Engagement Models in IT Services

As I interact with my young colleagues at work I get constantly questioned on non-linear models and the challenges being faced by the industry. So I thought of writing 3 separate articles. This is Part 1 of the series on "Engagement Models in IT Services" that sets the context for Part 2 focusing on "Why Traditional IT Services Engagement Models are Failing" and Part 3 "Business Function Outsourcing" discusses the roadblocks that organisations face while implementing non-linear models.

Trust you will find this informative and if you do, please forward this to your colleagues so that they can take advantage of this shared knowledge.

 a) T&M (Time & Material) - Perhaps one of the most widely used models for most consulting, development & testing assignments. Here revenue is directly proportional to number of hours being put by the employees (the more the merrier !) 

b) Fixed Price - Essentially not very different from T&M as it is as linear as any other model except the fact that a vendor has complete flexibility in staffing & managing the project given that most risks are assumed by the vendor. This model is mostly employed wherever business problem can be clearly and unambiguously defined. This is also perceived as a better model than T&M because of the control that vendors have over execution.

c) Fixed Monthly is a hybrid between T&M and Fixed Price where a certain "productivity" is promised every month for a fixed rate. It has advantages over T&M as it ensures constant source of revenue irrespective of holidays & vacations and yet unlike Fixed Price, one is not bound by hard deliverable.

In all the three models above there is a direct linear correlation between revenues & size of workforce (hence called linear models). This puts significant demand on continuous hiring. To circumvent poor availability of qualified talent in the market most large to mid sized organizations mastered the art of hiring, training & deploying talent directly from the colleges. I won't be off if I were to say that Indian IT services firms have mastered this art more than any other industry in the world.

Let's now spend some time on various non-linear models being practiced in the industry. As the name suggests, the primary goal is to break the linearity between revenue growth & employee headcount.

d) License Fee-based - Perhaps this is where one begins to make a start towards nonlinear models. Few questions arise. Why would a services company create products ? In reality it is not the license revenue it is after but it is the services around the product like integration, maintenance, support & operations that is more rewarding. For now let's answer why would a services company even think of products or product based models. Two reasons 1) Product demonstrates competence and helps win related deals 2)  It is a stickier revenue, a huge plus for services company where revenue backlogs are always thin. But isn't services business quiet different from products ? Yes it is and this is one of the reasons why services companies have done a below average job of creating products as most of them have tried running both businesses together, a strategy that seems flawed at the onset. 

e) Platform based Services - This model marries product with services. Most product companies are not setup to provide more than basic professional services to integrate its product in a client's environment. Just like services companies struggle with products, product companies struggle with professional services because of two reasons 1) Inability to effectively create & manage bench and 2) High cost (and hence bill rates) due to lack of de-skilling of complex tasks. This is where traditional services companies do a good job by not only creating & managing bench strength but also build low cost talent pool that can not only provide integration services but also provide operations around it at an optimal cost by leveraging low-cost centers.

f) Gain Sharing - This is an outcome based model that tries to create a WIN-WIN for both clients & vendors. Clients get service usually at an upfront discount in lieu of sharing the gains with the vendors at a later time. My own experience has been that most clients like the idea initially but the proposal usually gets shot down at the CFO/CEO level as it is often considered as a risky proposition.

g) Joint IP Solutions - Clients & vendors agree to jointly work on solving a business problem. Typically most investments are made by the vendor who is able to create innovative solutions by combining client's sponsorship with it's own agility & lower cost. Vendors gain by not only learning domain but creating solutions that can be taken to market. Clients gain because get the required resources to create these solutions which otherwise wouldn't have been possible because of lack of funding & bandwidth.

h) SaaS - Software as a service isn't new and has been existing for last 15 years. In this model clients typically pays per use and are oblivious of what the vendor needs to do to provide a service. Salesforce.com is one of the most commonly used SaaS. This model allows Transaction based pricing.

At this point it makes sense to introduce the concept of Business Function Outsourcing (BFO), which in recent years is emerging to replace individual function outsourcing. In BFO a client outsources an entire business function rather than pieces of it.

Now consider a case where an investment bank decides to outsource a business function around middle office settlement operations for Listed Derivatives. In a traditional model following modus operandi is followed:

1. Engage a process consulting vendor to review the entire operations and recommend possible improvements.
2. Engage an IT firm to enhance/develop IT applications based on recommended process improvements
3. Setup an IT team to make required changes and same or a different team for ongoing maintenance.
4. Engage a QA team to test the changes.
5. Setup operations team to perform day-to-day operations.
6. Setup a production support team to support the IT applications

If you notice there are multiple touch points by the bank with one or more vendors to perform 1-6 above. There is also significant time spent by the bank in managing these relationships and in coordination, prioritization & in oversight.

Alternately the bank could decide to outsource the entire settlement function. In such situations vendor would either use the client's existing application suite or implement an industry known platform or even write a new one (rare). All IT & operations are therefore managed by a single vendor. SaaS & Platform based services are two models that allow complete business function outsourcing. This also helps the clients turn their fixed spend into variable spend as BFO makes transaction based pricing possible.

Headstrong (A GENPACT company) has a platform called Teevra that allows clients to do real-time settlement & allocation of listed derivatives. Teevra even provides connectivity to 15 different clearinghouses. This platform provides "Settlement as a service" to our clients.

Trust this provides a good overview of various models. Do write to me in case you have any feedback or would like to suggest a topic for future.