The brutal reality is that IT organizations are still on the hook to provide mission critical data to the business, whether the SaaS provider has failed to deliver, or whatever the reason. The challenge is identifying what is mission critical data, and then, what level of risk it is exposed to, and finally, how quickly will the business need access to it when there’s an issue.

What is mission critical data? Practically speaking, some critical attributes are: system of record and operationally critical. Data can have either attribute or in some cases, both. Examples:

1. System of Record: Last month’s sales receipts.
2. Operationally Critical: Urgent customer support tickets.

Although last month’s sales receipts are very important data, critical for financial reporting and needs to be preserved for years, if the system that stores this information goes offline, chances are that business can survive for a 48 hour period without access to this information. In contrast, an urgent support ticket is information that is needed immediately. There may be Service Level Agreement at stake. What do you do if your SaaS provider tells you that access to this information won’t be possible for a couple of days. What do you tell the business?

We have made the decision to back up both types of our data nightly to our own corporate databases. Great SaaS vendors make it easy to access this data… it’s OUR data. But if the worst happens, we will be able to create ad hoc reports to satisfy immediate needs.

If you have engaged a quality SaaS provider it’s unlikely they would be off the air for more than 12-24 hours. And failures like that shouldn’t occur often. Make sure you ask about Recovery Time Objectives for disaster recovery when you are shopping for SaaS services. Your prospective SaaS provider should be prepared with an answer!

There is an interesting debate taking place amongst the commentators of IT trends and enterprise applications, particularly ERP and supply chain management applications, about the suitability and adoption of cloud or software-as-a-service (SaaS) solution to this space.  This is exemplified by the article from Todd Morrison at SearchSap.com that appeared on Dec 01, ’10 in which contrasts opinions by Liz Herbert of Forrester and Ray Wang, principal analyst and CEO of Constellation Research Group.  This debate is nothing new in that it has been raging since the advent of SaaS some years ago.

On the one side Forrester repeats the usual arguments against broad adoption of SaaS for enterprise applications, which, according the article with my emphasis, state that:

Organizations can also benefit from the Software as a Service (Saas) products that complement the core SAP applications they’re running, according to Forrester.
…  In addition, some organizations are also reluctant to put mission-critical operations, such as supply chain management workloads, into the cloud.

Ray Wang, on the other hand is quoted as stating that…

what’s holding back the adoption of SAP’s on-demand/SaaS offerings is a “lack of ease of use” and cutting-edge functionality such as offered by Kinaxis and Plex.

Liz Herbert’s opinion is these applications…

don’t offer the breadth of capabilities that SAP does and that companies are reluctant to deploy software from Plex and Kinaxis because they don’t have the established reputation that SAP does, and certain functions like SCM don’t lend themselves to on-demand applications.

There is no doubt that SAP has a much greater market share than Kinaxis, but I challenge anyone to call on our customers to evaluate our ‘established reputation’.  And these customers include the likes of RIM, Flextronics, Lockheed Martin, and Amgen, all of whom use an SAP ERP backbone.

And, by the way, they are all very large multi-billion dollar companies with operations around the world.  I state this because the other usual pushback against the adoption of enterprise on-demand solution is that they are fine for the ‘little guys’ but don’t play in the ‘big boys’.

Let me also state that while many of our larger customers initially chose an on-premise deployment, many have either converted or are in discussion with us about converting to on-demand.  More significantly, the majority of new customers are choosing to go on-demand.  This trend is also being observed by SAP as announced at their SAP Influencer Summit yesterday.  While I have to agree that Business ByDesign is aimed at small to medium enterprises (so the ‘little guys’, not the big boys’) SAP stated that on-demand inquiries have gone from 1 in 10 last year to 1 in 2 this year.

Let me also concede that for many ERP functions the only incentive for putting things in the ‘cloud’ is one of cost effectiveness because many of the core ERP functions do not stretch outside the organizational boundaries.  And yet two of the functional areas that are seeing huge increases in adoption of on-demand solutions are customer relation management (CRM), or sales force automation (SFA), and human capital management (HCM) or human resources management (HRM).  I don’t know what could be more confidential than a company’s sales order pipeline or their employee records, including employee evaluations.  OK, maybe I will concede that the core financials are more confidential.

