Computing Applications Viewpoint

Where Should Your IT Constraint Be?: The Case of the Financial Services Industry

Locating the strategic location of the IT junction constraint.
  1. Introduction
  2. Recent Tactical Work
  3. Constraint Positioning "Immediate Suspects"
  4. Analysis of Persona Resource Usage Profile
  5. Discussion
  6. Conclusion
  7. References
  8. Authors
  9. Footnotes
colored graphs

The financial services industry is in a state of turmoil due to technological innovation and the industry’s move from brick-and-mortar to digital and particularly mobile service delivery. In most cases, information technology (IT) function is the constraint blocking the organization from innovating rapidly.

In this Viewpoint, we seek to locate where the constraint of the IT function should strategically be. In recent work2,3,4,5,6 the focus was on the tactical management of a system using the theory-of-constraints (TOC). Research done so far has not addressed the strategic location of the constraint.

Here, we focus on the financial services industry. Though digitalization is applicable to a multitude of industries, we focus on the financial services industry due to its particular characteristics: dramatic reduction of human resources, significant increase of self-service, end-to-end computerization due to the lack of physical entities, increasing IT budgets, and a similarity of problems between banks, insurance, and credit card companies.

Back to Top

Recent Tactical Work

The theory-of-constraints3 has developed a methodology to identify the system’s current constraint and manage it, commonly using the seven focusing steps (see Figure 1) as the tool for this purpose.6

Figure 1. The seven focusing steps.

The seven steps set the organization’s goal and performance measures, identify the current system constraint, exploit the constraint, subordinate other functions to the constraint and elevate (offload) the constraint. If a constraint is broken, attention is turned to the new constraint. These techniques have been documented to increase the organization’s throughput, significantly improving its performance. This Viewpoint deals with the strategic management of the constraint, addressing the question of where the constraint should be. A common graphical tool to identify the system’s constraint is the Cost-Utilization analysis,6 which looks at the cost of each organizational function and at the extent to which it is utilized. Functions that are fully utilized are the organization’s bottlenecks. The constraint should be located in the most expensive/scarce resource.6

Figure 2 presents the Cost-Utilization diagram of the financial industry. Previous research5 has shown that IT is a permanent bottleneck, because demand for IT services is X3 to X5 the capacity of the IT function and because of the scarcity of good IT professionals and partners. Here, “IT function” applies to supporting the whole software life cycle. As seen in Figure 2, the cost (represented by the width of the horizontal axis) is large and the utilization is 100%. Later on, we will drill down to identify specifically where in the IT function the constraint should be located. Demand is the result of the needs of all organizational functions (marketing, operations, sales, financial markets, and so forth), with compliance, maintenance of existing systems, change requests, fix-it jobs, technological project requirement, and so forth, accounting for a significant portion of the demand.

Figure 2. Cost-Utilization diagram of the financial industry.

As shown in Figure 2, sales is a permanent bottleneck because demand is endless (upsell, cross-sell, new customers and so forth) Figure 2 illustrates that other functions such as operations, logistics, legal services, and others, should never be the organization’s constraint.

A current-reality-tree (CRT) of the financial services industry is presented in Grosfeld-Nir et al.;4 this is based on Goldratt’s work.2 For this Viewpoint, we analyzed the phenomenon that ‘Insufficient value is generated through digitalization” and drew the fCRT (focused-Current-Reality-Tree), a lighter version of Goldratt’s CRT6 (see Figure 3).

Figure 3. Focused Current Reality Tree.

The fCRT presents the leading Undesirable-Effect (UDE) that insufficient value generated through digitalization. All other UDEs account directly or via other cause-and-effect relationships to the leading UDE and stem from the root-causes: ‘Not enough differentiation of users’ and ‘No clear identification of constraint location’. Here, we deal with these root causes.

Back to Top

Constraint Positioning “Immediate Suspects”

Given the digitalization flow, we now search for potential resources constraining IT digitalization value creation. Five ‘immediate suspects’ are identified: Budgeting, Analysis and design, IT development, Assimilation in the organization, and Customer attention. Figure 4 describes bottleneck positioning ‘immediate suspects’. These are plausible constraints:

Figure 4. Constraint positioning ‘immediate suspects’.

