Misleading companies – part 3

The final part of my rant (see part 1 featuring ConnetU and part 2 featuring IX Reach as well as euNetworks) is all about C4L (A.K.A. Connexions4London), a company who are trying to take the take the practice of claiming to have PoPs which don’t really exist to a whole new level!

I should start this with a disclaimer. I have been a C4L customer, I have used their network services, I have experienced their technical support and their customer service. Equally, I have had both customers and friends who have used C4L’s services. Between these various experiences over a significant amount of time, I can safely say that I will never be a customer of C4L or knowingly use their services ever again.

Every time I have had anything to do with them, they have managed to reinforce my impression that they are an awful company with an old, unreliable, congested network, unhelpful staff, slow support and non-existent customer service. Suffice to say, I may be considered somewhat biased. Anyway…

On the front page of their web site, C4L state:

C4L is a Colocation, Connectivity, Cloud and Communications provider headquartered on the South Coast, providing access to over 100 UK data centres and more than 300 globally.

Now, this is not the only misleading thing on their web site, or even on the home page, but as the other items aren’t specifically relevant to this rant, I’ve decided to leave them out (for now).

These impressive sounding numbers of over 100 UK and 300 global data centres appear repeatedly on the C4L web site as well as in their email newsletters, press releases and other marketing literature.

The thing is, that C4L don’t have anywhere near that many Points of Presence (PoPs). If you take a look at their network map (http://www.c4l.co.uk/resources/data-sheets/Network-Map-Connectivity-Data-Sheet-C4L.pdf), then you can very quickly see that the actual number is much smaller:

• Telecity Reynolds House
• Telecity Williams House
• Telecity Kilburn House
• “Manchester 1” (probably the M247 Ball Green data centre)
• “Birmingham” (god knows)
• “Derby” (Node4 Derby DC1 and DC2 – basically the same building)
• “Enfield” (Virtus London1 Enfield)
• “Milton Keynes” (Pulsant Milton Keynes, formerly BlueSquare MK)
• “Slough” (Virtustream)
• Telecity Soverign House
• Level(3) Goswell Road
• Telehouse Metro
• City Lifeline
• InterXion
• “Brick Lane” (probably Easynet)
• Global Switch 2
• Telehouse North
• Telehouse East
• Global Switch 1
• Telecity Meridian Gate
• Telecity Harbour Exchange 6&7
• Telecity Harbour Exchange 8&9
• Telecity Bonnington House
• “City Reach” (Tutis Point City Reach, formerly owned by QiComm, now Docklands Data Centre Ltd – DDCL)
• “Greenwich” (the former BIS Anchorage Point data centre, now owned by 6dg)
• “Bylfeet” (4D Data Centres Sirius II)
• “Park Royal” (probably Telecity Powergate)
• “Maidenhead” (Pulsant Maidenhead 1-3, formerly BlueSquare House and BlueSquare 2-3)
• “Bournemouth” (C4L’s own data centre with their offices at County Gates)
• Evoswitch Amsterdam
• “Isle of Man” (Either Netcetera or Wi-Manx)

Now, that is still a fairly impressive PoP list, but at 31 data centres it is a lot less than the 100 that they are claiming in the UK, let alone the 300 that they are claiming globally – particularly when you consider that there’s only a single international data centres on there, and that is Evoswitch in Amsterdam!

This all comes back to my earlier question – what is the point of climbing more PoPs than you actually have? What do you stand to gain from it?

If you do somehow manage to deceive a customer into taking services from you at a datacentre where you don’t already have a PoP, then the likelihood is that you won’t be providing any kind of useful service – you’ll just be buying a single rack from the data centre, adding a markup onto it and passing it on to the customer. Now the customer has to go through you for access requests and remote hands, slowing them down and adding an unnecessary layer of complication.

Likewise with connectivity, you probably aren’t going to be able to build out a full scale PoP for a single customer, so you’re just buying a circuit from someone who does actually have a PoP in that data centre and then backhauling it to somewhere on your existing network. This adds a potential point of failure as well as complicating any troubleshooting as now three (or more!) parties are involved in any investigations.

In both instances, the customer could have got the same (or better) service from going direct and it would have cost them less. So all you have done is made it more expensive and less reliable for your customer!

There is no problem with reselling services, as long as you are adding some value to them. In these cases however, it seems that the only value being added is to the reseller’s bank account!