Agiles Projektmanagement in Data Science Projekten

Four Pitfalls of Agile Work in Data Science Teams (And How You Can Avoid Them)

Share post:

Background

The Agile Manifesto was written in 2001 by 17 software engineers who wanted to give the typical software develo­p­ment projects of their day an overhaul: instead of detailed planning for tasks years in the future, unrea­li­stic timelines and a plethora of unneces­sary documen­ta­tion, they advocated for delive­ring software that works and meets the customer’s needs in a timely fashion.

Over two decades later, agile metho­do­lo­gies like Scrum have become the de-facto standard not only in software develo­p­ment but also in adjacent fields like, for instance, data science.

Whoever has practiced agile methods in both software and data science, will have noticed that the effects on data science products and projects can differ from those on software teams quite a bit. There are a number of common pitfalls that can lead to diffi­cul­ties delive­ring the desired value to stake­hol­ders and clients and might yield plenty of frustra­tion for everyone involved in the meantime. In this blog post we will explore four common patterns and some ways to keep them from raining on your agile data science parade, regard­less of your role in the agile team.

Big tasks are hard to break down and estimate

Due to their explo­ra­tive nature, the early phases of data science projects are parti­cu­larly diffi­cult to estimate!

Especi­ally the early phases of any data science endeavor are explo­ra­tory in nature. Whether it is the actual data explo­ra­tion phase or ridding the datasets of outliers and other impuri­ties: more often than not you just cannot tell how long it will take. In the same vein, it can be equally challen­ging to break down a big task into smaller incre­ments, since you just cannot antici­pate what you will end up dealing with. This problem is typically not as big of an issue in software develo­p­ment and agile frame­works do not come forth with straight­for­ward remedies to this. One way to deal with these issues is to let go of some “comme il faut” prescrip­tions on how tickets are to be written and estimated, regard­less of how useful and well-inten­tioned they can be in software develo­p­ment teams.

DO:

Give it your best shot at making the stories and tasks as granular as you can and let “good enough” be good enough. Define crystal clear accep­tance criteria for each incre­ment and trust whoever will work on the task to figure out their way to reach the stated goal. If it helps to have the data scien­tists define subtasks as they progress with their task, in order to increase trans­pa­rency and struc­ture, go for it.

DON'T:

Let a task become a never-ending story (pun only semi-intended) without a clear outcome in sight. Just because the path to a goal is not comple­tely charted, does not mean that time and resources cannot be managed. If the goal is clearly speci­fied, effec­tive and frequent commu­ni­ca­tion within the team will show if meaningful progress is being made. Timeboxing certain tasks (e.g. “let’s give this idea a try for two days maximum to see if it works”) is also a helpful way to be prudent about precious develo­p­ment time while not negle­c­ting experi­men­ta­tion.

Depen­den­cies to other teams slow down progress

Data science teams are especi­ally prone to be depen­dent on other teams to deliver neces­sary inputs or feedback. The data might be managed in a dedicated DWH team, the cloud infra­struc­ture is in the hands of the internal IT and business stake­hol­ders are scattered all over the place. Maybe you work with a client that strug­gles to provide infor­ma­tion in a timely fashion. The impedi­ments that result from these circum­s­tances are a staple complaint in agile retro­s­pec­tives. Aside from re-desig­ning the entire organiza­tion to serve the needs of the data science unit, what could be done to remedy this pain point?

Team Abhängigkeiten
Data science teams are parti­cu­larly suscep­tible to being depen­dent on other teams

DO:

Clarify within the team what you need from other people or depart­ments on a regular basis and propose a low-effort way for the other party to provide these items. Effec­tive commu­ni­ca­tion with the right people can absolutely make or break a data science project or data product develo­p­ment. This holds equally true for client projects and any initia­tives within a company.

DON'T:

Host pity parties in dailies and retro­s­pec­tives without taking owner­ship of the success of the endeavor and propo­sing ways to make it happen despite organiza­tional deficits. This is where empowered agile coaches can make a big diffe­rence to the team’s impact on their company or client.