  1. Budgeting. Managers will always complain the budget is insufficient. This is often the case, but we have encountered a dozen large financial institutions that did not utilize all of their budgets by year-end. The financial mind-set is that if you provide enough of it, money will solve resource shortage. In reality, the pace of growth is limited. The budgeting process should be divided into two steps: first, how much funding is management and the board of directors willing/capable to allocate, given large, multi-year infrastructure projects; and second, how much to allocate to each department/project.
  2. Analysis and design. In many organizations, the system analysts are part of the business-relationship-management (BRM) function, and constrain new project development. It is not cost effective that analysts and designers should be the system’s constraint. Imagine the board of directors’ reaction should the CIO argue that projects are late due to the shortage of 10 analysts or designers. Despite their scarcity, their numbers are not large and their cost is manageable. This problem is solvable within the budget.
  3. IT Development. Development resources are expensive (whether internal or outsourced) and often block development throughput. Once the budget is allocated, projects should go through a value-based strategic gating process, eliminating projects that do not generate enough value or return-on-investment (ROI). In some cases, with a positive ROI management is unwilling to increase its IT headcount or budget. From our experience, the throughput of the IT development function can be significantly improved by using tactical methodologies such as the complete-kit concept, TOC, tactical gating, proper measurement, introduction of DevOps tools, and so forth.5
  4. Assimilation in the organization. There is a limit to the amount of new systems the organization can absorb at a given time (see Davenport and Beck1). We have not encountered an instance where assimilation was the constraint and the whole organization was subordinated to it. Assimilation should be planned and staggered over time. Agile methodology can be used to sequentially launch features in an effective pace. The introduction of too many systems at a time results in chaos and bad multitasking. Thus, application of scheduling tools can smooth the load on the organization.
  5. Customer attention. In today’s disruptive arena, many customers fail to become familiar with, and use, available applications introduced by the organization. More systems/functions are introduced than the customer is capable of adopting. The customer’s journey illustrates customer personas aware of a new digital service, examining it, registering for it, actually using it and continuing doing so. Figure 5 illustrates the journeys of three personas: seniors (persona A), performing their interactions physically in the branch; baby boomers (persona B), searching the Web, examining by phone, being coached in the branch and then using the mobile app.; millennials (persona C) performing all journey phases via their mobile phones (see Figure 5).

Figure 5. Customer journeys.

Pitfalls and hurdles along the way result in customer usage churn. Consider the paradox that after spending scarce, expensive capital, effort and budgeting the IT department, solving specification bottlenecks, screening projects in the development department, improving the development department, solving assimilation problems, and then—being blocked on the customer attention bottleneck; we next analyze the three personas in terms of their attention and usage profile.

Back to Top

Analysis of Persona Resource Usage Profile

Pareto analysis is one of the most powerful tools in management stating that 20% of the phenomena account for 80% of the effect, also known as the 80/20 rule. The Pareto methodology is an extension of the Pareto rule stating that improving a system is accomplished by the following three steps: classification, differentiation, and resource allocation.4 Dealing with customer attention, we can apply this methodology as follows:

Step 1: Classification. We classify the customers as follows: “A” customers are 30% of the customers that account for 70% of face-to-face and phone support. These are mostly senior customers unaccustomed with many technologies. “B” customers are 40% of the customers that account for 20% of face-to-face and phone support. These are mostly baby boomers who will use digital services if provided with friendly assistance. “C” customers are 30% of the population that account for 10% of face-to-face and telephone traffic. They are technically savvy millennials (who would rather go to the dentist than visit a banka). They can handle it alone, needing occasional assistance via Google, chat, or videos.

