Human project management in the age of machine learning.

“Alexa, please design and launch the new corporate website!” Until that day is possible, why does web design and development benefit from a human project manager? Why can’t we just streamline the process and expedite results by building code right away and launch ASAP? Why not just have software itself figure out the best solutions as well as the most streamlined way to get to the solution? With all the talk of AI and machine learning, can’t we just feed all our current website and marketing issues into a software application and have it tell us the right website design and marketing strategy?

There are inherent trade-offs in the real world that make it such that there is no one “right answer”. A human being working with other human beings is the best way currently to navigate the trade-offs of product design and development. A person acting as referee and coach can help unveil the questions, conflicts, risks and benefits of different potential solutions.

A project manager working as referee can help a team to see potential answers within complex and conflicting situations. A project manager can help teams to get to a solution, to force decisions, or help the team agree on a best path forward. A project manager can help create the social fabric that gives rise to decisions and then helps support those decisions once made. A project manager does not do this alone, yet by understanding and helping to bring out the talents and perspectives of the contributors and stakeholders.

First, let us start with the understanding that some problems are “Wicked Problems”1. These problems differ from “Tame Problems”. Wicked Problems have no easy answer, or often a way to know if any answer is correct or not. Tame problems are things like “1 + 1 =”, or “Who won the 2017 NBA Championship?”. Tame Problems have agreed upon answers. In a Wicked Problem multiple factors make it hard to answer. For instance, take the problem of a family wanting to remodel their house by putting a new window facing-west to take advantage of the view of the ocean. The family wants the window to be as large as possible to have the best view, yet they also mandate that it be as small as possible so as to avoid heat-loss through the window. What is the best size for the window? It depends on who you ask. The husband really wants it bigger. The wife really wants it smaller? Who wins? What is the best answer? Again, it is difficult to decide whether aesthetics or finance are most important. Will the family benefit most from enjoying an amazing view, or by saving money so that they can enjoy more vacations?

Problems: Tame vs Wicked
Wicked problems have no single correct answer and require teams to discuss them and make choices.

Computers and software are good at solving Tame Problems. To date, software has a harder time with Wicked Problems. A project manager can help groups of people to make decisions when faced with Wicked Problems. Take some web design scenarios: Should we have a green button or an orange button? Why that photograph versus the other photo? Should we wait to add cart personalization until Phase-2, or it is critical for Phase-1? The CEO really wants the pants category at the top, even though the CMO wants the skirts category at the top. Which is the best decision? Which way should we do the pagination? With the AJAX-ing “load more”, “load more”, “load more”, or with the “Page X of Y” and the left and right arrows? There is no single correct answer.

Machine learning and AI promise many things. In the future, smart systems and more advanced software will increasingly support humans in making decisions. Yet, as of 2018, there is no software to which you can turn over your all your team decision making. People still need to get together and hash out their perspectives, defend what matters to their constituencies, and then come to agreement about what will be in a product and what will be out. Plus, though it may sound axiomatic, there is no better way for people to work together, than with people.

The Problem Space - Exploring Alternatives
The number of inputs and unknowns actually grows during a project timeline, making it impossible to know all the conditions at the start of a project.

As a client services professional and project manager I have some self-interest in pointing out the value of human project managers. Besides this selfishness though, I challenge anyone or any organization to successfully have software alone design and launch a complex public website. Especially, within a mid- to large sized organization with different stakeholder groups and different agendas. There is value in the human touch. In listening. In interpreting. In helping to foster the dialogue that results in decisions.

One might say that advanced AI fueled by “all the facts” could make the decisions, or do A/B testing and find out the right choices, and then implement them. Yet, for now, AI does not have the insights. From the examples above: The orange buttons are the correct choice because Mary the creative director knows that next summer all the branding is changing to orange for indicating emphasis. Green would be a bad choice; the founder hates the color green (a fact that is not in any database). As for the question of pants or skirts as the top menu category? It was agreed to go with skirts at top after learning that an article is coming out in Vanity Fair featuring the skirts, even though the article has not yet been announced. All of the research and decisions were fostered by human-to-human communication.

Without interim demonstrations or prototypes to collectively review, the various parties in a project may be envisioning different desires.
Though projects and project management may not be ready to be handed off completely to robots, perhaps projects are ready for cyborgs. That is, actual people whose power is amplified by emerging tech. Whether by using Slack, Jira, or web analytics, the human project manager can become the Six Million Dollar Man by being modded out with super tools that let them rock. Optimal design process and optimal communication can be aided by tools, as opposed to replaced by them. People bring value to the results. Designers are key to good design. Developers are key to good software. Project managers are key to good projects. When planning your next web or app update, find some smart developers and designers, know your market, … and find a good project manager.

1. “Wicked Problem”, Wikipedia

Previous Post Creating Useful Style Guides for Web Projects - Part 1
Next Post Project Momager