This is my first shot at live blogging. I am going to give it a shot at writing and posting in real-time so please excuse any typos etc.
I am here at A4U with Sam and Tom from Distilled talking on a range of topics - if you haven't seen us speak yet, both Sam and I are talking tomorrow morning (competitive analysis and data feed SEO respectively). Also, if you want to see more from the Distilled team, there are still a few tickets left for our Pro Seminar in a couple of weeks.
I am sat in the hall listening to the legend that is Ralph Tegtmeier (Fantomaster) talk about his experiences in building his Link Whisperer service. I would expect that you should not expect everything in this talk (and hence in this post) to be fully white hat :)
## Onto Ralph's talk:
Refresh this page to get updates as I write *aaaand.... I'm done
The negatives of conventional linkbuilding are
If you want to use a third-party network
- it is expensive - you don't have control over the links and where they are placed - if you are paying for the links, it will be sold as (e.g. PR5) when that is just the strength of the homepage and the link you are buying is going to go on a rapidly-disappearing post - you are risking all kinds of penalties
If you want to link build in house
- recruiting and budgeting is hard - training is hard and maintenance costs are high - you still don't have control over the links(!)
If you want to outsource your linkbuilding
- Quality control is hard - If you lose control, penalisation is a risk - you still don't have control over the links(!)
Ralph clarifies that he isn't against black hat stuff (if you hadn't noticed) but that he feels that the risks are not always disclosed to clients. We definitely agree on that point.
On-site factors are getting more important again - it's not all about links these days. But to the extent that you need links, Ralph feels you need to:
- Determine the location of your inlinks(*) - Specify your own anchor text - Know when the links will appear - scale up as you need and change direction as you see what is working - modify your links to accommodate future changes to the algorithms
General point - be careful with your anchor text. Over-optimisation and artificial link patterns are an increasing problem (another point of agreement). You want (some) links that say "click here", that link to your homepage and brand links.
Ralph favours something like 60:40 homepage vs. deep links and brand vs. keyword-rich. Points out that if you are getting lots of links naturally, your linkbuilding can be more aggressive with optimised anchor text.
## Ralph's solution: proprietary linkbuilding network
Here is what you will need:
- An initial competitive link analysis - what are you going to need to compete? - Regular follow-up analyses of the competitive link environment - Proprietary topical websites (this is still over-rated in terms of "link juice" - most people mean "authoritative" rather than "topical") - Unique quality content - Vast range of hosts, IPs and IP C classes (there is only anecdotal evidence that this is needed, but from a technical perspective, it seems that it may be important. In particular, there are no hard and fast rules about how diverse you need to be, but if everything is on one host, you are going to struggle) - No artificial linking structures (in particular, avoiding having your sites linking to each other) - Organic deployment methodologies - you don't want to launch and update them all at the same time on the same day. In countries where you can't use whois privacy, you also need to worry about who registered your domains - Randomisation and structural variance to avoid a footprint - Consistent freshness and natural linking patterns - Linking out to other natural sites e.g. authorities, competitors etc. - Other monetisation plans. Plans might include selling links (carefully!), selling off the whole network - Your network should appreciate in value over time if you do it right
In the US, web hosts will sell you IP addresses for between $1 and $4 / month. This is good business for them as they are essentially free to the host. In Europe, hosts are less likely to sell you IP addresses on their own.
In doing this, you can essentially be flipping domains which can be a source of revenue in its own right. You are helping your core business while building an asset.
## Wrapping up
Ralph says, to finish up:
> If you want to do this, go for it. Analyse the situation and determine what effort and cost it is going to involve. It is likely to take at least a year before your link building network is giving you full value, but if you do it right, it is always going to bring you more value than it costs.
Should you place the link when the article is published or added later?
> Let's put it this way, if you set up a site with 100 pages of content and later add links into that content, I don't think that's a great idea. We will generally publish the content immediately including those links.
But you want central control?
> Yes, you may want to pull some of them later.
Administration sounds like a nightmare. Is there any software we can use to do this?
> For our own network, we have developed some software that automates the upload and distribution of content. But for what we have outlined here, there is no software that does this end to end to my knowledge. You will have to build on what is available or develop your own stuff. This is one of the reasons we are offering this as a service because it is painful. If you have 500 hosts, they are going to want to be paid monthly, they are all going to have different admin consoles, different configurations, different logins. It can be quite an issue to monitor and maintain this in a rational way.
Relative to keeping sites fresh - what would you call fresh?
> It depends. A live blog will probably have between daily posts and 2-3 posts / week. There are exceptions even in live blogs (his own blog is updated ~once a week). You don't want to have patterns - some sites you want to update daily, some weekly, some monthly to avoid having easily identifiable footprints.
Do you have any practical advice for setting up the sites? Do you set them up from scratch or do you buy existing domains and sites?
> We have our own service - silver, gold and platinum. I'm not saying you will want to do it this way - there are many ways to skin a cat (pulls up a slide showing the costs of buying this kind of proprietary network from his company - with costs ranging from upfront EUR145k --> EUR557k and ongoing EUR18k --> EUR46k and showing from 500 to 2,500 websites in the network) [Will - unless I missed it, he didn't answer the question here directly, but the information on that slide was HUGE - sorry I don't have a photo - if anyone snapped one, email me it and I'll include it here]