But I do find it interesting that Herbert considers the core ERP capabilities to be most applicable inside the four walls of a company. What surprises me is that she considers supply chain management to be restricted to inside the four walls of a company.  Our observation is that that for most of our customers outsourcing has meant that increasingly little of the supply chain is owned by the companies, especially the brand owners.  If these companies had the myopic view of the supply chain being only that part of the process in which they own the materials or product they wouldn’t need any of the SAP supply chain management suite because they outsource most of their manufacturing, in many cases never touching the product.

It is our observation that leading brand owners in high-tech electronics, consumer electronics, aerospace and defense, and life sciences need greater visibility and control of the end-to-end supply chain in order to provide reliable promise dates and to be able to deliver according to the promises made.  As inventories have been squeezed out of the retail and supply chains locations, the need for greater responsiveness and agility has increased enormously.  Ask any electronics manufacturer about the availability of key components over the past 12-18 months and you will hear stories about commodity items that could be purchased with a 2 week lead time as little as 2 years ago that now has 13 week lead times if you can find a supplier.  Given the buy-sell relationship many electronics OEM’s have with component suppliers (in which they buy items from the component suppliers and sell these items to the contract manufacturers) it is clear that they need a lot more visibility into and control over the supply chain than is provided in ERP systems, which is the core reason for their increased interest in on-demand SCM solutions.

What about your company?  What sort of ‘cloud’ strategy are they adopting?  What sort of applications will they be most interested in deploying on-demand?

The way projects are run have changed significantly since I started consulting 19 years ago.  In those days, it was unheard of not to be onsite at the client Monday through Friday working on a project.  Even though during that time you may only spend a few hours directly with the client.  Of course with the advances in technology (laptops issued to everyone, VPN, mobile/smart phones, VOIP, etc.) the consulting world has evolved.  While we once had to be onsite 100%, that then changed to 75-80% (7-10 years ago).  Clients sometimes let the consultants work from home one day a week.  The clients still had the mindset that they wanted to see the people who were doing work for them to ensure that the work was being done and they got what they paid for.

Fast forward a few more years to implementing Software as a Service (Saas).  The world changed again.  Most Saas vendors have an implementation approach that has much less onsite activities than a traditional software implementation.  Consultants typically will be onsite for specific activities or milestones while the remainder of the work will be performed remotely.  With access to servers in the cloud, web conferencing products, and more standard implementations due to less customizations, the need to be onsite is less critical. 

I certainly do believe there is less need for onsite work for a Saas project; however, there are critical activities which I do believe should be performed onsite:

1. Project Kick-off – the entire team should be onsite to set and hear expectations for the project.  Also to build rapport amongst the team and document how the team will work together during the project.
2. Requirements and Design – all requirement review and design sessions should be performed onsite so the teams can ensure both sides have an accurate understanding of the requirement and how it will be satisfied.
3. Testing – during end-to-end and acceptance testing the teams should be together to provide knowledge transfer of the solution, answer questions and troubleshoot and resolve issues quickly.
4. Production – during production cut over it is important for the teams to be together to resolve any production issues quickly and answer questions.

These may not be the only times when the teams should be together, but I believe these are the most critical.  If the client is struggling with understanding the solution and taking ownership of the solution, then more onsite work may be required.  Consultants need to watch for this and be flexible to change the schedule to best meet customer needs.

In order to make this model work, it is critical to set the appropriate expectations in the sales cycle.  Most clients will agree with the approach once they understand how it will work.  Clients are also increasingly budget challenged and the cost savings of not having all consultants onsite every week will be attractive.  Although, cost concerns should not outweigh whatever is required to make the project successful.

What are your thoughts and experience with performing remote work?

I agree with the author that the future of supply chain software is on-demand.  But the real question is why? 

I personally think it stems from the fact that companies no longer want to risk huge, costly, time intensive implementations of software with no predictable end result.  On-demand software can offer the customer a whole new future. 

Low up front software costs (no huge multi-million dollar investment), no hardware cost and generally speaking the subscription pricing is user-based so customers have the ability to try the software with a few users, get some user adoption before committing to a huge roll-out across the organization.  This helps to de-risk the investment.  And honestly, customer executives would no longer have to potentially put their job on the line for a supply chain solution. 

Also, I believe on-demand software companies have to provide better customer service than traditional on-premises licensed software because the software vendor is constantly trying to earn and retain a customer’s business.  There is no perpetual forever license. The software company must listen to customer needs and adapt to them or face the customer going off subscription.  As everyone knows, the supply chain industry is dynamic so the way to solve a problem today can be different tomorrow.  Wouldn’t you want to work with a vendor that would actually listen to your customer requirements and put them into the product rather than to have a bunch of custom code the vendor won’t support?  I think you get more of that with an on-demand software vendor.

