Samuru: A Search Engine Designed By An SEO

Apr 22, 2013 • 8:59 am | comments (23) by twitter Google+ | Filed Under Other Search Engines

SamuruBrandon Wirtz, an SEO who started his own search engine named Samuru, aims at making a search engine that is spam free. Doesn't every search engine have that aim.

Someone posted the search engine to Hacker News where Brandon explained how Samuru is different:

Samuru doesn't use link authority, it analyzes pages and matches what you queried to the types of pages and picks the best matches.

Let me give you an example. You search for "How to Make cupcakes" Google says give me the pages that have the most inbound linkes (over simplification) that contain all those words. The winner is Brandon's Cupcakes (not really but play along for a minute) because it says, "We know how to make the best cupcakes, because we have been doing it for 25 years"

That is not a useful result. Samuru on the other hand says "how to make cupcakes is a search for instructions" and it looks for pages that match the words, and are written as instructions.

We weigh other factors, like is there an author associated with the article. Do they routinely write about the topic?

We do this for reviews, products and other things as well.

To be a full replacement for Google we need Driving directions, and image search and a lot of things. But in order to do all the other things we are doing we needed a search engine. (related content, analysis, speed testing, building a corpus of words)

Responses get better if you search something someone else has searched or do a second search 30 seconds later. This is because we haven't deep indexed the entire Internet yet, and so we don't have all the deep data.

Matt Cutts, the head of Google search spam, responded to this post asking something the SEO community would appreciate:

Hey Brandon, congrats on launching Samuru! I'll be curious what you think of running a search engine after being an SEO for so many years.

Brandon said he likes it, "we went with the approach of how can we make this impossible for Brandon Wirtz to game."

For more about this search engine, see this page.

Forum discussion at Hacker News.

Previous story: Google Captures Murder: Body Being Dragged Into Lake?



04/22/2013 01:10 pm

Nice attempt. But Google is too better and advanced right now, let's hope you can reach there soon.


04/22/2013 02:41 pm

Someone doesn't have to routinely write about a topic to have the best content. That's where links come in. People have voted for the best content by linking to it, their algorithm doesn't know anything about cupcakes and can't bake them, so the idea of relying on an algorithm that does direct matches is flawed from the start. You need links and/or social to rank the data. Although basic search engines like this worked back 1995 when the web was tiny, times have changed just a little.


04/22/2013 03:03 pm

Looks like one of their main ranking factors is domain age. Followed by direct name matching, then by complete nonsense and then again followed by domain age. Just from my first 10 minutes of observations. Not very exciting.


04/22/2013 03:49 pm

His example "how to make cupcakes" is a TERRIBLE example. All of the results on the first page of a Google search provide guides or step-by-step instructions for making cupcakes. I'm not going to check, but I assume they are all useful and reliable pages. Wirtz, the creator of Samuru, is creating a straw-man argument that Google is not good at providing you results, when in actuality, it's doing a fine job of it.


04/22/2013 11:17 pm

I assume that not much money is behind this project so people should keep that in mind when they comment on the quality of the search engine. As far as I can see, Samuru seems pretty nice - again with budget kept in mind. I do have to comment about the spam thing though... Many of these new search engines claim to be on a mission to reduce spam but in reality, they are just using spam as an excuse for not being able to index many pages. When you think about it, spam depends on the vertical so to say that you are on a mission to eliminate spam, is like saying that you aren't willing to provide answers to every single query.

Brandon Wirtz

04/23/2013 05:13 am

The example is meant to show something people can understand. You are right, the user tracking of clicks in results negates this, for this search. How ever when you start looking at all the content that is milled, this does work. If I said "how to convert a 67 mustang to shelby tail lights" you would argue it was a contrived example. Google thinks this is a shopping search. Google does fine for things the mainstream searches for. It sucks if you are doing real research, or wander off the beaten path to things that are more obscure. We don't have billions of users so we can't Google Dance using click through. So if we are 90% as good with no users think what we could be with users.

Brandon Wirtz

04/23/2013 05:15 am

We raised $2m. That is spread across all of our technologies of which search is just a small part. We index everything. We don't deep index everything.

Brandon Wirtz

04/23/2013 05:17 am

The best content is often not well linked to. A professor who publishes the definitive guide to a topic is often cited by others who have much more rank. Think of all the ArsTechnica stuff that is dumbing down of more collegant papers. We track that link and bubble it up to the surface. We also do this with things like press releases that broke stories.

Brandon Wirtz

04/23/2013 05:18 am

We don't use Domain age at all. We do have too much bonus for domain matches. This was added because we penalize Brand Pages too much, and we attempted to compensate. Early stage the Scoring needs tweaks from time to time.

Brandon Wirtz

04/23/2013 05:21 am

Google has hit the limit of what they can do with Page and Text rank. Panda was really the proof of that. We don't have to replace them as long as we can prove we can do the same things without these traditional methods of ranking. We don't want to replace Bing or Google. We want acquired by them. If we can be 95% as good on 90% of searches. Better on 5% and worse on 5% we can be the deciding factor in who is the number one search. Also because we don't care about link authority we can crush Google at internal document search. Because our scoring still works in a linkless environment, and theirs does not.

