Leaving The Day Job

Archive for October, 2007

Whether they’re the chocolate kind or the internet kind, cookies are great.

When I was at school a million years ago the most popular kid in the playground was the one who brought the packet of mini chocolate cookies for a break time snack. Don’t ask me why but it was always more satisfying to scrounge a cookie than a handful of Wotsits, Square crisps or Snaps.

Here in the internet age cookies have acquired a new meaning. They’re the little snippets of data a website sends to your browser so it can recognise you when you come back again. They’re essential to affiliate marketing. When a visitor clicks on your affiliate link they receive a cookie from the merchant’s website and then, if they buy something, the merchant uses the cookie data to know which affiliate they should be paying.

The best thing about cookies is that they don’t go away if the visitor doesn’t buy straight away. It stays in the browser and most merchants will pay out if the visitor comes back to buy something at a later date (generally up to 30 days). I’ve been discovering in the last couple of days just how good this can be. I’ve had several sales now where the actual transaction took place weeks after the visitor originally clicked my link. It means that the long barren period earlier in the month where I was making virtually nothing isn’t as bad as it seemed. It also shows that you shouldn’t write off an idea too soon. The customer might have bookmarked the offer to return to after payday.

There are a couple of complications after all nothing is ever simple. Firstly, merchants only pay out for the last affiliate to refer a customer. Kirsty explains it well in her post here so I won’t bother repeating it.

Secondly, some people don’t like the idea of being tracked on the internet so they’ll set their browser to reject cookies or to delete them when they log off.

Third,  some merchants don’t set tracking cookies at all so you only get one chance at converting. Amazon, I believe, are notable for not setting cookies.

I like cookies and I’ll be paying closer attention in future to merchants’ cookie policies. It also backs up my idea that it’s better to concentrate on more niche products from smaller merchants as it’s more likely that the cookie-carrying visitor will return to the same site to complete their purchase.

I’m not having a very productive day so I thought I’d make it even less productive by spending a few minutes pointing you towards my top three funny things that I like to look at online instead of working.

Dilbert

Every geek’s favourite anti-hero. Offers me some relief from the woes of the day job.

Basic Instructions

Currently my absolute favourite online comic. Updated once or twice a week, Basic Instructions provides witty, sarcastic and surreal instructions on how to do all manner of mundane tasks.

I Can Has Cheezburger?

The internet meme of the year. Pictures of cute animals captioned with bizarrely syntaxed comments.

Enjoy. By the way, shouldn’t you be working?

Techcrunch is reporting that Microsoft has bought a share of 1.6% of Facebook for $240 million. The deal values Facebook at a stonking $15 billion.

Microsoft already has the gig selling advertising to Facebook’s US users and this investment will see that deal expanding internationally. Assuming Microsoft are going to tie this to their AdCenter product this might finally give the volume of impressions and clicks that they need to start attracting some serious ad spend away from Google.

Facebook is attractive to advertisers for a number of reasons. Firstly, the sheer number of users is huge. Secondly, Facebook by definition has lots of demographic data about their users making ad targetting much more accurate. (Microsoft already allow AdCenter advertisers to pay a premium to target certain demographics based on users MSN profiles). Thirdly, Facebook stills has its college roots and this late teen-early twenties demoographic is an ideal market for high value shiny gadgets (think iPods, mobile phones etc)

The question though is whether Facebook users will actually click ads. Anecdotal evidence suggests that clickthroughs on social network sites in general is poor. Users are there for the social element not to go shopping.

You also have to ask if this is more evidence of a second dotcom bust approaching. Is Facebook really worth $15 billion?

I’ve been trying to be an affiliate tycoon with varying degrees of success for four months now so I thought I’d put together a list of some of the things I’ve learned so far.

1) Don’t follow the herd

If there’s one thing I keep hearing over and over again it’s “find a niche”. Affiliate marketing is getting crowded and more sophisticated. Steer clear of the big shiny commissions that gambling, mobile phones and loans offer. Many first time affiliates go down those roads and quickly get disheartened when they find their site at number 10 million on Google for “cheap mobile phones”. Find a niche for yourself when you’re starting out. There’s a lot to learn before you can try to take on the big boys.

2) Read blogs

There’s a wealth of knowledge out there on the net for free in the shape of successful affiliates’ blogs. Take advantage of their tips. Use a feed reader or aggregation service so you can check them all in one go. Try Affiliates4u blogs or internet marketing blogs for a quick start.

3) Don’t be seduced by big name merchants

It can be reassuring to see a well-known high street name in amongst the huge list of merchants some networks have. Brand recognition counts for a lot and it might make your affiliate links more attractive to your users but generally the big name merchants pay the lowest commissions. Also there are more affiliates promoting them so customers who are shopping around are more likely to pick up a competing affiliate’s cookie before they buy. Smaller merchants are likely to pay higher commission and to be more niche in their product offerings.

