You can view the slide set here.
What I’m going to talk about today is what we at CTT have
been doing in developing our IT infrastructure. In particular, how we have
moved to a position where we can operate the whole organisation from our homes
i.e. without needing to come into the office.
The original driver behind our approach was our disaster
recovery strategy. Basically asking the question, if we lost the office, or we
couldn’t get access to it, what would we do?
We are a technology charity. Everything we do relies on
technology. So the question of what would happen if we lost the office, or
access to it, is a central issue for us.
While we were thinking about this, a couple of events served
to focus our minds a bit more sharply. We are based in Old Street on the
north-east corner of the City, and while the riots last year never came that
close to us, we had to take into account the travel needs of our staff. Were
the riots going to put them in danger as they travelled to and from the office?
Were indeed, the riots likely to engulf the area we are in – one afternoon, the
police did advise that that was a possibility. Were the outworkings of the ‘Occupy’
and other protests in the City likely to make reaching the office potentially hazardous?
Then we started to receive the information about the
possible effects of the Olympics on London’s travel infrastructure and the
possibility that our question “could we lose access to the office” became a
serious reality.
The other side of this question has also been a desire to
provide a flexible working environment for our staff. Like most organisations
in London, commuting times are a real issue – not to mention transports
strikes. Like everyone else, we employ part time staff and staff with young families.
Being able to offer a more flexible working environment and the ability to work
at home once or twice a week becomes a differentiator when we are trying to
attract staff – we can’t afford to pay high wages to compensate any more than
anyone else in the sector.
So developing an infrastructure that allows us to run the
organisation remotely seemed like a good aspiration to inform our planning.
After all, we are a technology organisation so trying some of the stuff we
offer to others seemed like a good thing to do as well.
Now this hasn’t been an overnight change. It’s been a
journey and what I want to share through the rest of this session is to take
you on some of that journey. Let me stress here that this is our journey. It’s
not the only way, but it’s the way we chose and I hope it will give you some ideas
and insights that will help you on yours.
So let me take you back four years to when I joined CTT. We
run IT services and the live systems that deliver these run through websites –
so they were already ‘in the cloud’ and I’m not going to argue definitions of
that term because I think it is so broad and so abused as to be next to
worthless. I have been involved in running business applications that were
delivered through the internet since 1998. What we see now is the outworking of
those first applications as the technology to support them has moved on. I
would urge you to forget about ‘cloud’ as a term and focus on the applications
and (if you develop your own applications) the platforms and evaluate them for
what they are, not because they’ve got some trendy marketing term attached to
them.
What I would say, however, to declare my bias as much as
anything, is that I spent many years measuring the lifetime cost of technology
– not just the upfront investment, but all the costs, including all the people
costs both of IT professionals and the impact on non-IT staff – and not having
to manage servers and all the paraphernalia that goes with them, is a big
attraction for me of the “cloud” model.
But as I say, our main client applications ran through
websites, hosted by major providers and developed to deliver our day to day
business. Our internal systems were another matter. Email, accounting system,
file storage all ran on a server in the office under the protection of a large
rubber plant. Yes the server sat in a corner and the rubber plant stopped
people falling over it. Backup was a manual tape and the process of checking
around that was poor to non-existant.
Well that got fixed. We invested in a proper server &
automated backups and also Remote Desktop Connection through the Windows server
that sat on it. At the same time we moved the hosting of CTX after the then
provider told as they had a disk problem (something we had been telling them
for three months) and that they had just taken it down and we could expect it
back if everything went well in 48 hours. This was at 10am one Tuesday morning!
Lesson one, if a hosting provider doesn’t
provide a professional service, move.
So that was the starting point we were at when looking at an
‘officeless’ operation. We had access to operational systems and remote access
to files and email.
And actually, that’s pretty good. It means that staff can
work away from the office. However, we were in a bit of a cleft stick. If we lost
access to the office (e.g. the internet), then inside the office you could get
at the file server but not the operational systems; outside the office you
could get at the operational systems but not the file server; and email didn’t
work for either (unless you wanted to send an email to a colleague who was in
the office with you - we’re only 11 people who sit in the same area, so that’s
not really a big benefit.
The other thing from my point of view was that servers
require support and maintenance (we have lost a few disks). And remote desktop
under Windows Server is not a great environment to use for any length of time.
In particular, it doesn’t provide for any ‘connectionless’ access and is very
awkward to use to download a file to the desktop. Essentially you have to work
inside Remote Desktop which I personally find very clunky.
So there are three areas we really set out to work on to get
to “Officeless”:
- The phone system
- File sharing and management
Phone System
We run a number of client services all of which have some
form of customer service line. So in running in an ‘officeless’ environment,
being able to manage these effectively is essential. The answer we have opted
for is to install a Voice Over IP phone system. Because of our relationship
with Cisco, we were able to get a full system donated. That’s not always an
option and there are a number of other VOIP services that are available. I
haven’t evaluated them because CISCO donated their system to us but, as I said
before, this is just our way and not the only way.
What does the VOIP system provide us with? Well a full
function phone system that can be extended out to our staff’s homes either on a
physical phone or softphone on the PC. In effect this puts an extension on the
network in those locations.
Our phones are all configured to show all the major service
incoming lines plus the owner’s number. We regularly run with customer service
staff working from home (most staff work at least one day a week from home) and
they are able to answer the customer service lines along with the team in the
office.
We’re not overwhelmed by calls so this works fine for us.
Essentially we have a code of practice so that everyone knows who is answering
and who is there and we use IM to manage this.