The author also made a very interesting comment “Increasingly, you will be able to try the software before you buy it! What a dramatic, game changing impact that will have.”  Just think about that for a minute.  What a huge statement.  If all those companies who had failed supply chain implementations could have just tried it before they made a huge investment would it have made a difference?  For some, yes and others maybe not, but the key point is that consumers of supply chain software can be more educated prior to embarking on a software implementation. 

What would be even more provocative would be a company who is willing to not only let you try the software for no cost, but also perform the implementation with no initial services fees and no cost at all if you don’t like it…  That would really change the game, the customer could completely de-risk their supply chain software decision.  Now that would be bold.

I have been in the enterprise software market, principally with supply chain management vendors, for 20 years so have lived through the rapid adoption of this technology. There is no doubt in my mind that the biggest disruption caused by SaaS is transitioning to a subscription model. This is so tied up with cash flow and maintenance, not to mention Wall Street expectations and methods of evaluating performance, that the transition to subscription pricing has to be handled extremely carefully. We have only to witness the backlash from SAP and Oracle customers on their recent attempt to increase maintenance fees to over 25%, not to mention the consequences to the SAP management board. These companies are hooked on maintenance fees, not only the initial license fees.  According to their 10-K, for the 3 months ending Feb 28, 2010 Oracle reported $1,718M in “New software licenses” and $3,297M in “Software license updates and product support”.  In other words, Oracle’s maintenance fees were nearly twice as much as their revenue from new licenses. How do you transition from that ratio?  In their annual report for 2009, SAP reported €2,607M in “Software revenue” and €5,285M in “Support revenue”, which is very close to twice as much as their new license revenue.

The technology issues are related to cost and therefore more in the control of the software vendor. One can still deliver on-premise solutions, but use subscription pricing. Without a doubt, over the long term a transition must be made to web technologies so that the solution can be hosted.  While client/server technology facilitates hosting to some extent, as Ron points out there is a lot of time and cost associated with installing a client on everyone desk who needs to use the application.

For solutions that are computationally heavy, I am not yet convinced that virtualization is not a suitable substitute for multi-tenancy. The operating costs will be higher because of the need to deploy software updates on each of the instances, but this can be automated. Multi-tenancy has gained a strong foothold in computationally light applications such as CRM. I am not sure what would happen if an MRP run was thrown at a multi-tenant engine. Can the clock cycles be managed as effectively in a multitenant model as they are in a virtualization model?

We are seeing quite a bit of interest in SaaS from large multi-billion dollar companies. Some legacy customers have moved from on-premise to on-demand, and there is increasing interest in on-demand from prospects. As your graph shows, large enterprises are slower to adopt SaaS.

We are becoming more and more familiar with the key benefits of SaaS:

• Richer user experience and user productivity
• Rapid IT implementation and optional higher quality deployment
• More frequent cycles of innovation
• Minimal upgrade hassles
• Access to the “always-on” SaaS tools regardless of location
• Subscription pricing
• Anytime scalability and dynamic capacity
  Source: Altimeter Group October 2009 Customer Bill of Rights: Software-as-a-Service.

To deliver these benefits, should a SaaS provider use a multi-tenant architecture? What are the pros and cons of this architecture?

Multi-tenancy, if done correctly, can drastically reduce the costs for the SaaS provider. If an application can service multiple customers on ONE piece of hardware, with ONE OS license and ONE database license it is far less expensive to operate than a multi-instance architecture. In addition, the cost of resiliency investments like server clustering can be kept much lower. All in all, the provider can keep his costs lower… and pass these savings on to his customers. (One hopes!)

The ultimate expression of this model is Google. They have driven the cost to zero for email and personal productivity applications. They leverage a massively multi-tenant architecture that is free to users. Talk about a value proposition. We can all see what’s in it for the customers… the very best price going!

But as we all know, Google is providing a ‘commodity as a service’ which implies a very different business model from other software/service vendors. What about supply chain applications deployed to meet a particular company’s needs? What value does a multi-tenant architecture bring to them?

Well if supply chain services were free, the case would be simple. However, there may be factors at play that are more important than cost: security, availability, performance and scalability to name a few. How does multi-tenancy improve value to the customer in these areas?

