Project prioritization

This is a long read, skip to “Prioritizing the projects and changes” for the approach details...

Organizations and companies generally have an IT workload (dare I say, backlog?) which needs to be properly assessed, prioritized and taken up. Sometimes, the IT team(s) get an amount of budget and HR resources to "do their thing", while others need to continuously ask for approval to launch a new project or instantiate a change.

Sizeable organizations even require engineering and development effort on IT projects which are not readily available: specialized teams exist, but they are governance-wise assigned to projects. And as everyone thinks their project is the top-most priority one, many will be disappointed when they hear there are no resources available for their pet project.

So... how should organizations prioritize such projects?

Structure your workload, the SAFe approach

A first exercise you want to implement is to structure the workload, ideas or projects. Some changes are small, others are large. Some are disruptive, others are evolutionary. Trying to prioritize all different types of ideas and changes in the same way is not feasible.

Structuring workload is a common approach. Changes are grouped in projects, projects grouped in programs, programs grouped in strategic tracks. Lately, with the rise in Agile projects, a similar layering approach is suggested in the form of SAFe.

In the Scaled Agile Framework a structure is suggested that uses, as a top-level approach, value streams. These are strategically aligned steps that an organization wants to use to build solutions that provide a continuous flow of value to a customer (which can be internal or external). For instance, for a financial service organization, a value stream could focus on 'Risk Management and Analytics'.

SAFe full framework

SAFe full framework overview, picture courtesy of www.scaledagileframework.com

The value streams are supported through solution trains, which implement particular solutions. This could be a final product for a customer (fitting in a particular value stream) or a set of systems which enable capabilities for a value stream. It is at this level, imo, that the benefits exercises from IT portfolio management and benefits realization management research plays its role (more about that later). For instance, a solution train could focus on an 'Advanced Analytics Platform'.

Within a solution train, agile release trains provide continuous delivery for the various components or services needed within one or more solutions. Here, the necessary solutions are continuously delivered in support of the solution trains. At this level, focus is given on the culture within the organization (think DevOps), and the relatively short-lived delivery delivery periods. This is the level where I see 'projects' come into play.

Finally, you have the individual teams working on deliverables supporting a particular project.

SAFe is just one of the many methods for organization and development/delivery management. It is a good blueprint to look into, although I fear that larger organizations will find it challenging to dedicate resources in a manageable way. For instance, how to deal with specific expertise across solutions which you can't dedicate to a single solution at a time? What if your organization only has two telco experts to support dozens of projects? Keep that in mind, I'll come back to that later...

Get non-content information about the value streams and solutions

Next to the structuring of the workload, you need to obtain information about the solutions that you want to implement (keeping with the SAFe terminology). And bear in mind that seemingly dull things such as ensuring your firewalls are up to date are also deliverables within a larger ecosystem. Now, with information about the solutions, I don't mean the content-wise information, but instead focus on other areas.

Way back, in 1952, Harry Markowitz introduced Modern portfolio theory as a mathematical framework for assembling a portfolio of assets such that the expected return is maximized for a given level of risk (quoted from Wikipedia). This was later used in an IT portfolio approach by McFarlan in his Portfolio Approach to Information Systems article, published in September 1981.

There it was already introduced that risk and return shouldn't be looked at from an individual project viewpoint, but how it contributes to the overall risk and return. A balance, if you wish. His article attempts to categorize projects based on risk profiles on various areas. Personally, I see the suggested categorization more as a way of supporting workload assessments (how many mandays of work will this be), but I digress.

Since then, other publications came up which tried to document frameworks and methodologies that facilitate project portfolio prioritization and management. The focus often boils down to value or benefits realization. In The Information Paradox John Thorp comes up with a benefits realization approach, which enables organizations to better define and track benefits realization - although it again boils down on larger transformation exercises rather than the lower-level backlogs. The realm of IT portfolio management and Benefits realization management gives interesting pointers as to the lecture part of prioritizing projects.

Still, although one can hardly state the resources are incorrect, a common question is how to make this tangible. Personally, I tend to view the above on the value stream level and solution train level. Here, we have a strong alignment with benefits and value for customers, and we can leverage the ideas of past research.

The information needed at this level often boils down to strategic insights and business benefits, coarse-grained resource assessments, with an important focus on quality of the resources. For instance, a solution delivery might take up 500 days of work (rough estimation) but will also require significant back-end development resources.

Handling value streams and solutions

As we implement this on the highest level in the structure, it should be conceivable that the overview of the value streams (a dozen or so) and solutions (a handful per value stream) is manageable, and something that at an executive level is feasible to work with. These are the larger efforts for structuring and making strategic alignment. Formal methods for prioritization are generally not implemented or described.

