I’ve touched on microsites before, but lets go a bit more into the concept. The idea is essentially this, build seperate, small websites that are good for several reasons:

• Being Separate from your mainsite has its advantages. You can be more specific, where you couldnt build a 5 page detailed look at one product easily on your oscommerce you can on a seperate site. Also by being seperate you can highly target your seo more blatently without worrying about jeperdising your main branding. Similarly you can express other thoughts on a topic, for keyword placement and to attract different end users only to lead them to your main site.
• A setup microsite can act as a bridge drawing people to your main site.
• A setup microsite can feed PR to your main site (hosted with a seperate company)
• Quick to set up
• Done correctly it can provide free advertising, free customers to your site, free PR boost for your site and even create a revenue.

Effectively you decide on your main site, 2 sites or small network of sites - your flagship big push project/s and then build a network of smaller, supportive ‘pusher’ microsites to aid the growth of your main site. They are designed to be left alone with little to none maintenance post development (which should only be a day or so) and self sufficient, paying their own way.

Here’s a few steps I would advise you read before proceeding with developing a good microsite network as a base to a main site.

Steps:

1. Get a different host - Hosting accounts are key to making use of microsites. It is probably not nescessary to have a different host for each site - maybe 2 microsites per hosting account in different niches. This is because each different hosting account will have different IP (shared hosting might have a load on one ip.) Different IP’s will help defuse your spread to the search engines. (it will help but I am sure google is not stupid to this!)
2. Get a good keyword relative domain - “candle-wax.com” would fit a candle wax subsite.
3. Go Niche - Choose an absolute niche - go niche and then niche of that niche. The better the niche the more chance your microsite has - if you sell a range of candles, make a microsite about candle wax types.
4. Build around a main site - using the previous example if you ran an online candle shop you could build microsites on: candle wax, bees wax, candle products, candle gallery, homemade candles, original candles, world candle records.
5. Time Manage microsite production - Limit your time on each project - I tend to allow a day every few weeks to building a microsite. Its not always constrained to a day, and doesnt always take a day - but don’t waste years making it stunning, stick to the basics - good original content, good link structure and clean code.
6. 1-15 pages - its only a microsite, start it as a microsite, plan how many pages you are going to have and fill them with good original content.
7. Links - Stealth links work - make it a ‘fake’ online shop that when the person clicks ‘add to cart’ it sends them to your main site, its also a great chance to get some real contextual links to your main site. Use words that are what the pages they are linking too.
8. Monetize - I will explain further down but done well will actually make money as well as helping your main site.

So if you manage the above you should be able to create 1 microsite in a day, 1 effective, targetted, long tail niche micro site in around 8 hours. Then what? Well then you push it - as you would any other site - but provided you have chosen your niche well and seo’d your pages this shouldnt be too hard. Strategic linking of your microsites together can help although you want google to think them as different as possible so use different sources for linkage, dont typically link them all together and try and differentiate them all from each other. Also do not use your main site to help them - this is not benificial. Use the microsites to boost your main site not the other way round!!

Monetizing your microsites should more than cover the costs associated with setting them up, therefor allowing you a free place to advertise, draw customers and further dominate your niche.

First things first stick some pay per click or similar on there - although typically you are trying to get people to your main site - so not over the top otherwise you might loose customers for only a few pennies. Using Matched.co.uk you should be able to make a minimum of £15 per month, unless your site is badly designed and poorly linked too (and you don’t have 5 good pages on it). You give them 5 urls that you will stick links on and they will approve/not approve you and give you a bit of code. Having this ad on your site will then get you £3 a month. Not much but 5 of them and you get £180 a year - that should more than cover hosting and registration fees . You are limited to 5 sites with each account on matched though (each with 5 pages = £75 a month)

Kindly use this link if you are going to use Matched so we can all benefit

Further monetization with amazon affiliates, Commision junction or ebay auctions is all good progress but as the site is so niche you are not likely to get huge returns on the effort. Personally I keep it reasonably simple but I have got a few micro sites running all of the aforementioned.

So try it out - microsites as a base for a larger web project! - If you require further consultation regarding seo powered by creative microsites then let me know

---

Clicks are effective viewers, completely targeted and ready to be entertained. You pay google, yahoo or one of the other world dominating company’s to dice up their success and give you a billionth of a billionth of the result.

