Starting my own search engine

just an idea, would like your opinions.

[penfold25]

Hi there, ive asked a few people around in here and gotten there opinion and would like to know what you would think in the general public.
Im gonna start my own search engine soon, just a general one. Id ideally like to model it on yahoo.
I know it may have been done before and everything but i wanna try anyways :)
Not really in it to make money, i would really just like it to be popular.

My idea is that i would like to offer to webmasters the opportunity to choose a keyword of there choice , first in best dressed in whatever area. They have link permenantly, but instead of money i would like a link from them providing a minumum 100-500 a week or month..

My initial aim would be the top 500 keywords.

Ill try to get a 500 customers, which they will be first per keyword. Also as a bonus incentive, if it ever takes off, they will be permanently be in the top 5 keyword as a thank you.

if i can 500 customers times 300 , hits a months
150,000 hits a month minimum from other ppl
plus my own traffic, which will make it more.

Its just an idea, any thoughts
PS, even if it doesnt take off, the amount of traffic lost is not too significant for anyone thats why i thought it might be worth it taking a risk.

[John_Caius]

The whole point of a search engine is that it provides the most appropriate results at the top. Or have I got the wrong end of the stick?

[Shak]

I think you are a referring to a cross between Directory/Top sites type of thingymijig.

Shak

[penfold25]

well what i mean is that, initially to start up
instead bidding for top position or pay for a sponsored listing at the top,
there will be keywords available to like 500 sites, they each can technically choose a keyword of their choice and in return they provide hits to the website.

it is no way a top site, just a way of initally bringing in alot of popularity, because if many sites provide hits at one point.
its just an idea

it will function exactly like a human based search engine, except if it becomes popular the first 500 sites who bidded for the keywords and assisted with traffic will have permanent top 5 position for selected keyword

Remember google used to advertise to get hits.

[stevegpan2]

seems this is not a spder/robot search engine but a human edited directory.

Is this correct?
Like overture?

[edit_g]

Here's 14 reasons to devote your energy to something else:

Deja.com
Direct Hit
Excite
Excite News
Go
Go (Infoseek) News
Infoseek
iWon
Magellan
NBCi (formerly Snap)
Northern Light
Northern Light Current News
WebCrawler
WebTop

Thanks to searchengineshowdowndotcom for those... There are probably 1000's of others.

[penfold25]

this is a human edited directory.
yahoo was always the best when it first started, i mean google is cool , but the old yahoo structure was my favourite.

It doesnt matter how long it takes i will and maybe i few ppl i know can build it pretty quickly, just cos google lists 10 trillion sites does not mean everyone is going to view them all.
I bet out of all the sites listed there are certain ppl visit again and again and some ppl never visit.

I know there are alot of other search engines out there, but look i plan to only add interesting stuff to it, stuff ppl will actually bother to read.
They can keep all the boring stuff on the other search engines, i want to specialise in things that ppl search for.

[stevegpan2]

this will take a lot of human resource.

keep begging people to use your site at beginning.

free? people wont even spend a minute on your site?

my suggestion is to use my staff in China to help this. As labor cost there is low and they have plenty of time to beg people use the search engine

remember those dot com days?

buy eyeballs, revenue not profit?

[penfold25]

well the point is , if i do have an existing customer base of 500
thats 500 times per 300 hits from each of them to be listed each month, which is hardly much.

but that equals to 150,000 hits in one month.
that would be a good beginning.
out of those 150,000 is 500 like it. That is a great start.

Look im not in it for the money, i dont care whether it fails, its something fun to do and if even just 1 person uses it ill be happy.

at least the bills will be cheaper then
:)

[penfold25]

its more like a team effort than a single effort.
having initial traffic of 150,000 a month will provide an inital interest hopefully

ppl will try something new at least once, in any case i will not bother with this project if aint going to be unique
i dont want to re hash.

in terms of content , having web sites ppl actually go to rather adding rubbish will help a big deal in keeping number of sites down.

[jeremy goodrich]

my advice?

use [dmoz.org...] as a starting point. There are some ways you can use the data for free - and that is as high a quality database of human edited meta information as can be found online.

Then you can build from there. :)

[Brad]

penfold, if it gives you pleasure than do it. :) You never know what people will take a liking too. Just be aware that it is not a business that many suceed at.

I would dig around in some of the old threads in the Directory forum - there are some real gems there about running a directory.

My two suggestions to you are:

1. Get good directory software. Do your research on that. It will save you time and money down the road.

2. See if you can get a backfill for your directory (like Yahoo used to get from Inktomi back in the old days.) Because unless you start out with something like the ODP data Jeremy was talking about, your users will expect some sort of result to their searches.

Good luck and enjoy!

[jimbeetle]

penfold25, you might also consider narrowing the scope a bit -- or a lot! Maybe pick one area of interest, a hot industry say, and build something solid, then add related areas later on. Don't think I would go for the whole world to start.

Getting frustrated trying to get to dmoz just now, but don't they have a category for "internet directories" or something similar? That might be a good starting point for ideas.

Just my 2 cents,

Jim

[penfold25]

thank you all for the great advice
this is seriously why i joined, such good people

Dmoz actually is really what i like to be honest, that is a great structure.

but ill take your advice and do my research

[Blue Gravity]

If you hope to mimic the ODP the one thing that you can do that will boost you way over their spot as "the human editor directory" is funding. That's something they lack greatly in. You'll also need a staff, there is NO WAY you can manage it yourself. If you're serious about it, think in terms of search results quality, then think cashflow. Human edited directories are great ideas for engines/directories that want clean, spam-free, results, but again, you'll need a pretty large staff to start off with, as well as a big 'ol business plan, and yet again, funding is a must. Goodluck!