Neuma White Paper:
CM: THE NEXT GENERATION of ALM is on Our Doorstep
2008 brought us the buyout of Telelogic by IBM, not long after IBM had already acquired and integrated Rational. However, 2008 also brought us a global recession which has governments working hard to avoid having the R-word turn into the D-word. These events will help shape the coming year of CM and ALM.
To start with, companies are downsizing. Large, expensive CM/ALM systems are adding to the pressure not only to keep the bottom line in the black, but to keep companies afloat. The total cost of operation for CM/ALM systems, which can be quite significant, will not be justified by the technical need when there are other, less costly, and in some cases more capable, solutions out there, and when there are more pressing financial needs. So the shift from big-IT will start in earnest this year. CM/ALM will only be one segment of big-IT affected. Watch for SAP and Oracle to be hit hard. Off-site training and conferences will be hit hard, with a move to put more on-line.
2009: An Opportunity for CM/ALM Industry
Nevertheless, the CM/ALM industry will prosper, as companies look at how they can reduce costs. The proof of reducing operational costs will have to be there, up front, before anyone commits to new capital purchases however. For that reason, much of the CM/ALM industry action will involve companies migrating from existing costly solutions, including big-IT CM/ALM solutions, to more lean solutions. This is the real beginning of the 3rd Generation of CM.
How will this happen? Organizations will begin to look at their costs, and then their IT costs, and then their CM/ALM costs. Typically these involve:
- New license acquisitions (not much of that this year with falling employment)
- Annual maintenance/support fees
- New Release Upgrade Fees
- CM/ALM System Administration Costs
- CM/ALM Process/Tool Customization Costs
- CM/ALM Training Costs
- Glue development and maintenance costs for tool integration
- Cost of supporting infrastructure, such as databases and hardware, including cost of operation.
- Cost of down-time/outages, recovery, backups (especially as teams become lean)
- The capability for productivity improvements for developers, managers, CM Managers, Project and Product Managers, etc.
- The ability to make good, accurate decisions based on up-to-date information, to avoid rework, vaporware, etc.
For some vendors, this will be a real challenge. IBM is a service company. Will they be able to adjust to the changing CM/ALM market? For other vendors, this is a real opportunity to entice prospects using competitive solutions to test drive their own. These vendors will have to have something to offer that can deal with the requirements of these prospects at a lower cost of operation, and even after factoring in the cost of new licenses.
On the other side of the economic fence is a significant wild card. As governments attempt to stimulate the economy, how much will they loosen their purse strings? 2009 could see the biggest push from defense, aerospace and other government and government-contracted organizations to improve their operational processes, as the coffers swell from stimulus efforts. It's a great opportunity for government to build and/or upgrade its IT infrastructure, and CM/ALM is a significant part of that infrastructure, especially where process improvement is needed.
And the wild card extends to industry, at least to those organizations that have to come to grips with being more competitive. As stimulus programs are unveiled, there will be requirements for significant structural improvements attached. So the automotive and aerospace industries could easily be looking to realize process improvement. They may look to proven process improvement methods such as Six Sigma and to a lesser extent CM II, as well as to more generic process improvement methodologies/monitors, such as CMMI. An interesting note about CM II is that this fall, the ICM (Institute of Configuration Management) officially launched a rebranding effort to cover not only traditional CM, but also the wider set of enterprise processes.
The Proof is in the Pudding
Now with all of this opportunity to improve process, organizations are going to want some assurance that as they reduce costs, they're not jumping from the frying pan into the fire with respect to both their infrastructure and their operating processes. This is where CM consultants and vendors will have to prove themselves. They'll have to deal with issues such as:
- What guarantees will you give that your solution will work for us?
- Can you demonstrate that it will really take only a few weeks to change our CM/ALM infrastructure?
- How do I know that I won't be caught in requiring expensive and extensive consulting resources?
- What are the productivity improvements that I'll see that I haven't seen before?
- What will training cost and can it be short, for productivity purposes, and flexible, for incrementally bringing users up to speed?
- Your solution won't work if you can't show me better levels of availability - can you?
- Can we continue to evolve our process to improve more and more over time with your solution?
- How quickly can you customize the tools and process support to meet our needs?
- How easy is it to evaluate your solution with our existing data, and at what cost and risk?
- How is your solution going to affect our product and process quality?
- How will our users react to the proposed changes? What about our managers and executives?