It almost annoys me that I use google adwords, although with competition as strong as it is you need to use all the tools available to you. Anyway I was not part of the now legendary 1p a click generation - I was online then I was probably just doing something less useful - like command and conquer or writing qbasic. But I am online now, trading and making money. So to get these bums in seats, get people there and looking, snag them into using my sites I use a range of tools - one being pay per click. Pay per click is not by any means my favourite type of advertising. Its effectively well over priced in comparison to the other alternatives, should you know where to look. But for new projects I usually allow a certain amount of play money and some of that gets thrown down the google hole.

But to get the most out of your pay per click, that is spend 6p on a good viewer as a pose to 30p is not actually that hard. First I would recommend diversifying. If your target audience is not essentially british (mine usually are but if yours are not) then you have a vast amount of choice in pay per click sellers.

I would recommend trying a variety of different ppc operators to see which works most effectively for your target audience and keywords. Definitely start with Google Adwords, Yahoo Marketing (£60 free coupon if you click this link :D) and Looksmart.

For me Looksmart provides very cheap traffic, although I am not to sure at this point how relevant it is in comparison with more british targeting equivalents. Yahoo IMO is quite expensive but only relatively.

Google Adwords - if your not paying attention will be very expensive - prepare to pay 80p per click in a mass market for open keywords and still only hit 2nd place. However I suggest the following to get much cheaper adwords clicks (works with all ppc):

Long tail your keywords - that’s it - simple you say - well yes actually. By longtailing my keywords I am now paying 2p-8p for 2nd/3rd/4th position, more specific relevant clicks. This works in most markets, but requires some time and a regular adjustment. Essentially you want to create 3 or 4 word lines in adsense - e.g. instead of “used cars” you want to get “2nd hand cars in uk” - Depending on your localisation (or Hyper localisation) you can then intersect your locations to provide you with a huge list of 3/4/5 word keyphrases. These longer keyphrases will be used less often but as a result will be a lot cheaper. Spend the time to set yourself up with huge long tail lists of phrases relative to your target.

Here’s how I recommend going about getting cheaper adwords clicks:

• Get a database or list of your products/locations/keywords/descriptives. I say get a database because its a lot quicker if you do so.
• Create a php (or whatever language) script to write you keyword lists. Aim to have each line use at least 3 words. Phrases like “used cars west london” rather than “used cars”, other good examples would be “second hand cars west london”, “cars for sale west london” etc.
• If you have a good list of locations you may end up with a very long list of variated keywords. This is only good, just don’t create 20mb files - that’s too many.
• Stick them into google and run them in line with your current broader ads e.g. run “used cars *AREA*” as one campaign in google and another campaign with just the keywords “used cars”. The used cars one will likely get as many hits as you could want, but at a much higher cost. The more long winded list will achieve modest numbers but for a 2nd position it will cost you pennys.
• Do this with all your static keywords, ones that don’t change with time.
• Monitor them - Some will work well others will be no shows, but eventually almost every one will probably be typed in, depending on how long tail your going! Adjust the bid prices in as much detail as possible. If you have lists of thousands build this into the php script to output the bid value on the line *bid*.
• Sticking at double the minimum bid of 3p or 4p will usually get you top 3 position in my experience (with long tail adwords.)

There are numerous advantages to this, but depending on how much your spending it could be more cost effective to just use the broad term. Factor in your hourly wage in creating the files.

If you manage some low cost per clicks - Post them in the comments here - what’s the lowest cost per click you can get in google adwords?

Just 4 quick essential points to remember when running a website - fairly random but essential points for today! Part of a series looking at essential aspects of running a successful website.

1. Create lists - Email subscriber lists help you build your business, lists of people to rep about your business/promote too, lists of what you need to fix/create/re-design, lists of people/ip’s you don’t want on the site, spam artists!

for example we have had spam from 6 different IP’s in the last 5 days, but using .htaccess we can essentially block them for life:

deny from 209.47.94.52
deny from 211.172.225.137
deny from 211.174.63.148
deny from 24.166.249.218
deny from 76.214.29.123
deny from 218.234.21.33

Only problem is that they could be using pooled ip’s or dynamic ones at least - so its best to check them up before they go on your list!

