Keep it small and focused. The worst thing is these very general, empty directories that are no use to anyone. Start with the most specific niche, and expand outwards as necessary. It's really easy to underestimate how much work is involved, and end up with loads of gaps.

It's really easy to underestimate how much work is involved, and end up with loads of gaps.

I think I am starting to discover that very problem now...

It's really easy to underestimate how much work is involved, and end up with loads of gaps.

This depends on the scope of your goals. If your goal is to make a "directory of everything" this is obviously an unachievable goal.

However, if you start with something like a town or city directory, it is very easy to get some useful categories up and running on the site within a few days or less. The same thing would apple to creating a niche directory.

If you are going to attempt to allow the submission of just about any link, you need to contemplate how you can introduce some limiting factors. Here are a couple of examples:
1. Increase the amount of work required to submit a listing, so that each and every listing will be created by someone who put a lot of thought behind the listing. Just getting the title, description and url is insuffient!
2. Create a system or triage that only promotes the best content to the top (ex. digg). For example, maybe the content will not be indexable by google until certain criteria are met (votes, comments, other popularity or quality measures).

So the categories would be based on who NEEDS what, not who IS what?

Link partner categories should mirror the types of sites you want to get links from. Sometimes you have to reverse engineer and ask yourself "who is my target market and where do they hang out on the web?"

For example, if the wedding photographer wants to link with other businesses in their local demographic, the photographer should have link categories that match those markets. If the wedding florist in the same city as the wedding photographer visits the photographer's website but does NOT see a wedding florists category, the florist is LESS LIKELY to request a link with the photographer. If the florist sees a link category that matches her business, not only is she MORE LIKELY to request a link with the photographer but she is also more likely to go ahead and link to the photographer's website because she realizes the photographer has a category for her and is likely to approve her link request.

The photographer now has a relevant one way link from a relevant business. The photographer can link back if he wants to. He doesnt have to.

There is psychology at play when it comes to all of this. If you create link categories that mirror the kinds of sites you want to get links from, you will get one way links from those markets. Many times someone who links to you first will leave the link up because they are hoping you will link back to them in the future.

Keep your link categorization relevant to your business. Here is an example list of link categories that the wedding photographer may want to use. This is just an sample list. Many times we will see a webmaster insert city name into the link categories to make them even more relevant and category specific.

Wedding Invitations and Printing
Wedding Travel Agents
Wedding Jewlery
Bridal Fashion Accessories (there is a personalized wedding and prom garter website on the internet which I will never forget)
Wedding Florists
Bridal Party Gifts
Officiant & Vows Resources
Wedding Media & Vicdeo Production
Wedding Favors
Wedding Planning
Wedding Limo Drivers

Do not create a "miscellaneous" or "other" category as that only will generate irrelevant link requests from junk sites.

Hope this helps.

Link partner categories should mirror the types of sites you want to get links from.

thank you cnvi. That will be my philosophy.

My problem I think is the way that I may have done the structural design of the directory.

I only have five main categories, and would probably expand it only to about six or seven total main categories in the future.

But the structure is:

type -> country -> state (or province) -> city -> listing

So while that is fine for where I have states / or cities with lots of listings, for those states or cities that don't yet have any listings, google bot is indexing a page with minimal unique content on it.

So for Category 1, I have about 12 states with a handful of listings each, while the other 38 US states have minimal unique content yet.

Multiply that by the five main categories I have, and that means I currently have about 200 pages with minimal unique content on them.

So I will have to figure out what to do with those pages until I can get some good quality listings for them.

Thanks again.

@Planet13 - What we have been doing recently is only populating the database with locations for which links exist. To start with a huge city/state db with thousands of empty categories makes for a less user friendly site. But to build a region specific feature that is separate from the categories and always leads to pages with data makes a lot more sense.

Thank you, dvduval!

I'm doing the same thing right now!

Does anyone have any evidence that having a niche directory like that on your website (that does attract plenty of inbound links (from sites other than those listed, of course))...does, or does not help with SE rankings?

I didn't really consider this. If Google understands it is a directory on your site that attracts the links, could they think "oh..the links are there because of a (useful) directory, not because the person has written great content herself..thus she's not an expert, thus we wont give it a ranking boost"?

Anyone got any evidence that it helps a site rank if a truly useful niche directory is on it?

@Makavelli2007 - I try to read carefully what people like Matt Cutts say about directories, and based on that I would make sure you have reviewed these points carefully:
1. You should be adding quality content too, not only depending on user submissions.
2. You should be making sure titles and descriptions are not spammy, but reflective of the site.
3. Having a review process means that you don't let any link into your site, and this includes people who pay you.
4. It is okay to charge for the review process (for your time reviewing sites), but you need to be very clear what you are charging for.

Additionally, I would add that directories can have all kinds of additional features from video to articles that make them special. It is your job to make the directory special. If you do a good enough job, you will attract some natural inlinks as well.

With regard to building a niche directory, I think you are right on track, because it limits the scope, and increases the potential for quality.

@dvduval:

Thanks for your reply.

I wouldn't try to build a directory that doesnt add value to anyone. This directory (I'm selecting the links for it myself by doing deep research, rather than having people submit them) fills a true need in my niche, and there are people who should want to share this with their audience (as they have a problem this directory will solve).

In other words, Im not trying to build some random directory, but something that should lead to quality inbound links (if it doesn't, then I fail, but that's another issue ;-)).

I was simply wondering if...Google doesnt like directories for some reason - e.g. if you have a niche directory on your site vs. having a content (text) page on your site....and either one received the same number and quality of links.......Would Google allow the text content to help the overall site much more, than it allows the directory help the overall site?

In other words Im wondering if they might have some kind of filter.

I didnt really think so, but reading through this thread made me wonder just that, with I think 2 people stating, that you shouldn't call it a "directory" (b/c of the search engines).

If I interpret your question correctly, you are wanting to know if a particular script or software would lead to google treating the site differently. I would say the answer is most certainly "NO". There have been multiple occasions where people like Matt Cutts have expressed this. The only caveat would be if the software violates the google terms of service.

I would be wary of things like bidding directories that are created for the sole purpose of buying and selling pagerank, or any tools built for directories that would query the google servers in an unacceptable way (ex. trying to get the pagerank of 10,000 links updated).

It doesn't sound like you have any intentions like these, and I think there are a number of different script you could use that would be just fine.

it will probably be more of a carefully selected list, not a directory (thus no need for directory software, i think)..but you answered my questions, thx