Email
At the beginning of the year we moved to Office 365 and now
use exchange online. This was a bit of an adventure, but I have to say, I’m
really pleased with it. It has certainly met our needs and made administration
of email a lot easier.
I think the biggest challenge was migrating the mailboxes
from the on premise exchange server into the cloud server.
One of the key things in moving to an “officeless”
environment was that we invested in a 10MB SDSL internet connection to the
office. That means we get 10Mbps upload speed as well as download. For an
individual at home, we’ve found that a standard ADSL line from the usual
suspects is fine. But we felt that as we’re putting all our eggs in the
internet basket we should invest in as much bandwidth into the office as we
could afford. To date we’ve had no significant issues.
Office 365 Exchange Online offers a coexistence mode to help
migrate from an on-premises Exchange Server. It does a lot of account
synchronisation and other fancy things. However, it’s complicated to set up as
you need to install modules onto your existing server, so we went for a
straight ‘let’s move everyone as fast as we can’ approach. It’s actually
possible to configure things quite well so that those still on the on-premise
server don’t really notice that other users are now using Exchange Online. It’s
just a question of setting up a few temporary workarounds and planning which
groups of people we should move when.
For us, and as I said we’re 11 people, moving fast was a
much better option than trying to be clever! As I say, the really issue was
moving some very large mailboxes (9GB was the biggest). We also decided to move
all the data then run the archiving functions to reduce the amount of data that
needs to be synced with
Outlook, but that was just our choice and you could opt
to archive old mail on the on-premise server and just move recent stuff. But
running the archiving so the very large mailboxes got shrunk and we now sync a
lot less data was a key part of the tidying up that we wanted to do.
In terms of use and management, I really like Exchange
Online. In particular using email through the browser has moved on a quantum
leap. For anyone who doesn’t want to be synchronising Outlook at home, it is
almost as good. Being able to open other mailboxes for which I have permission
e.g. shared customer service mailboxes is really straightforward and again
makes the ongoing management much easier.
Day to day admin is pretty easy. The Admin console’s very quick
to pick up. However, during the setup, I was surprised at how quickly I got in
to using Powershell to do things, especially if I wanted to do things across
all users.
File sharing
This is the final piece for us. And we’ve opted to use
Sharepoint Online to manage it.
We have a lot of old data on the file server – as most people
do. So we are doing a cull of what we use on a day to day basis and moving
those directories into Sharepoint libraries. We have had some fun getting the
local data sync to work, particularly from one user’s point of view as they
want everything on the server to be available offline on their IPhone! But
we’re getting there.
What this gives us is effectively a server with control of
who is able to access what while enabling local file storage for users like me
who are constantly on a laptop.
What it avoids is large volumes of files finishing up on the
local drives of our staff’s home PCs. When they are working from home, they
work with Email through a web broswer and Sharepoint Online, again through a
web browser. I can configure the web view of a library in Sharepoint as a
Windows Explorer file site, so if I attach or save documents, I can do it
directly into Sharepoint.
However, for anything that is client data, we attach that
directly to the CRM record in our operational systems which are already hosted
outside the organisation. That means it isn’t going into that file store anyway.
Other
Talking of client ‘paperwork’, we also invested in a
multi-purpose printer/fax scanner which scans to email, so that documents go to
the relevant users email as an electronic document and can then be filed from
there.
We also subscribed to fax-to-email for incoming fax, which
means that all fax documents arrive as email and can be filed appropriately as
well.
Finally
You will note two flaws in the plan if we lose access to the
office. One is that the phone is still physically in the office. So we are
moving to non-geographic numbers as our main contact numbers. These can be
rerouted very quickly. So we lose the office entirely, we can divert them to
another number completely.
The other is that the postman still delivers letters to the
office. So someone needs to go in and deal with the post from time to time
(which includes cheques so it is very important).
Does it work?
Can we really operate with no one in the office? Well today
is the big day. No one is there. We’re doing it to see if we can run for most
of the Olympics without people in the office. And, so far, my phone hasn’t lit
up with panicked calls, so, so good!
Learnings
I guess the main one has been trial and test. The technology
to do this is out there. For many charities, there is a lot of accessible
pricing on it.
There are choices, we did it this way. It’s possible to do
and, for us, it works well. But you should choose the solutions that work best
for your users and for you from a technology management point of view.
I think investing in bandwidth is going to be key if you
have a number of people in one place. As I said, a domestic ADSL is adequate
for a single user, but I wouldn’t push too many people through that sort of
pipe.
I would urge you to identify the full costs of your technology.
One of the things that I observe in this sector is organisations make an
investment then run it into the ground. There are costs associated with that which
people typically ignore. If I can give you an analogy (and please don’t push it
too far) but a fifteen year old car is a lot less fuel efficient that a new
hybrid. You need to ask yourself the question, is it time to move on.
The other
point with “cloud” solutions is that they run on a rental model. So the upfront
investment is a lot lower (you still need to invest in implementation), so it’s
not so hard to change - compared to buying a new server. But what they do make
much more explicit are the ongoing running costs which many people ignore or
neglect (to use the car analogy, they never get it serviced).
The rest are the classic managing technology change issues which
revolve around taking people with you, helping them change the way they work,
and thinking about the tools and training from their perspective.
It’s also a great opportunity to review all the emails and
data you have and cull a whole load.
Future
One of the areas we are really investigating closely is
managing our customer facing work much more from within our CRM systems. So
sending emails to clients from within the CRM system and directing mail back
into it. This has the major benefit is that we finish up with everything
related to that client sitting in the CRM system, not spread across email,
document libraries and hard copy folders. That’s the next adventure.