These aren't new issues brought on by an economic crisis. They're just amplified by it. Somebody's trying to cut costs, be more competitive and improve quality. Instead of doing it incrementally, they're looking for a giant leap forward. That's the difference and that where the consultants and vendors will have to come through in spades.
The Next Generation - Vendors Step Up To The Plate
So this is where we really start to move into the 3rd Generation (3G) of CM/ALM tools. I've talked a lot about what that means in my history of articles, especially in November's article on evaluating and selecting a CM tool. So let's review some of the 3G advances that will be most applicable to the current year, the current crisis and the above issues. Focus on these issues will be apparent through the coming year. Over the next 3 to 5 years, only the vendors that step up to the plate will remain key players.
What guarantees will you give that your solution will work for us?
Vendors will have to put their money where their mouth is. Money is tight. Nobody wants to pay until they've proven the solution will work. What can vendors do? Well, they will have to come up with some creative alternatives for the current times. Here are some examples:
- Pay only annual maintenance/support costs for the first year - you don't have to make a buy decision for 12 months.
- Put 25% down, 25% in 6 months and the balance in 12 months. If it doesn't meet your needs, you can bail out of the contract at any time.
- You may lease the solution on a monthly basis - on a lease to acquire the licenses over 5 years. You can bail out at any time.
Demonstration of Timely Cutover
Can you demonstrate that it will really take only a few weeks to change our CM/ALM infrastructure?
How easy is it to evaluate your solution with our existing data, and at what cost and risk?
Simple demos won't cut it anymore. Project teams want to evaluate the tool with all of their data. Ideally, they'd like to run it in parallel with the existing system for a month or two before committing to buy. If cutover is really promised to be easy, they'll want an attractive cost-sharing, fixed-price, cut-over contract before buying.
Another approach might be to run a fixed price pilot. This puts everyone through the paces - not only cutover, but training, productivity, ease-of-use, etc. If it's a small project, this approach is really not a pilot. Instead, you're asking for a fixed price initial contract with options to extend the operation and acquire licenses at a pre-determined price.
Gone are the days when you buy the tool, or use up your budget, before knowing it's the right choice. This is a critical demonstration. If cutover is not easy, it's likely the total cost of operation won't be low. They'll want to throw in some customization tasks in the trial. If the solution is not replacing all of the tools, vendors will have to demonstrate integration with existing tools as well, as that will definitely be part of the cutover process. Ideally though, the 3G tool set will be an end-to-end set of tools so that there will be no need to do this integration and maintain the glue.
How do I know that I won't be caught in requiring expensive and extensive consulting resources?
How quickly can you customize the tools and process support to meet our needs?
Expect to see 3 requirements up front from a consulting perspective. Customers will say:
"First, I want to operate the tool and see that it requires almost no administration. We'll force a couple of outages to be sure, and throw in a multiple site operation as well.
Next, I'll give you some time-constrained customizations and you do them on site.
Finally, you'll train us and we'll do a second set of customizations on site.
If you pass the time and dollar constraints, I'll believe you. Otherwise, it's too big a risk for us."
Risk will become unacceptable. And customers will start to look at customization in a new light. It's not only a cost for benefit thing. It's really, "How responsive can we be to our needs? What confidence do we have that the existing technology and budget will get us through to the end of the project?" If a customer cannot control these factors, the risk is substantial.
What will training cost and can it be short, for productivity purposes, and flexible, for incrementally bringing users up to speed?
The first rule of a good, efficient, comprehensive training program is... ease of use. That doesn't sound much like training, but if the product requires too much training, forget it. Organizations can't afford the per diem course charge. But equally, they can't afford the lost productivity which comes out to about $2K/person/week. Besides that, try keeping someone's attention for a few days learning a new tool they didn't really want in the first place. End-user training will now be measured in hours, not days.
The second rule: course materials will need to be customized to organization terminology and process. It's fine if the vendor wants to call a change package a "changeset" or an "update". It's fine to call an "issue" a "problem report", a "bug", or a "defect". But don't try to train users on the flow of a problem report by using different state names than they are already familiar with. Your course will turn into a terminology session.
And finally, course material will need to be available on-demand. This means video and/or computer-based training - still customized mind you. In some cases, the vendor may help the customer customize the course and present it themselves, for a much lower royalty fee instead of a course fee. In other cases, vendors may provide web-based live customized training, which can later be viewed by others, at a nominal charge per view.