In my company, there are exercises that are aligning with SAFe, but it isn't company-wide. Still, there is a structure in place that (within IT) one could map to value streams (with some twisting ;-) and, within value streams, there are structures in place that one could map to the solution train exercises.

We could assume that the enterprise knows about its resources (people, budget ...) and makes a high-level suggestion on how to distribute the resources in the mid-term (such as the next 6 months to a year). This distribution is challenged and worked out with the value stream owners. See also "lean budgeting" in the SAFe approach for one way of dealing with this.

There is no prioritization of value streams. The enterprise has already made its decision on what it finds to be the important values and benefits and decided those in value streams.

Within a value stream, the owner works together with the customers (internal or external) to position and bring out solutions. My experience here is that prioritization is generally based on timings and expectations from the customer. In case of resource contention, the most challenging decision to make here is to put a solution down (meaning, not to pursue the delivery of a solution), and such decisions are hardly taken.

Prioritizing the projects and changes

In the lower echelons of the project portfolio structure, we have the projects and changes. Let's say that the levels here are projects (agile release trains) and changes (team-level). Here, I tend to look at prioritization on project level, and this is the level that has a more formal approach for prioritization.

Why? Because unlike the higher levels, where the prioritization is generally quality-oriented on a manageable amount of streams and solutions, we have a large quantity of projects and ideas. Hence, prioritization is more quantity-oriented in which formal methods are more efficient to handle.

The method that is used in my company uses scoring criteria on a per-project level. This is not innovative per se, as past research has also revealed that project categorization and mapping is a powerful approach for handling project portfolio's. Just look for "categorizing priority projects it portfolio" in Google and you'll find ample resources. Kendal's Advanced Project Portfolio Management and the PMO (book) has several example project scoring criteria's. But allow me to explain our approach.

It basically is like so:

  1. Each project selects three value drivers (list decided up front)
  2. For the value drivers, the projects check if they contribute to it slightly (low), moderately (medium) or fully (high)
  3. The value drivers have weights, as do the values. Sum the resulting products to get a priority score
  4. Have the priority score validated by a scoring team

Let's get to the details of it.

For the IT projects within the infrastructure area (which is what I'm active in), we have around 5 scoring criteria (value drivers) that are value-stream agnostic, and then 3 to 5 scoring criteria that are value-stream specific. Each scoring criteria has three potential values: low (2), medium (4) and high (9). The numbers are the weights that are given to the value.

A scoring criteria also has a weight. For instance, we have a scoring criteria on efficiency (read: business case) which has a weight of 15, so a score of medium within that criteria gives a total value of 60 (4 times 15). The potential values here are based on the "return on investment" value, with low being a return less than 2 years, medium within a year, and high within a few months (don't hold me on the actual values, but you get the idea).

The sum of all values gives a priority score. Now, hold your horses, because we're not done yet. There is a scoring rule that says a project can only be scored by at most 3 scoring criteria. Hence, project owners need to see what scoring areas their project is mostly visible in, and use those scoring criteria. This rule supports the notion that people don't bring around ideas that will fix world hunger and make a cure for cancer, but specific, well scoped ideas (the former are generally huge projects, while the latter requires much less resources).

OK, so you have a score - is that your priority? No. As a project always falls within a particular value stream, we have a "scoring team" for each value stream which does a number of things. First, it checks if your project really belongs in the right value stream (but that's generally implied) and has a deliverable that fits the solution or target within that stream. Projects that don't give any value or aren't asked by customers are eliminated.

Next, the team validates if the scoring that was used is correct: did you select the right values (low, medium or high) matching the methodology for said criteria? If not, then the score is adjusted.

Finally, the team validates if the resulting score is perceived to be OK or not. Sometimes, ideas just don't map correctly on scoring criteria, and even though a project has a huge strategic importance or deliverable it might score low. In those cases, the scoring team can adjust the score manually. However, this is more of a fail-safe (due to the methodology) rather than the norm. About one in 20 projects gets its score adjusted. If too many adjustments come up, the scoring team will suggest a change in methodology to rectify the situation.

With the score obtained and validated by the scoring team, the project is given a "go" to move to the project governance. It is the portfolio manager that then uses the scores to see when a project can start.

Providing levers to management

Now, these scoring criteria are not established from a random number generator. An initial suggestion was made on the scoring criteria, and their associated weights, to the higher levels within the organization (read: the people in charge of the prioritization and challenging of value streams and solutions).