4) Blog

Writing your own blog helps to channel and focus your thoughts and helps you to build networking connections with other affiliates. Try to be honest and interesting and don’t use your blog as an excuse to push affiliate links - other affiliates will see straight through you. A good blog will attract inbound links which will help build Pagerank; you can then link out to your affiliate sites and pass on some of that Pagerank.

5) Keep a very tight grip on PPC

PPC advertising (Google Adwords etc) can be a great way of delivering targetted traffic to your sites but it can also burn through a silly amount of money very quickly. Make sure you set a daily spending limit you are happy to lose if it all goes wrong. Don’t just add every keyword in the world on day one. Focus on one product and keywords specific to it. See what works and build out from there.

6) Don’t be afraid to make mistakes

You will not get it right first time. You will make mistakes. Don’t let it get you down. That which does not destroy you makes you stronger.

7) Don’t expect anyone to understand what you are doing

Outside of affiliate marketing nobody will understand what you are spending your time doing. You can show them your sites and they still won’t understand. They’ll ask if it’s legal. They’ll ask if it’s like spamming. Most annoyingly they’ll ask you to show them how to do it cause it sounds like an easy way to make money. Come up with a stock answer to their questions which is suitably dull and vague that they’ll lose interest and leave you to get on with it.

8 ) Keep proper records

You might not think you’re earning enough to be bothered with proper records now but you need to get into the habit before it’s too late. You don’t need an accountant when you’re just starting out but make sure you keep copies of all invoices and receipts and keep a simple set of accounts in a spreadsheet or a package such as QuickBooks. Register with Inland Revenue as self-employed. Read the tips at High Royd Business Services

9) Learn how the different networks and link types work

Each affiliate network is different. They have different portfolios of merchants, different tools, different payment procedures and different levels of support. Take time to understand each one. Four months in and I’m still finding features I didn’t know existed. Even if you don’t understand what all the different link types and content units and so forth are now at least you will know that they exist so when you find a need for them you’ll know what to look for.

10) Don’t expect to make a million overnight

I can’t emphasize this one enough. It takes time to be successful at anything. You won’t make a fortune overnight. I thought I’d be able to quit the day job within six months when I started; now I’m thinking one year or more like two and only then if I work really hard and have more than a fair amount of luck. This is the toughest of my rules for me to stick with. I get really frustrated if I can’t do something the first time I try.

It can be done. But the odds are against you. You need to work hard to succeed.

Yesterday I reviewed three solutions for making your blog accessible from mobile devices. One of these was MoFuse and I highlighted a problem I’d had with my RSS feed being truncated. I’m pleased to say that MoFuse have got back to me today to say they’ve identified and fixed the problem.

You can view the MoFuse version of my site here and I’ve also got ten invites to join the beta program to give away. If you’d like to receive one just request it in the comments below and I’ll send by email.

I have a rather lovely Nokia N95 mobile phone which sports a powerful (if memory intensive) web browser. I often read web sites on it during my lunch break and I’m always pleased when a website automatically downgrades itself to a mobile-friendly version when it detects my phone.

I wanted the same thing for this blog and thought I’d found it when I got a beta invite for MoFuse via an article at Read/Write Web. MoFuse takes your RSS feed and dynamically generates a mobile version of your website. Unfortunately they don’t know how to parse the RSS from Wordpress properly so all my posts ended up truncated at 80 or so words. After several emails back and forth to their support desk who insisted the problem was with Feedburner despite me sending them the raw XML from Feedburner which clearly has the full post in it, I gave up.

Next to try out was Winksite. This is basically the same idea as MoFuse (and indeed predates MoFuse). This worked out fine and was able to display my posts in full (using exactly the same RSS feed which MoFuse failed with). It’s also quite customizable in terms of design and adds various social network type features. My main problem with it was that I couldn’t see any way easily to integrate with my existing blog. I want people to be able to type in the same URL on whatever device and see the correct version of my site.

Finally I found Andy Moore’s Wordpress Mobile Plugin. It’s a standard Wordpress plugin which as usual is a doddle to set up. It does the automatic detection that I wanted and also has various customization options including the ability to post from your mobile and to add advertising to your mobile blog. You can repay the author for his hard work by configuring a revenue share percentage within the plugin (I’ll do this once I figure out how the AdMob advertising platform works, Andy!)

So there you have it. Next time you’re out and about with your mobile or PDA and have a few minutes to kill check out the mobile version of my blog at the usual address : www.leavingthedayjob.com

I’ve been playing around with openads. It’s an incredibly powerful ad delivery platform which is open source and available for the princely sum of zero. The best feature for me though is the ability to target your ads depending on the geographic location of your site’s visitors.

Like most UK affiliates I’m primarily promoting UK merchants to UK customers. With PPC it’s easy to make sure that you’re only bringing in UK visitors but if your SEO is good enough or your content compelling enough you’ll soon have visitors from all over the world. These international visitors will most probably not be able to take advantage of the offers you’re promoting so you won’t be making any money from them.