Step 2: Differentiation. Adopting a different policy for each group. No amount of effort will convert “A” customers to digital. Our challenge is to shift them to the telephone channel. Branch personnel should patiently train them to use the telephone to reduce the load on the branch. They are not a target for digital banking, and no resources are allocated to convert them. They are therefore not a constraint. The attention span of “B” customers limits their adoption of new technologies. They are the focus of technology introduction as they account for the lion’s share of our target population (estimated 40% of the population that account for merely 20% of the face-to-face traffic). “C” customers crave digital services, can assimilate all the changes and are not a constraint.

Step 3: Resource Allocation. “A” customers are allocated telephone services delivering basic bread-and-butter services and dedicated staff in the branch training them to use these services. “B” customers are allocated technology facilitation resources and are the focal point of the digital world. “C” customers are provided with virtually paperless mobile banking services.

The IT-development potential bottleneck can be resolved by a value-based strategic gating process and applying tactical tools such as the complete-kit concept and others.

Back to Top


In our journey, we have analyzed several ‘immediate suspects’: budgeting, analysis and design, IT development, assimilation in the organization, and customer attention. Analysis and design is not the constraint as reasonable investment can resolve the problem. The IT-development potential bottleneck can be resolved by a value-based strategic gating process and applying tactical tools such as the complete-kit concept and others, and sometimes budget increment. Assimilation in the organization should not be the bottleneck and can be resolved by smoothing the peak demand. Customer attention should not be the bottleneck as only B population requires organizational resources to adopt new technologies.

Traveling through potential bottlenecks has brought us to the realization the budget should be the constraint. However, management should ensure:

  • the budget is allocated through a structured and value-based strategic gating process;
  • analysis and design is budgeted for and should not be a bottleneck;
  • development must undergo drastic improvement that will reveal capacity to the system;
  • assimilation should not be the bottleneck, and if it is, then management should implement smoothing of feature launch; and
  • from all personas only the “B” population (baby boomers) consume system resources, they are not considered the constraint since they constitute only 20% of the population.

Back to Top


There is a set of issues regarding the strategic positioning of a constraint. Managers should resolve the three following issues:

  • Where the constraint should be.
  • Where the constraint is now.
  • How to move the constraint to the desired position.

Analyzing the IT development process, backed by well-accepted methodologies such as TOC and Pareto analysis, we have come to the conclusion that the budget should be the constraint. Executives must generate a well thought-out budget and at the same time ensure other functions do not become the system’s constraint, according to the guidelines here.

IT executives have a significant message to convey: the budget is the constraint. All other issues must be resolved in light of this constraint. The board of directors and management define the budget and the role of the IT executives is to participate in the budgeting process, resolve problems, and lead the analysts and developers according to the budget.

Similar analysis of other industries such as telecom, retail, tourism, and so forth and may yield other results. A company that follows these guidelines will better compete in the new market.

    1. Davenport, T.H. and Beck, J.C. The Attention Economy: Understanding the New Currency of Business. Harvard Business School Press, Boston, MA, 2001.

    2. Goldratt, E.M. It's Not Luck. North River Press, Croton on Hudson, NY, 1994.

    3. Goldratt, E.M. and Cox, J. The Goal: A Process of Ongoing Improvement (2nd ed.). North River Press, Croton on Hudson, NY, 1986.

    4. Grosfeld-Nir, A., Ronen, B. and Kozlovsky, N. The Pareto managerial principle: When does it apply? International Journal of Production Research 45, 10 (Oct. 2007), 2317–2325.

    5. Pass, S. and Ronen, B. Reducing the software value gap, Commun. ACM 57, 5 (May 2014), 80–87.

    6. Ronen, B. and Pass S. Focused Operations Management. John Wiley and Sons, Inc. Hoboken, NJ, 2008.

Join the Discussion (0)

Become a Member or Sign In to Post a Comment

The Latest from CACM

Shape the Future of Computing

ACM encourages its members to take a direct hand in shaping the future of the association. There are more ways than ever to get involved.

Get Involved

Communications of the ACM (CACM) is now a fully Open Access publication.

By opening CACM to the world, we hope to increase engagement among the broader computer science community and encourage non-members to discover the rich resources ACM has to offer.

Learn More