2. Used good opensource software, add plugins - Its good to use opensource software to run a website, wordpress, phpbb, gallery2, oscommerce are all excellent examples of quality releases that will take all of the pain out of running a website. But almost as good as the software for a link is the contributions from the community’s that surround them. SEO mods are plentiful and you can get mods for wordpress that will turn your blog from a blog into almost anything!

3. Write your own additions, plugins, addons - Your opensource software is great - but there will surely be some improvement you can add, but make sure someone hasn’t already done it well as a release. Submitting additions to the community can only help move your website along.

4. Be outrageous - the loud, crazy, quirky all stand out online - ‘another business blog’ won’t attract half as many people as ‘Blogathon - why jesus wore dresses’ (or similar.)

That’s it -four random but essential tips to running a website/e business!

---

Like this little snippet of gold? Read some of my other amazing quick guides:

• Organise your online Reading - Use google reader, firefox bookmarks and be free from wasted reading time
• 10 Steps to submitting your site for a start - ten quick steps to getting your new born site about
• Sandmonkey - Wordpress theme that doesn’t quit

A lot of data can place a lot of load on a server. As I discussed in a post a few weeks ago Data can translate nicely into cash, so if you have a nice good source of data then take it while you can. Free data is great and can be good to feed to pages, but as its free other people are likely to be using it - upping the likelihood of competition, google thinking your replicating and thus lower traffic. Free data not in the public eye is even better. You have users on your site? a lot of them? Then think how to use that - from users logging in and out, registering, chatting - whatever - use everything you can to produce good sources of information.

Anyway in this post I will discuss ways to manipulate, play with and organise data - in large chunks specifically. Lets say if you have a source with 40k lines of detail, with 10-30 fields for each line, and you want to mass update a few fields.

First the format: My skill base covers SQL so I will discuss with this as the basis, MSSQL is great, if you are using that then updates are easy as its presumably all local, good connection speeds and processor/memory control will allow for easy updating and manipulating. In a web situation Mysql is the obvious choice, and always my recommendation. But there are problems that may potentially crop up with using mysql to feed 40,000+ records, edit/update them and generally manipulate that much data usually when hosted by another company, that is running mysql locally is almost better than mssql for data manipulation.

Shared hosting and even quite a lot of virtual servers are heavily restricted in memory, script running times and processor load. There are as ever ways around this.

1. Transfer the database to a local machine and rape the memory/processor/bandwidth like its yours (because it is) then transfer it back up.
2. Get better hosting - a good dedicated server is not cheap - and depending on your office location / IT knowhow it can often be as cost effective to run your own server (e.g london offices.)

or…Trickle updating - When you have thousands of lines to update and you really want to get messy with each line e.g. create long manipulations of all the fields and then update them.

Here’s how I do it:

1. Work out what you really want to do to the data - by planning how you would like it in an ideal way, titles here, dates in a good format here etc.
2. Break the process into as many chunks as possible - if you can - do manipulations in small chunks - more likely to get through shared server load problems!
3. When you understand what you want to do, add a single field to the data table structure - e.g. ‘process_status’ - From here then start a loop updating records doing all you need to - perhaps in steps and then update this field as you go. so for example: 5 steps to do to 200,000 lines, start a loop splitting the 200k into as bigger lumps as doable on the server. e.g. grab lumps of 200 lines from the server and do one step and line at the time and update the field as you go.
4. To get round your shared hosting having limits on how long a script can run for - (often crashes on 1and1 when you run a script that updates loads of lines and takes more than a minute), because you have already split the load into 200’s or a relevant figure (depending on how long it takes to update a single line) simply let it do this 200 and then spit out a meta refresh line. Then you can leave it running over night and once its done each 200 records it will refresh, do the next 200 records and loop until done.

In essence:

• Create a php script that splits the mass of data into smaller chunks e.g. 100, 200 - whatever
• Let the script then spit out a meta refresh line which refreshes the page after a few seconds
• It will then do its chunk, refresh and do the next - until its done.

Php and mysql can get round most shared hosting hold backs - with a little elbow grease!!!!

If anyone wants help with mass manipulating data, sql I can offer freelance data warehousing, manipulation and general processing - from back end to crystal reports.

---

If you like this post on mass data then you may be interested in the following:

• Data = Dollar ? - An analysis on whether data sources can be money sources - Microsites Specifically