While the glitzy e-consultancy design work steals the headlines, the business that really looks set to go from strength to strength over the next few years lies deep in the heart of the average blue chip data centre. Enterprise Applications Management, or plain Applications Management (AM) is enjoying a 22% year-on-year growth rate and while some sceptics predict a modest decline by 2004 for this sector, most industry watchers believe that this game will run and run. IDC, for example, predicts that the AM market in Western Europe will be worth $5.6bn by 2004, eight times more than the volume that it anticipates for the Applications Service Provider (ASP) market.
The big consultancies are starting to take a real interest. KPMG, for example, recently opened up a brand new AM business. However, the front runners continue to be the big IT outsourcers of old. Companies like EDS, IBM, Cap Gemini and Sema all have a long track record in AM and all are busy flexing their businesses to cope as AM takes on a more dynamic form.
As companies look to get e-enabled and move to more customer-focused processes, even the most staid and static legacy applications tend to get caught up in the frenzy. AM providers find that service level agreements couched in more stable times have little bearing on new demands for integration projects that seek to web-enable these old systems.
In the bad old days of outsourcing, where supplier/customer confrontational tactics were the order of the day, this kind of departure from the script laid down in the original contract would have been manna from heaven for the outsourcer. Unscripted additional work, particularly complex, mission critical work, was often regarded as a licence to fleece the unlucky client to the bone. In this new era of caring, sharing partnerships, market discipline and the AM provider’s desire to protect its reputation for reasonable dealing, go some way towards protecting the client company. However, the real protection will increasingly come from the industry’s maturing sense that in dynamic markets, static scenarios will be a rarity. If the AM market is going to work, SLAs will increasingly have to cover a variety of outcomes. Metrics will continue to be important, but contracts will increasingly have to reflect the real, partnering nature of these activities.
Robert Brant, partner responsible for Enterprise Applications Management service at KPMG
“All the big research houses, Gartner, Forrester and IDC, for example, are predicting strong growth for AM, and we are putting some large bets down on this area. One driver is undoubtedly going to be Enterprise Requirement Planning, Phase II. We are already seeing ERP bouncing back from the doldrums.
As large companies grapple with these 17 to 20 module systems, it becomes clear to many of them that it would be easier to have an AM provider take care of the day-to-day running, upgrading and bug fixing of these systems.
We see the AM market segmenting in at least three ways. First, there is the established AM market, consisting of legacy and heritage applications.
The second segment comprises new implementations which we will look to provide a complete service on, from initial consultancy to implementation and the provision of a full hosted service. The third market consists of the new, up and coming ASP type services, which we may or may not get involved with. There are very few companies offering a true ASP package service in Europe now. It takes a very long time for a provider to recover the costs of setting up this kind of a service. Even in the US, the 6,000 or so ASPs that initially launched services are now down to just a tenth of that number.
We are the first of the Big Five to launch an AM service, aimed mainly at this stage at the ERP market. Where clients want to carry out a major integration exercise between these types of systems and legacy mainframe applications, we will do the consultancy on the middleware side and look to host the ERP suite where applicable. However, we direct the client towards mainframe AM providers such as IBM or Sema for the long-term care of the legacy systems.
Clients are definitely more amenable now to contracts that seek to reflect more of a risk/reward relationship. What we often look to do to help clients, particularly with e-ventures, is to take the initial costs of setting up the implementation, plus the ongoing costs for the duration of the contract, and average them to find a yearly charge. This smoothes the fee so that the client is not paying a large bill up front, for front end loaded implementation costs. What we do not favour particularly, is getting involved with the wilder dotcom ventures on a per transaction value recovery basis, which locks you in to their success or failure.”
Ian Denning, associate director, Cap Gemini, with responsibility for outsourcing
“This is potentially a huge market. We have long provided AM for a whole raft of enterprise systems, and many of the businesses and processes, plus the management rigour that we bring to bear in those exercises, come into play with new style AM, dealing with e-systems and ERP integration.