During the 3rd generation and into the 4th, tools will evolve so that the training is fully integrated with the tool, and in such a way that the organization can control the content. For those vendors looking to training revenues as a significant source of revenue, don't rely on your basic user operation courses any more. Sure CM is still a complex business, but the ease-of-use has got to be there regardless. That means smarter technology, and in-tool process-guidance.
Reliability, Availability and Performance (RAP)
Your solution won't work if you can't show me better levels of availability - can you?
Good question, especially if your tools rely on infrastructure in the layer below the CM functionality. As the project scales, will the vendor solution continue to perform well and reliably. Can consistent backups be done without affecting availability, even on very large projects? If there's a need to switch CM/ALM tools once again because of performance issues, the costs have just doubled.
What happens to multiple site operations when networks go down? Is data still available? And what happens when the network is restored? What happens when a disk crashes, or a disaster takes out one of your data centers? Non-availability of data is not an option in some projects, even if users have to cope with a no-update capability for a couple of hours. What about during an upgrade, or an upgrade of the infrastructure layer? Is data still fully available? What if there's a need to switch from a Linux architecture to the next OS or to Windows, or to different hardware? Or what if the vendor software has a problem? What if there's data sabotage from within the project? The easier it is to cope with these changes, the less it will cost down the road.
A related issue is security. Is data availability easily restricted? Can multiple contractors work on the same project without seeing one another's work, while the integration team sees all of it?
Better Processes Mean Higher Quality
Can we continue to evolve our process to improve more and more over time with your solution?
How is your solution going to affect our product and process quality?
Software CM varies considerably among developers. This is not just because different tools are being used. It's because different market and contract demands require differing processes: Agile, Regulated, Mission Critical, Safety Critical, One-time deploy, Aircraft/Vessel Deployed, COTS, etc. These practices all have differing process requirements. And the process changes from the beginning of the release cycle to the end of it. Continuous improvement is more and more critical both to remain financially stable and to remain competitive. When organizations switch infrastructure, they'll be looking for CM/ALM tools that support the processes they need.
3G Features and Ease-of-Use
What are the productivity improvements that I'll see that I haven't seen before?
How will our users react to the proposed changes? What about our managers and executives?
Software CM is complex. This complexity, along with the traditionally technical user base, have resulted in complex tools and processes, with long learning curves. Tools which are easier to use will have a definite competitive advantage in any churn of the CM market. And tools that can let managers and even executives have easy access to project information so that timely decisions can be made, will have a further advantage. The rest of the software industry has focused on ease-of-use to a much greater extent than the CM niche. People are aware of the differences between easy-to-use and hard-to-use software, and so we'll start to see a clear shift in user interface functions as new vendors try to replace existing deployments.
Some of the things organizations will be looking for include:
- End-to-end integrated solution with point-and-click navigation along traceability links
- Interactive build and release comparisons
- Management and Role-based Dashboards with drill-down capability
- Tool driven CCB meetings and peer reviews
- Clear To-Do Lists for all roles/users
Whatever makes life easier will be an important consideration. Each user role will have their own impressions, so vendors will have to evaluate their offerings to ensure that they address each role in the ALM function.
2009 - A Turning Point
If IT infrastructure budgets are significant, it may be a significant turning point for CM/ALM tool and process infrastructure. This is an opportunity to which both vendors and consultants must be ready to respond. A crisis is a point of change, so expect to see a lot of it this year.
However, I've been around long enough to know that I shouldn't expect a lot of motion in the industry in a single year, but the current times do set the table for significant change, driven by competitive and financial requirements of the corporate world. Still, I'd be happy to see the shift to 3rd Generation CM platforms take place over the next 2 or 3 years. If 2009 serves only to get CM Managers to look around at their options, that will be a major advance. If they don't, they may find their trimmed budgets are unable to cope with their existing tools and processes.
Joe Farah is the President and CEO of Neuma Technology . Prior to co-founding Neuma in 1990 and directing the development of CM+, Joe was Director of Software Architecture and Technology at Mitel, and in the 1970s a Development Manager at Nortel (Bell-Northern Research) where he developed the Program Library System (PLS) still heavily in use by Nortel's largest projects. A software developer since the late 1960s, Joe holds a B.A.Sc. degree in Engineering Science from the University of Toronto. You can contact Joe by email at firstname.lastname@example.org