't track users.

I think that the fact that DDG has a no tracking policy allows users to turn their ad blocking off, this might contribute something towards profitability...

I definitely whitelist ddg on all my computers.

As an advertiser, I would love to advertise on DDG, but to do so, I have to use Bing which means 99% of my ad spend does NOT go to DDG. I don't care about tracking - just want to reach a great audience that is using DDG. Will it ever happen (DDG native ads)?

We don't see too many Philly area firms getting attention on HN. I try to promote the area when I can (here or on Reddit). What would be your top reasons software engineers should consider the Philly tech scene as a place to work and live?

Lifestyle -- we moved here 9 years ago not knowing anyone -- after a rather complicated search for a great place to live and raise a family. There's a lot to it, but it is climate, affordability, architecture, culture, etc. Granted, some of that (maybe a lot) is personal preference but I think a lot of people share similar preferences.

Cheaper -- you can bootstrap really easily here compared to other places even nearby (DC, NYC, BOS).

Community -- it is really a nice community where you can really reach anyone really quickly and they're willing to help.

Would you add anything?

With lifestyle, I'd add that the Philly tech market is a bit different than many other cities in that the companies and startups are mixed between the city and the suburbs. If you prefer more suburban or rural areas for your home, you can find affordable places to live and choose to work in or outside the city and still access the culture the city provides (not to mention being ~1 hour from NYC and Baltimore). Or if you like to live in the city, you can still find commutable places to work outside the city. I think that's a bit unique.

I'll second community as well. Really helpful and approachable people in the startup community of Philly. I've never spoken to you, but have read that you are among the most approachable in the area.

I might add the available feeder system of colleges in the region - Penn being obvious, but some less obvious schools like Lehigh and Villanova that Philly firms can tap for talent. Unfortunately those students often leave Philly, but as they get more options to say they will do so.

I think the area is still in need of one or two big name companies to keep talent in the area. A larger dev satellite office of a Google/FB or even Netflix type would help keep those grads in the area after graduation, and at some point they may have grown deeper roots (families, kids in school) that might keep them in the area. I'm starting to see more and more local engineers moving west, as they feel they are "outgrowing" the area. I think a company with a good employer brand (as it relates to software) would come in and be able to hire a ton of great talent.

Thanks for answering, and best to your continued success.

Gabriel, thank you for DDG I have been using and promoting for years.

In terms of mobile any plans for the future you can reveal? How do you feel about Siri, Google Now and Cortana is it something you think DDG can do or can be used as a backend/source ?

Thank you! We're of course very focused on making the best mobile experience we can. We think our instant answers can be a big part of that since answers are even better on mobile where space is limited and browsing sites is slower. How to exactly showcase them is a bit unclear, however. So nothing major to reveal yet, sorry :)

Thanks for the answer I'm crossing my fingers for a major player to use DDG for their personal assistants. The first step was already done when they set DDG has a search option :)

Thanks for publishing your book it has been really useful for me.

When you were first working on DDG who did you show it to before you "launched" on here? What kind of early feedback did you get?

These interviews (I referenced above) were really the first things I did: https://www.youtube.com/user/tractionbook

I blogged all of them on Hacker News (you can search the archives to see), and then started blogging ideas that would become the book and got feedback from those on my blog and on HN.

This was all before really writing. Once we actually started writing we recruited "early readers" periodically to read and give feedback up until we had a full draft more ready to go, in which case we hit up our whole list and got another round of fresh-eyes feedback.

Sorry, I meant when you were working on your search engine

Oh man, double face palm :)

I would go around to local meetup groups here in Philly and continually show it to people, as well as friends and family.

Great job w/ DDG and thanks for sharing this encouraging thread.

You're very welcome!

Congrats and all the best with the book! Happy to participate and hope it can help get my experiences with Reddit, Hipmunk and beyond into as many people's heads as possible.

Thanks! I would too love to spread those stories to new audiences. We are really trying to branch out beyond the startup world but are honestly new to it and unsure so far what will work. So any ideas from you or anyone would be appreciated! Open to try almost anything :). Also, unjust sent you a DM on Twitter about a related matter.

I'd just like to say thank you for making the search box the first thing that gets focus when pressing Tab.

Google got this one wrong (though I suspect that google does it on purpose).

Is there any chance that you guys could implement a feature like your major competitor's "scholar" search for searching academic articles?

We've had a lot of requests for that, but it is not on our immediate roadmap yet. If there is a good source out there (vertical search engine), it could be incorporated via DuckDuckHack (http://duckduckhack.com) -- that would be the best bet.

Was this book updated for the Oct 2015 release? Or is this more or less the same as the book you asked when you originally solicited feedback?

Major update, though the overall messaging is the same. It is a much more high-quality book. The first one was self-published and this one is published by Penguin Random House. So the first thing is it re-edited all the way through. We ended up cutting out about 50 pages, even though we added a couple of sections.

Second, the sections we added were a preface about my struggles with getting traction and a testing addendum giving 1-3 suggestions of how to test each of the nineteen channels (something people continually requested).

Third and finally, I really focused on rewriting the five introductory chapters to focus on things I saw readers of the first edition struggle with. We simplified the framework from 5 to 3 steps, and tried to shed light on the more non-intuitive aspects (like don't prematurely optimize marketing tests, etc.).