I recently discovered that a long neglected site of mine was still getting a fair bit of traffic from organic search and that the majority of the visitors were US-based. I do use Adsense on the site which automatically handles showing the right ads to the right international visitors but I also have some image banner space and I wanted to target appropriate ads to those US visitors - they tend not to have much need for Sky TV in the States.

Openads is a PHP/MySQL solution. Installation is relatively straightforward with a wizard which steps you through setting up the database. If you’ve done a Wordpress installation you should probably be OK.

Once you’re logged in things can look a bit daunting. Openads is able to handle the ads for very large networks of sites and so there’s more features there than you need. Openads refers to each site as a publisher and each banner space within that site as a zone. I just set up one site and one zone within it to start with. You tell openads the size of each zone eg 468×60 so it knows which ads can fit in which space. You get a choice of “invocation codes” to paste into the appropriate place on your site as you would for Adsense or any other banner. The choice of code types is a bit confusing but I found IFRAME worked best on my sites.

On the other side of the equation you set up advertisers and the campaigns for each advertiser. I set up Amazon US as an advertiser and just created a default campaign. Then you add one or more banners within the campaign. Banners can be images, text or HTML. I pasted in the HTML that Amazon gave me.

You then assign banners to zones. As I only had one banner and one zone it was a simple case of linking the two together. If you’ve got multiple sites though there are loads of options and permutations for deciding which ads go on which sites.

After setting up a second advertiser of Amazon UK with a UK centric banner it was time do the geotargetting magic. This bit’s not very well documented. You need to download the free GeoLite Country database from Maxmind (get the binary format) and copy it into a directory on your website. Then in the openads geotargetting settings page enter the location of the data file in the ‘MaxMind GeoIP Country Database Location’ box. This should be the filesystem path eg /var/www/html/GeoIP.dat not a web address.

Finally I went back to my Amazon US banner and set a Delivery restriction by selecting ‘Geo - Country’ and picking the USA. Now my site is displaying an Amazon US banner to US visitors and an Amazon UK one to everyone else.

Now this may seem like a lot of work to achieve something relatively minor and you may wonder if its worth bothering given the appallingly low banner clickthrough rates. But openads is much more powerful than just a banner server. The ability to use HTML banners means you can effectively target any kind of content to visitors based on a number of criteria. For example, if you are building a postal mailing list you may wish only to show the sign up form to UK visitors. Or you might want to show adult offers late at night and family friendly ones during the day.

I’ve barely scratched the surface of what’s possible with openads. I’m sure I’ll find many more features as I play around with it some more. I also haven’t tried integrating it with this blog but I’ll have a go real soon now.

Well, it’s going to be a long time until I can go fulltime as an affiliate but I took the day off from my “proper” job today to try to get a good chunk of affiliate work done. I started out the way I always imagined working from home would be like by sleeping through the alarm and getting up an hour and a half later than intended.

I moved on to reviewing stats, reading blogs and answering emails over a cup of freshly brewed coffee then settled down to have a long, bitter and swearword-infested fight with PHP and XML. I was trying to convert the XML from Amazon’s web service into a nice CSV file to use with Affilistore. I quickly learned that Amazon provide you with easily enough XML rope to hang yourself with. There’s far more data in their feed than I could ever need and so I spent ages trying just to figure out which bits of it were which. Then my rustiness with PHP paid me a visit and I spent a similar number of ages referring to the manual trying to work out how to manipulate said XML.

In the end I got something not entirely dissimilar to what I was trying to create but was not at all pleased that I’d wasted half the day trying to get it. It’s my intention to create a more general Amazon to CSV tool to put on this website but it’ll have to go sit on the backburner where so many of my other bright ideas are waiting. As much as I love Linux my years spent coding .NET at the day job mean development would be a heck of a lot quicker if I just bit the bullet and switched to hosting my sites on Windows.

By this point it was lunchtime already. In the day job, lunch usually marks the point at which my productivity takes a serious downturn but working at home I’ve actually got more done in the afternoon than the morning. This is a good sign as it kind of suggests that working for myself suits me.

I’ve discovered that one of my old, neglected sites is still seeing a fair bit of traffic and that most of it comes from the US. So I’ve installed openAds to rotate the banners on the site and geo-target them to the visitors. I’m thinking of revitalising and relaunching the whole site since people seem to still like visiting it.

I’ve also set up a PC at home with Fedora Linux so I can do development work, launched some PPC campaigns and done some tweaks to another one of my sites. All in all I feel I’ve been quite productive and I’ve certainly got more done in one day than I would have done in a week of just doing affiliate work in the evenings. Unfortunately you don’t get quick results in affiliate marketing (at least I don’t) so it’ll be a while before I see if my hard work pays off. In the meantime it’s back to the day job tomorrow - only four days until the weekend