The same people are those that approve the weights on the scoring criteria. If management (as this is often the level at which this is decided) feels that business case is, overall, more important than risk reduction, then they will be able to put a higher value in the business case scoring than in the risk reduction.

The only constraint is that the total value of all scoring criteria must be fixed. So an increase on one scoring criteria implies a reduction on at least one other scoring criteria. Also, changing the weights (or even the scoring criteria themselves) cannot be done frequently. There is some inertia in project prioritization: not the implementation (because that is a matter of following through) but the support it will get in the organization itself.

Management can then use external benchmarks and other sources to gauge the level that an organization is at, and then - if needed - adjust the scoring weights to fit their needs.

Resource allocation in teams

Portfolio managers use the scores assigned to the projects to drive their decisions as to when (and which) projects to launch. The trivial approach is to always pick the projects with the highest scores. But that's not all.

Projects can have dependencies on other projects. If these dependencies are "hard" and non-negotiable, then the upstream project (the one being dependent on) inherits the priority of the downstream project (the one depending on the first) if the downstream project has a higher priority. Soft dependencies however need to validate if they can (or have to) wait, or can implement workarounds if needed.

Projects also have specific resource requirements. A project might have a high priority, but if it requires expertise (say DBA knowledge) which is unavailable (because those resources are already assigned to other ongoing projects) then the project will need to wait (once resources are fully allocated and the projects are started, then they need to finish - another reason why projects have a narrow scope and an established timeframe).

For engineers, operators, developers and other roles, this approach allows them to see which workload is more important versus others. When their scope is always within a single value stream, then the mentioned method is sufficient. But what if a resource has two projects, each of a different value stream? As each value stream has its own scoring criteria it can use (and weight), one value stream could systematically have higher scores than others...

Mixing and matching multiple value streams

To allow projects to be somewhat comparable in priority values, an additional rule has been made in the scoring methodology: value streams must have a comparable amount of scoring criteria (value drivers), and the total value of all criteria must be fixed (as was already mentioned before). So if there are four scoring criteria and the total value is fixed at 20, then one value stream can have its criteria at (5,3,8,4) while another has it at (5,5,5,5).

This is still not fully adequate, as one value stream could use a single criteria with the maximum amount (20,0,0,0). However, we elected not to put in an additional constraint, and have management work things out if the situation ever comes out. Luckily, even managers are just human and they tend to follow the notion of well-balanced value drivers.

The result is that two projects will have priority values that are currently sufficiently comparable to allow cross-value-stream experts to be exchangeable without monopolizing these important resources to a single value stream portfolio.

Current state

The scoring methodology has been around for a few years already. Initially, it had fixed scoring criteria used by three value streams (out of seven, the other ones did not use the same methodology), but this year we switched to support both value stream agnostic criteria (like in the past) as well as value stream specific ones.

The methodology is furthest progressed in one value stream (with focus of around 1000 projects) and is being taken up by two others (they are still looking at what their stream-specific criteria are before switching).

more ...

Structuring infrastructural deployments

Many organizations struggle with the all-time increase in IP address allocation and the accompanying need for segmentation. In the past, governing the segments within the organization means keeping close control over the service deployments, firewall rules, etc.

Lately, the idea of micro-segmentation, supported through software-defined networking solutions, seems to defy the need for a segmentation governance. However, I think that that is a very short-sighted sales proposition. Even with micro-segmentation, or even pure point-to-point / peer2peer communication flow control, you'll still be needing a high level overview of the services within your scope.

In this blog post, I'll give some insights in how we are approaching this in the company I work for. In short, it starts with requirements gathering, creating labels to assign to deployments, creating groups based on one or two labels in a layered approach, and finally fixating the resulting schema and start mapping guidance documents (policies) toward the presented architecture.

more ...

Switching focus at work

Since 2010, I was at work responsible for the infrastructure architecture of a couple of technological domains, namely databases and scheduling/workload automation. It brought me in contact with many vendors, many technologies and most importantly, many teams within the organization. The focus domain was challenging, as I had to deal with the strategy on how the organization, which is a financial institution, will deal with databases and scheduling in the long term.

more ...

Making the case for multi-instance support

With the high attention that technologies such as Docker, Rocket and the like get (I recommend to look at Bocker by Peter Wilmott as well ;-), I still find it important that technologies are well capable of supporting a multi-instance environment.

Being able to run multiple instances makes for great consolidation. The system can be optimized for the technology, access to the system limited to the admins of said technology while still providing isolation between instances. For some technologies, running on commodity hardware just doesn't cut it (not all software is written for such hardware platforms) and consolidation allows for reducing (hardware/licensing) costs.

more ...