Congratulations, Gabriel on all the success you've had so far with DDG and Traction. I was an early user of DDG and although I didn't stick around, I truly do admire you for what you stand for and your courage and perseverance to push through it.

I do believe that it is critical that we do have choices. Choices that can free us from a single establishment. So, thank you!

You're an inspiration. Keep up the great work!

Hey Gabriel,

Just wanted to say thanks for DDG, it's been my default search engine for going on a year now and absolutely love it. I'm also a developer(DDG is also the reason I started to dabble with Perl) from PA (about 10 minutes from DDG) and have to say it's very exciting to see something like DuckDuckGo created in my hometown area. Thanks again for the great work!

Oh that's awesome! You might want to check out our DuckDuckHack meetups (there is of course a local one) http://duckduckgo.meetup.com

Are you working on any new business dev for DuckDuckGo? I feel that no-tracking is a soft value add and I imagine keyword-based targeted ads is limiting when competing with large search engines that tracks more information about the users. Which direction are you taking DuckDuckGo that further differentiates your product?

Our long-term product vision is community-driven instant answers for most searches: DuckDuckHack - http://duckduckhack.com/

We think partnering with thousands of the best sites on the Web with great answers in all niche categories (lego parts, municipal bonds, anything) is a win-win-win for everyone (us, partners, users). Right now anyone can suggest an instant answer source, and anyone can develop it. The answers themselves and the entire platform is open source.

I think there is real value in consumers being able to glean value they want from their own data, would you or have you ever considered a program that allowed users to share data with you in exchange for services, ads or other products they continually found relevant?

We have anonymous cloud save now for DuckDuckGo settings, which is sort of like this already: https://duckduckgo.com/settings -- for example, you can set a language or region setting.

I think more generally you're right privacy is about control and choice, as I tried to explain in https://medium.com/life-learning/privacy-is-at-a-crossroads-...

totally agreed on that post (its one I've read before) - i think a consumer truly having control (I'd even envision a dashboard where you could give or retract access to your data instantly) is what would be most helpful for the consumer to truly take more care into what is happening in their lives as the physical digital worlds quickly merge into one.

That said, I'd also love to see a google apps like product from DDG that could really compete but promise not to sell my company's info or use it for useless ads and general spam

We would also like to see similar services across the spectrum and really believe people want private alternatives that they can switch to with little sacrifice. That said, search is difficult enough for us!

How would you self describe the general search quality of DuckDuckGo, excluding the (brilliant) Instant Answers?

How do you evaluate the general search quality?

Do you have any focus on expanding crawling in general? And localized content in particular?

Our long-term focus is really on the instant answers, and we hope to really build up the DuckDuckHack platform to a point where the majority of searches have instant answers on par or better than the links.

Of course, basic link quality is still par for the course. We have worked a lot this year on better language support so if you type say in Finnish we recognize that and adjust the results appropriately. For truly localized stuff (like restaurants, transit, for example), we really want to attack that via the instant answers.

Perhaps it's completely out of reach (fiscally or ethically) but have you considered partnering with Mozilla to become Firefox's default search engine?

We'd love to become the default search wherever we can!

Are there any features that you wanted or thought about implementing into DuckDuckGo but could not for technical reasons or otherwise?

Tons. There are so many constraints like getting access to data in a cost-sustainable manner. We try not to pay to data, for example, but some data is expensive. That's why we still don't have movie showtimes, for example.

Interesting. Sounds like there could be space for a client-side extension that lets users enter API keys for rate-limited APIs, etc.

Oops I missed it. In case if you check this sometime, I wanted to know how much of Python is being used at DDG

when will you add instant search results in the browser's search box? For example, if i use Firefox and have Google ad the default, and i type "32 * (13 - 9)" it will instantly show a dropdown with the answer 128. I would really like DDG to do this too.

Hey Gabriel,

Do you have some plans to open source the search engine of DuckDuckGo one day and let people contribute?

Yes! Check out http://duckduckhack.com/ -- it is our open source instant answer platform, which is becoming more and more of DuckDuckGo. The whole platform is open source itself (in addition to the answers).

Thank you for the reply, but I already knew about duckduckhack - I made a small contribution to it in 2013. What I meant however is the core search engine, that is currently closed source.

Thank you for the contribution! Would love to get you back :).

We've been actually open sourcing more and more because what I meant is that as DuckDuckHack grows it is becoming more and more of the search engine. Check out https://github.com/duckduckgo -- bunch of new repos since 2013.

Just switched to DDG 3 weeks ago and can't see me going back. :-)

Where did the name for DuckDuckGo come from?

Do you think DDG would have still been a success without the Snowden revelations?

Depends what you think of as success :)

Snowden accelerated our growth by years and we are really grateful for that. However, I believe we were already a success and I was happy with where we had gotten before then.

There are still no good limits on online tracking and as a result it continues to get crazier and crazier and more people are reacting. It was already heading in that direction pre-Snowden. For example, in 2012, a year and half before, we saw a huge uptick when Google changed their privacy policy to allow tracking across all their properties. And a year before that, we saw another huge uptick when we did a privacy-focused billboard and microsite (http://donttrack.us/).

Where is the current best place to buy the book in an electronic format?

How does DuckDuckGo index content? Does it use its own spider?

Great book Gabriel! Excited to help you promote it.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact