Outdoor clothing and equipment company Kathmandu was in the process of putting an additional chunk of its IT into the cloud when the February 22 earthquake hit its Christchurch head office. Most of the company's processing is already in other centres. "We run the point-of-sale systems out of Melbourne" so sales were not interrupted, says CIO Grant Taylor.The ERP, based on Lawson's Movex software, was hosted by SAS IT in Auckland. Kathmandu is currently upgrading to Lawson's latest version, known as M3.Since he joined the company in September last year as its first CIO, Taylor has been working with Gen-i to move the company's core systems into the cloud and set up a full infrastructure-as-a-service environment. "That Gen-i link came up the day before the second quake," he says. "Our head office was only a few kilometres from the epicentre and took a hell of a knock," he says, but fortunately the server room didn't suffer a lot of damage.The parts of the system still in Christchurch provide a crucial link between sales and financials and other basic services such as email, he says. "Once we got power into the building, it only took us a couple of days to get our integration and our core server environment -- our email and intranet, comms, firewall and remote access -- up quite quickly. From 2pm Thursday afternoon, from an IT perspective, it was business as usual. "Because of that infrastructure-as-a-service model we have and are continuing to move to, the stores could continue trading and the business wasn't impacted at all, really," he says.A conventional hosted environment or even an environment sold as "cloud", but based largely on one datacentre, can still have a single point of failure, he says. "But there are technologies out there now based on virtualisation and I think you'll see a lot of businesses in the future lean more towards a virtualised, replicated infrastructure that lives in the cloud. As opposed to having single points of failure in physical datacentres, the trend is now going to be to have replication for the purposes of performance, continuity and disaster recovery." SAS IT has facilities for disaster recovery in Auckland, he says. "We've been fortunate that we never had to failover." The IT side of Christchurch-based businesses has been impacted in various ways. "No two stories are the same," says Taylor. "I saw someone coming out of an office with two computers under his arms. His business livelihood was obviously on those machines; they were the most important thing he carried out of the building."All businesses in the future will have that challenge, he says. "How do they make those systems highly available from anywhere and avoid that sort of thing in the future?"Moving towards that objective has been part of Kathmandu's plan "but I think there's much more urgency from our business now to invest in IT and meet that challenge," says Taylor.Most of the company's systems are in the cloud now and over "the next few weeks or months" the rest will be moved to the same model.Glidepath, headquartered in Auckland, runs an international business in machines for handling airport baggage and cargo. The company has offices around the world and manufacturing facilities in Auckland and Dallas, Texas, and has completed more than 550 projects in 60 countries.With disparate systems in the different offices globally, and a lack of external access to company data and emails, Glidepath recognised the digital environment for collaboration among its staff could be improved."The cost of telehousing the Exchange 2007 server we had in place was rising and we needed to upgrade our hardware," says Glidepath's chief financial officer Wayne Brown. So it made sense to investigate a new way of doing things that would add to the effectiveness of co-working and save Glidepath money, as well as accommodating future expansion he says.The company has started cautiously by entrusting its email to Microsoft's Business Productivity Online Services (BPOS), working through Microsoft partner Code Blue.Previously it was not possible for Glidepath staff in remote locations to collect their email securely, without setting up complex and costly virtual private network links. Under BPOS emails are encrypted automatically and transparently. Glidepath says the implementation also includes Microsoft SharePoint Online for intranet portals and document sharing, but Brown says this has not yet been implemented, "It's just email at present". The company is taking further development slowly and waiting for the first stage to prove itself, he says.Cloud email has only been in for five months, he says. Collaboration through shared documents is "one area of possible extension for the future."The email implementation has resulted in definite cost gains and operational gains, Brown says. Staff can access emails wherever they are, though performance varies with location.
When life insurance company Sovereign took its first steps into the cloud four years ago, with salesforce.com, the company saw the move as a potential remedy for the "business pain" of scattered information on the agents who sell Sovereign's policies, says Sharron Botica, general manager of customer experience. Information on agents was scattered through 18 legacy systems, she says, and it was a lengthy job for an analyst to draw the threads together to show the comparative performance and history of agents.Sovereign had an in-house system that performed some consolidation, but "that wasn't agile enough", Botica says. The company was not coming to salesforce.com entirely "cold"; its parent, the Commonwealth Bank of Australia already had about two years' experience with the company's service. Sovereign did comparative assessments of functionality against other solutions, but the ability to leverage CBA's experience was an additional factor in Salesforce.com's appeal.There were a few "teething problems" in the early stages of implementation; the company wasn't getting the response times it had been led to expect; but the snag proved to be located in the company's own network and not at salesforce.com, she says.The consolidation of its agents' details was completed in 2009 and now all details are readily available in one place.Some customisation of the solution was required at that stage, but since then the service has evolved and there are more options available "out of the box" to cope with Sovereign's ways of operating, Botica says.Encouraged by the initial success, Sovereign put its claims-handling system in the cloud with Salesforce.com. "That was a one-and-a-half-year project and went live last year," Botica says. Again, putting the system into the cloud allowed Sovereign to centralise information that had previously been difficult to access or even at times proved to be contradictory. "Now there is one version of the truth," says Botica. Turnaround time has improved and the company has the complete history of a claim and related claims from the same customer on hand. This year the company plans to put new business -- sign-up and processing of any new clients -- into the cloud, leaving only maintenance of current business in-house.As a developer