SECURITY: There is always the possibility of a ‘leak’ in the security model for a multi-tenant model that could be exploited by a creative individual. Data is living in physical proximity, in the same database. Multi-instance architecture by contrast is more secure by its insular nature.

AVAILABILITY: A reputable SaaS provider takes great care to meet the Service Level Agreement in the contract with its customers. It’s an axiomatic part of the operations of a SaaS company. Customers can depend on it. How does multi-tenancy help deliver value to customers here? Other than the risk of a massive hit to the reputation when hundreds or thousands of customers are without service, there’s nothing inherent in a multi-tenant architecture that delivers better availability. High availability is all about great design, great equipment and great process. (As I write this I just got notified that our Salesforce.com service is down- ouch.)

PERFORMANCE: All multi-tenant architectures are built to govern and limit read/write access to prevent resource over-utilization. This type of design can restrict flexibility whenever user extensions to standard applications are built. This is a clear limitation for customers in the multi-tenant model. Also, there are application areas that don’t lend themselves to this type of ‘governed’ model. Multi-instance might be the only alternative that can deliver satisfactory performance without the risk of compromising another customer’s performance.

SCALABILITY:  Does multi-tenancy deliver scalability? It’s obvious that an application built for hundreds of customers using it simultaneously must be scalable. But, in which dimensions is it built to scale? Users, data size, transaction volume? Again, by nature the multi-tenant model does not deliver this scalability- the application whether multi-tenant or multi-instance must be designed to scale in these dimensions… with the cpu, network cycles, and memory space to back it all up.

UPSHOT: It is clear that well designed multi-tenancy is essential for delivering software service on a large scale where profit margins are thin. However, by its inherent nature, multi-tenancy does not have a customer value delivery advantage in many respects. On the contrary, for delivering a single customer software service, like a supply chain solution or even a social media solution, a great single instance design and operating environment might just be superior, with better security and dedicated resources allowing consistent performance, inherent in the model.

WHAT DO YOU THINK?? When is multi-tenancy an ‘anti-feature’? Are you willing to pay more for features or benefits that are only available in the multi-instance model?

I believe what I call SaaS 2.0 will change the enterprise software model forever. Software vendors will take the entire risk of deploying major SaaS based software offerings – before the customer has to commit any monies. The SaaS 2.0 vendors will have to prove the value of their offerings in a production environment – not just some trivial demonstration.

When you think about it – enterprise software vendors have been getting away with the business equivalent of murder. If you think this is an exaggeration – how about getting away with fraud?  There have been billions of dollars and millions of man years wasted on failed enterprise software projects. Consumers do not buy any kind of other products (homes, cars, buildings, airplanes, etc.) under such a crazy business model.

CIO’s are justifiably sick and tired of an enterprise software model which is ludicrous. CIO’s also hate enterprise software offerings which are only useful to a very small user community. SaaS vendors have already addressed this issue.

How could a vendor offer such a riskless value proposition? The SaaS vendor must have a very strong balance sheet (to fund initial deployments), shorter more efficient deployment cycles, and more importantly have an offering which they know works!

Look out for SaaS 2.0 – it is coming!

Deployments are still required for SaaS, but on average they are much less involved and costly.

Security is even a greater benefit of SaaS. An on-demand vendor has the greatest incentive to keep their service secure – it is called losing customers if they don’t. One of my favorite stories is about a large enterprise (years ago) who refused to go Saas because of “data security” concerns. Ironically, thirty days after this discussion they had a major breach of their own IT operations which became public knowledge.

I believe one of the greatest reasons to seek out SaaS offerings centers around system administration. Turnover in IT operations is common these days, so why wouldn’t you want the vendor who knows the offering to be responsible for system administration? As CEO of a SaaS company I am often asked why we don’t offer a lower cost on-premise version of our offering. I tell them I would have to charge more because we get more customer support calls due to increased system administration calls.

Having pointed out all of these indirect benefits – economics is still the largest reason to go the SaaS way. There were billions of IT monies wasted on what became “shelf-ware” – because vendors stuffed contracts with additional users or capabilities which were never used. What I have found with SaaS is customers only order the number of users they need in the short term, because it is easy to add users later on.

While there are still some “last remaining holdouts” on the movement to SaaS – it is now undeniable. A proof point is witnessing the “most secretive and data security oriented” large enterprises asking for SaaS!

With manufacturing enterprises, a variety of unforeseen events can occur on a daily basis related to customer demand and the supply of the manufactured goods, as well as the supply of required purchased components from vendors. Even with the best of plans going into a reporting period, actions need to be taken to adjust for these unplanned changes in demand and supply in order to meet the revenue target.

I read the recent report titled “Performance Management in the Midmarket“, by David Hatch and Max Gladstone, from the Aberdeen Group, dated November 2009. The thing that struck me is that based on data collected for the report, the top pressure related to performance management for both midmarket and large enterprises is “lag times and the inaccuracy of operational business decisions”. What is stated is that operational performance is negatively impacted when “the lag time for determining operation performance is greater than the decision window (the opportunity to affect performance based on taking or changing an action)”.

The report goes on to identify the major stumbling block is the identification of source data: “During a fiscal period review process, or within an operational performance review meeting, senior executives often ask for metrics and analyses that send their business managers and teams scurrying to find the data that supports the desired answer”. From my point of view, the problem is that data is not readily available to the decision makers. In addition, reports and analytics need to be available against the data for review and to enable effective operational responses. The data and reporting cannot just be looking at the past, but the analytics need to be forward looking to provide information to decision makers so that actions can be taken to meet the performance metrics.

The operational requirement to meet revenue goals for reporting periods is a common concern among midmarket enterprises and could also be viewed as a fairly standard process across entities. Also, given that the absence of source data and the required reporting and analytics is also common, there exists a real opportunity for a standard solution. Midmarket companies can increase their effectiveness at meeting revenue goals by implementing an out of the box solution. The best solution will have the following key attributes:

• Standard integrations with various MRP/ERP systems
• “Cloud computing” enabled – midmarket companies can subscribe to the software rather than having to stand up and manage the infrastructure and software
• Simple to use with little or no training required
• Easy to tailor and configure if needed

Of course, many other operational issues could be targeted other than revenue management and with today’s technology, many solutions that only large enterprises have the resources to invest in can be made available to the midmarket. What do you think?

So many problems were stirring, like a runaway train;
The 2010 initiatives were set in the boardroom with care,
In hopes that inventory reductions and delivery improvements soon would be there.

The executives were nested all snug in their beds,
While visions of corporate success danced in their heads;
And IT with their servers, and you with your supply chain cap,
Had just settled down for a 2009 nap.

When out in the market there arose such a clatter,
You awoke from your nap to see what was the matter;
Away to the computer you made with a dash,
Logged in to the system and saw a big flash.
The light on the screen with the words that appeared,
Gave the lustre of hope to supply chains next year;
When, what to your wondering eyes should you see,
But many new initiatives for businesses.. to your glee!

Software as a service is easy and quick,
Just log on to the web and that does the trick;
Rapid implementations are the name of the game,
Just whistle and shout and get rid of your mainframe.

Now, SIMPLE, now FAST now COST EFFECTIVE and PRACTICAL!
On SaaS, on CLOUD, on DISRUPTIVE and INNOVATIVE!
To the top of the supply chain! To the top of your stock!
Now dash away dash away costs - go plummet like a rock!

But as forecasts increase or your customers don’t buy,
You miss your targets and inventory mounts to the sky;
You’re left with poor systems and the consultants just flew,
With all of their knowledge and your confidence too.

And then, in a blog, or a website you heard,
The news of new technology …but it sounded absurd!
As you sat with bewilderment and was turning around,
Down the IT stack it came and oh, with a bound.

It was dressed in one solution from the Demand to the Supply,
And its interface was kind of Excel like…it really caught your eye;
But it also looked flexible and user friendly for many on your team,
You smiled and laughed and started to dream.

With this service you could change demand and watch inventory replan on the fly,
With this service you could say to the word ‘latency’ GOOD BYE;
When your business changes your processes can too,
And adoption will spread as quick as the flu.

Now wouldn’t it be nice to start 2010 without any troubles?
To have demand and supply so smooth without any bubbles?
Well we know that won’t happen; volatility is here to stay,
The question is what to do since it won’t go away?

Do your homework, ask questions, but don’t look too far and beyond,
There are technologies out there to help with the way that you respond;
Your competition may be doing it right so make sure and listen,
Many companies are using services that make their eyes glisten.

All the best in 2010,
Good luck with your plan;
But remember,
You need sensible sense and responsive response as fast as you can!

Start out your new year by attending SCM World Live – The Free-to-Attend Virtual Supply Chain Conference . Link here: http://www.scmworldlive.com