Jawad Latif

04/23/2013 07:26 am

I would like to appreciate the way Brandon Wirtz has answered everyone logically and detailed.


04/23/2013 08:19 am

@Brandon Wirtz First of great effort.I hope you succeed where others have failed :) The only problem I see here is that you guys are throwing important ranking factors/signals out the window here, which could be that an end-user would be getting low quality/spam SERPS, which we all know is exactly what he/she does not want. Unfortunately an authority on the web these days is determined by the amount of votes/links/citations it receives from people as they see it as a well respected/useful resource.And by the looks of things there will not be allot of this going around on your SE, which then still makes it an unattractive option.Same with domain matches, as we all know that there is a lot of low quality EMD's going around. But with that being said, if you guys are willing to provide a user with fresh content and other internal documents that people might find useful in a different context, it could be a good 2nd alternative to the big G.

JR Oakes

04/23/2013 12:53 pm

I cannot believe there are negative comments on this post. I remember Google when it first came out gave users something completely different from what was out there, intelligent search and not all the crap that yahoo, aol, and excite had going on on their home pages. I think the idea of attempting to analyze content and letting that be the determining factor is a terrific direction even if it is not 100% perfect now. Right now, serps are dominated by budgets instead of content. I agree with Brandon's response below that Panda and Penguin have exposed flaws/limitations in Google's machine. The fact that Google is going after websites that exploited this flaw for years is strange to me and let's me know a great purpose behind Samuru. Google had a decision to make, How do we get rid of spam in serps? Instead of asking how they could refine their engine to better understand/grade content, they decided to penalize sites for things like manipulative linking, etc. Everybody looks for an edge in business so it is kinda like giving candy to a 4 year old and telling him not to eat it. Samuru let's us see at least a glimpse that the former(understanding content) was/is possible. Thanks Brandon and cheers to your courage to battle the big boys!

Brandon Wirtz

04/23/2013 03:27 pm

We do use links from things we deemed relevant. If you write a paper that explains in simpler terms what another paper said, or you site 14 sources in your paper, we pass that authority to the sources but ONLY for things which both sources would be considered results for, and both benefit. In this way you can still "vote" for answers, and the most important works "rank up" but you can't as easily cheat your way to the top. Is a good example of this. Yahoo Blog would not normally be a top result. But NextWeb likely would. I have better examples but not on such main stream topics. I try to stick to examples that people feel they would have searched.

Jaspal Kalsi

04/23/2013 04:51 pm

Its a nice effort and reminds me of the time when Google had just started and everyone questioned how a clear page with just a single search box can substitute the then grand Yahoo! Did anyone else here try to search for "Samuru" on Google? I Did ... the site is not listed and other relevant posts and press releases are listed as more relevant than the site itself. Going by the way Google is designed to work taking into account the latest news and discussions which point to a particular entity/brand ... the fact the is not listed in the top results as a currently relevant result is quite surprising for me. In my personal view Google has been doing a lot and has been delivering value. I use it for almost all my searches as well but its reaching a point where quality is not maintained to higher levels and somehow they have more quality issues at any updates / new options in search ... Google Places and the many turns it took (an continues to take) is one such example. I feel there is still room for a good search engine and so welcome Samuru ... I may not switch to using you for my searches immediately .. but am watching for sure.


04/23/2013 05:09 pm

This search engine is a brother of BING.COM and this search can be easily manipulated. I tell you how? You can directly influence META TAGS, Tags, blogs, h1, h2 optimize it well all this factors, Anchor text linking highly important here. So as far i can sense this engine is dam easy to manipulate in ranking. If you have a good internal linking with Keyword based titles and H1, H2 tags then yes you can. LOL Anyway good search engine. How about traffic??

Angry Troll

04/23/2013 10:24 pm

From what are you basing the meta tag statement? Everything I have seen implies that is not what they are doing.


04/23/2013 11:19 pm

See what i mean to say is that if anybody wants's to rank on this search engine then it will be very easy to rank. Its an easy task!


04/24/2013 08:42 am

I don't think Google will agree that they have reached a limit with what they can do with Page and Text rank, I wouldn't expect them to just roll over and die :) I've found low-linked site with better content out-ranking better linked but poorer content content more and more over the last year or two in particular. I wish you all and best and hope you do well, but its not an easy mountain you've got to climb! Good Luck!


04/28/2013 03:48 pm

Relevance in results seems pretty nice, for a new SE despite the lack of deep indexing. However, if this search engine does gain popularity, I think it wouldn't be long before blackhats & impatient SEOs come up with loopholes and what not. Correct me if I'm wrong.. but since the major focus is on-site, doesn't that make it so much easier to manipulate?


04/28/2013 03:52 pm

Exactly what I was thinking. I think the concept is good, but chances are it would work only in a perfect world.


06/16/2013 01:20 pm

Two words: NSA and PRISM. Use Google on those and see what your get !!

Luana Spinetti

11/28/2013 06:42 pm

Brandon, you're a genius! :) I have Samuru for my default search engine and I'll keep my radars alert for news.

blog comments powered by Disqus