and provider of treasury management software, CNS Treasury has a record since 2004 of helping its clients to hedge their financial risks. Last year, "we decided to hedge our own risk", by moving fully into the cloud with Telecom and Gen-i, says team leader of product development Dean Forster.The company has been offering a hosted service to its own customers since its inception, so the change to Gen-i's ReadyCloud service was a question of moving an established SaaS business to infrastructure as a service (IaaS) mode.The transfer was, in fact, done in two phases. The service began in 2004 on equipment and software at CNS' own premises. In 2006 this was moved to a hosted service at Gen-i, but using dedicated software and hardware. The switch to IaaS means the environment is fully virtualised. Implementation of CNS's service with a client is now a much smoother and faster process, as compared with the pre-2006 situation, Forster says. "It's much quicker to get the software up and going and we can roll out changes much more swiftly." Conversation with customers is more about availability and security, than the technical territory of installing and maintaining software.Having the entire infrastructure housed at Gen-i means "a much stronger guarantee" of redundancy and recovery capability, than if CNS were hosting the application itself, says Forster.However, the change from SaaS to IaaS has not conferred any material advantage, he says, though it clearly gives Gen-i more flexibility. "We were told, 'this is the way we're doing it now; come on board'," he says. CNS made the shift initially thinking it would solve a periodic response-time problem, but it has made no visible difference. However, comparing the cloud in general (post-2006) with the in-house version, Forster says, CNS staff no longer need to be so concerned with mechanical maintenance, so can be more committed to front-line support of the customer from a business and service-quality point of view. Their job is now a matter of risk management, rather than day-to-day application of maintenance skills. Security is a big issue with CNS itself and its customers. "We needed a partner that had a robust solution in that area and someone who could keep up with the developments in security threats." This is another task that the cloud has taken off CNS staff's plate; though a periodic independent security audit of the system is performed to retain CNS and customer confidence.Having a local provider who can be consulted quickly in case of difficulty or degradation in standards is important, again for retaining customers' confidence. This aspect will perhaps not be quite as critical as CNS expands its services overseas, starting with a planned venture into the US some time in the next two years. Customers there are more familiar with cloud environments in general and trust them more, Forster believes. The company already has one client in the US running with the ReadyCloud environment. It is still looking for a cloud host in the US to service that market; US customers are more likely to trust someone in their own territory than processing in New Zealand, with uncertain long-distance telecommunication, Forster says.However, it is a basic principle of the cloud; that all the data is in one place and can be accessed from anywhere. Processing at two sites "breaks that model" and may present difficulties in marketing the service to companies sited in both countries, he says.The transition to IaaS fulfilled expectations, Forster says; there were no snags specific to the environment. "Of course there are always challenges in any new environment, but you will get some of those irrespective of what kind of transition it is."Firm service-level agreements are necessary. There are, he points out, applicable internationally-derived standards such as ISAE NZ 3402, for quality assurance and quality control in service of virtually any kind, with such standards applicable to computing service in the cloud mode. Customers and sales and presales staff offering cloud services should be acquainted with these standards, he says.In addition to formal standards, there is no substitute for visiting the supplier on its own premises -- "to see the footmarks in the carpet", says Forster.Personnel assessment company Aspeq's first attempt at cloud processing, with Gmail, was so unsuccessful and disruptive of work relationships "that for a while I was a bit uncertain whether the [IT team] would be allowed to go into the cloud again," says acting CIO Andrew Jamieson.The company delivers in-depth assessments to help organisations test personnel for positions requiring a high level of skill and responsibility. As the New Zealand-based firm has grown, it has expanded into new industries and regions, with operations in Australia and Singapore and small outposts in Mauritius and Macao. Framing assessment material demands close co-operation between staff writing the material and subject-matter experts. This was working tolerably under an in-house system based on Microsoft Outlook and Exchange, but taking the application into Google's Gmail cloud service made it impossible for the staff to use the shared Outlook folders that had been a key to their way of working. The change severely reduced the team's collaborative potential and decreased performance, detrimentally affecting productivity, Jamieson says.The second, more successful attempt utilised Microsoft's Exchange Online. Aspeq migrated its email environment to Exchange Online in March 2010, with the company also upgrading its computers with Microsoft Office 2007 and the Office 2010 suite. As Exchange Online is designed to operate seamlessly with Office Outlook 2007 and Outlook 2010, Aspeq could deliver the familiar email experience that its employees wanted, while providing enhanced features like Conversation View in Outlook 2010. Staff use this to view all the messages in a conversation thread as a single item.Getting reasonable commitment from the provider is one of the battles of cloud computing, he says. "A number of security issues are involved, particularly in serving a company like ours. I would like to see more commitment [from cloud providers] as you need to know who is looking after your data and what degree of commitment they have to its custody and security and to the availability of their service." The major difference to Aspeq's business arising from implementation of the cloud is a substitution of an outsourcing agreement for investment in in-house equipment and support staff. Jamieson describes this as "a difference in ownership and a move from capital to operational expenditure. "We have a clean desk and we can concentrate on what we are good at, without being distracted by technical considerations." Implementation of email and document sharing to begin with has allowed staff to become accustomed to operating in the cloud but with a familiar desktop environment, he says. After 15 months operating in this mode, Aspeq has won back some of the confidence of its staff and has near-term plans to put more in the cloud, including internally written applications, to assist the development of testing materials.