What we have found, however, is that the SLAs that used to govern legacy AM have had to undergo considerable changes to reflect the nature of many of today’s projects. The old style applications were very stable and the business processes that they reflected were mature and not subject to much change. In the e-business world it is a sort of mirror image inversion of this. You have applications that are very often highly unstable. Many of them are put together in a rush with speed to market being the primary driver. That is fine, but when you pick up the responsibility for the ongoing support, maintenance and delivery of these systems, it is much more like a build-and-operate deal.
Development work has to go hand in hand with support. It is usually fair to say that neither the customer nor the supplier really knows what they are getting into with these systems, so there has to be a great deal of mutual trust and maturity about the way the contract is managed from either side. All the good old traditional wisdom about thoroughly testing systems, about scalability and robustness, tends to go out the window in the initial design and it all has to be built back in. However, apart from initial design flaws caused by the dash to the web, we are also dealing with a market that is inherently more difficult to model.
Two decades ago, you knew how many accounts and users you were going to have and you could model all this in a sensible way. Now everything is much more dynamic. In our view the best way of safeguarding the interests of both AM provider and client is to move into a shared risk/reward relationship.
Generally you have to come up with a deal that allows the system to be built at a low initial cost and you look for your reward on the up side.
We have to be prepared to run the usual commercial risks with the customer as to whether the whole venture flies in the long term or not. I believe the industry will see a lot more of these kinds of partnering, shared risk deals going forward.”
Neil Barton, manager IT consulting services, at Compass Management Consulting
“I can’t see much room for the market to get confused about the respective merits and nature of AM and ASP offerings. With AM we are essentially talking about a bespoke or heavily customised application where the provider does all the programming, upgrading, bug fixing and end user wish list development and so on. This is vastly different from the typical ASP relationship, where the provider is looking to sell the same service, with absolutely no bespoke work or customised tweaks, to as many different user organisations as possible. The way an ASP gets economies into the service is precisely by avoiding customisation. This in turn leads to an argument from the business’s perspective that says, if you can get 90% of your functionality 50% cheaper, is this not the better business call? We are going to see many companies choosing this route and living with the constraints of a standard package solution delivered via the ASP route, rather than going down a heavily customised AM approach.
We see four different sectors emerging. The first will be pure functional ASP offerings, such as Microsoft Exchange on a per seat basis. United Airlines, for example, just outsourced mail to an ASP for 250,000 users precisely because this was the more efficient option. Another sector will be based on outsourcing the whole desktop to an ASP. We see very few big corporations going down this route, though it might suit a 50-person professional services firm, or a dotcom operation. Third we see certain ASPs targeting SMEs with SAP, Oracle and other ERP solutions, where the client wants to put in a big name ERP offering, but can’t justify the huge man day effort involved in implementing in-house. Finally, we see AM as the one to one, rather than one to many solution, where the AM provider enables the client company to have vast amounts of customisation to safeguard the client’s unique value added proposition.
What we do not believe companies should outsource is application portfolio management. It is very much for the business to decide what mix of applications it needs. This involves questions about how long you run with legacy applications as against migrating to modern systems, allow certain business units to keep running legacy applications when everyone else has moved on, and so on. It is all about identifying the applications that give the company real value and real returns.”
Ian Charlesworth, senior research analyst, Butler Group
“We hear a great deal these days about the burdens and inefficiencies associated with legacy systems. Companies are under a lot of pressure to migrate their core systems to ‘new age’ web-based technology. However, I would argue strongly that a ‘rip and replace’ strategy for heritage systems, that encapsulate huge amounts of development effort and which contain much of the company’s value proposition, is fundamentally flawed.
An AM solution which maintains these systems while freeing up the company to focus on strategic issues is usually going to be far more beneficial.