Commu­ni­ca­tion and business objec­tives are misali­gned with the data science team

Sprache

Data science teams (similarly to software dev teams) at times speak quite a diffe­rent language than their stake­hol­ders or clients. It is not uncommon to measure the contri­bu­tion of a product owner by how effec­tive they are able to bridge this gap. Nevert­heless, everyone in an agile team can and should contri­bute to being in alignment with business objec­tives and efficient commu­ni­ca­tion with stake­hol­ders and clients.

Each member of the data science team should know his/her contri­bu­tion to the company's goals

DO:

Be very clear on what your stake­hol­ders or clients are trying to achieve and how the data science team contri­butes to those goals. Make sure that the reviews and other meetings enable the stake­hol­ders to under­stand what progress is being made and the benefit that is provided to the company or the client. If your stake­hol­ders would benefit from a crash course on data science termi­no­logy and concepts, do not hesitate to offer such training to them because it is bound to improve commu­ni­ca­tion going forward. In the end the data science team benefits the most from their stake­hol­ders parti­ci­pa­tion and sugges­tions.

DON'T:

Invite stake­hol­ders not familiar with the details of data science to the review and proceed to discuss highly technical concepts because they happen to be interes­ting to a hypothe­tical data science audience. The reviews are a valuable occasion to elicit stake­holder feedback and help to improve the outcomes for the business. Technical discus­sions ought to take place in a more appro­priate setting if the stake­hol­ders are not equipped to meaningfully contri­bute to them.

Adapting an agile mindset proves challen­ging

Data science as a disci­pline has been booming for just over a decade, and specia­lized courses have become incre­asingly common over the years. Nonetheless, the majority of data scien­tists are still recruited from STEM, econo­mics and other quanti­ta­tive fields taught in higher educa­tion. Such an academic background brings with it a set of values and attitudes that, for example, empha­size thorough­ness and indivi­dual perfor­mance over experi­men­ta­tion and teamwork. Getting comfor­table with the values and expec­ta­tions of agile work can be deman­ding and counter­in­tui­tive to many indivi­duals. It certainly takes time and a healthy dose of courage.

It takes courage to embrace the agile way of working
Mut

DO:

Be very clear on the team’s expec­ta­tions regar­ding values and cultural norms. Give construc­tive feedback and discuss openly what is working and what is not.

DON'T:

Fall prey to stereo­types about certain profes­sions or academic backgrounds. People learn and adapt to their environ­ment more than we sometimes think is possible. The culture of the group and the incen­tive struc­tures people find themselves in, shape how people interact with their environ­ments. Thus it is an exercise in leader­ship to shape said culture and incen­tives to produce the desired results.

Conclu­sion

Even though the agile metho­do­logy was not neces­s­a­rily designed with data science in mind, with a few tweaks it can be a natural fit. Commu­ni­ca­tion and a growth mindset are every bit as important to the success of a data science product or project as the code and the math involved.

This list is far from exhaus­tive but attempts to shed some light on very common patterns that baffle plenty of product owners, agile coaches and data scien­tists. With artifi­cial intel­li­gence gaining relevance these days, new pheno­mena are bound to emerge and might warrant a sequel to this article at a future point in time.

Picture of Natalia Grinberg

Natalia Grinberg

Lead Data Scien­tist

Projektanfrage

Vielen Dank für Ihr Interesse an den Leistungen von m²hycon. Wir freuen uns sehr, von Ihrem Projekt zu erfahren und legen großen Wert darauf, Sie ausführlich zu beraten.

Von Ihnen im Formular eingegebene Daten speichern und verwenden wir ausschließlich zur Bearbeitung Ihrer Anfrage. Ihre Daten werden verschlüsselt übermittelt. Wir verarbeiten Ihre personenbezogenen Daten im Einklang mit unserer Datenschutzerklärung.

Project request

Thank you for your interest in m²hycon’s services. We look forward to hearing about your project and attach great importance to providing you with detailed advice.

We store and use the data you enter in the form exclusively for processing your request. Your data is transmitted in encrypted form. We process your personal data in accordance with our privacy policy.