It is all very well for dotcom companies or for greenfield projects that start with a clean slate, to leap direct into object orientated, web and component-based technologies. By contrast, the experience the industry has to date shows that blue chip companies with heritage systems generally fail when they try to rip and replace these systems simply in order to ease enterprise management difficulties. There are many middleware and web-to-host tools available which simplify the task of integrating these legacy systems into new wave CRM or supply chain management solutions.
The idea that data gets irretrievably locked up in these legacy systems is now just plain wrong. The middleware Applications Enterprise Integration market is now very strong indeed, and it is already a central feature of AM strategies.
Fuelling the drive to rip and replace legacy systems is the sheer marketing pressure from ERP vendors on the importance of tight integration. However, ERP projects themselves have huge integration issues – and their costs are legendary. There is an obvious value proposition in having all your applications able to speak together, but migration to a homogenous applications platform is not the only route to achieve this end. Nor is it really possible when you look to join the enterprise up with suppliers and customers, all of whom have different systems.
There is no simple solution. It does not matter whether a company has AM’d its core systems or not, it will still have to swallow an expensive and difficult project to build enterprise wide and inter-enterprise applications integration capabilities.”
Andy Thompson, director, management consultancy services, PricewaterhouseCoopers
“We would agree that the AM and Applications Enterprise Integration (AEI) markets have merged. AEI middleware tool vendors are doing extremely well, since they help to eliminate the need for massive custom programming exercises as companies try to achieve integration both internally and with suppliers and customers.
Our belief, though, is that post Y2K, there are now far fewer legacy applications around than might be supposed. The big ERP vendors were conspicuously successful at using Y2K pressures to get companies to migrate to ERP solutions.
The major thrust today is to move from these essentially back office solutions to more customer facing, front office, value added systems. This creates a huge consultancy opportunity and a huge AM opportunity. We can take these systems on for people, maintain them and help them to complete the job of e-enabling and CRM-enabling their operations.
We see a move by businesses to accept global best practice for a sector, once it is properly encapsulated in standard package software. Business clients are increasingly trying to move away from bespoke systems since it is becoming harder and harder to prove a real value proposition in heavy customisation.
There is obviously an issue with a company approaching an AM provider for unexpected modifications, perhaps with an aim to gaining efficiencies from linking the AM application to new technologies. However, that company would have other, perhaps equally pressing, issues internally, if it was running everything in-house. Some AM providers will doubtless look to charge a premium under those circumstances. But there are plenty of flexible options for clients. They can hire third parties to come in and build a warehouse or data mart type solution to facilitate the link with the new system. In most contractual circumstances they will not be wholly bound to use the AM provider’s resources. This is yet another instance of how AM and AEI knit together, with perhaps different resources being used to do each job. Everything comes down to what the company wants to do, where its strategic vision is pointing, and how much resource it can afford. Contracts for AM deals have to give security to both parties, but also take into account that the future is increasingly unpredictable. Clients must avoid lock-in traps, but there is now much contract-writing experience in the market.”
Hugh Morris, partner, Accenture
“If one is looking for a distinction between AM and AEI, the time line involved is usually a clear guide. AM is for the long haul, while AEI is usually a get-in and get-out scenario. In today’s large AM contracts there are normally any number of third party contractors, in addition to the AM provider. What the client wants is a single organisation like ourselves to act as prime contractor through the contract.
In taking this role, the AM provider frees up a substantial amount of internal management resource in the client company. Clients can focus on what they want, not on the mechanics of how to get it. Of course, in our position as prime contractor on AM deals we are frequently working alongside our major competitors, such as EDS or IBM. We all realise that if we fight out our competitive battles within the framework of a client AM relationship everyone will lose massively.
It is clearly true that where unexpected elements come into play, an AM contract has to flex to cope with this. However, no contract will permit unreality to prevail. Even if clients retained a system in-house there is still no way that they could get jam for nothing, as it were. In an AM world, everything has a clear price tag attached to it. But this is no bad thing from a corporate perspective. It puts a sharp focus on the costs of IT and forces companies to make strategic decisions.”
Anthony Harrington is a freelance journalist