product design
enterprise financial software product

TABLE OF CONTENTS

case study

OVERVIEW

brief

process

DELIVERABLES

videos

links

 

OVERVIEW

Client

-  KinderCare Education
-  Children's Creative Learning Centers
-  Champions
-  The Partners Group

Location

-  Portland, OR

INDUSTRY

-  Child-care education and services
-  Financial / accounting
-  Tuition subsidy

COMPANY PROFILE

-  $1.6B company
-  World's largest early childhood education company

TECHNOLOGIES

-  Enterprise software
-  ERP software (enterprise resource planning) 
-  Financial and accounting software

Product(s)

-  Tuition subsidy financial software  

Project / feature / deliverable

-  Tuition Subsidy v1.0 (MVP)

duration / date

-  15 month release / 2015-2016

NOTE: Don't feel like reading? Watch the videos at the end of the page.
 

background

This case study highlights product design of the Tuition Subsidy v1.0 MVP software product. This background is extensive, but provides context for scope and complexity in all facets of this project. For that reason, I recommend reading it before any of my other KinderCare Education case studies. 

KinderCare Education (formerly Knowledge Universe) is the world's largest early childhood education company with approximately $1.6B in annual revenues. It was founded in 1996 and became the dominant force in the North American child-care education services market, eventually acquiring a number of other companies and business lines.

The company currently has more than 2,000 centers and 31,000 employees in the United States, split among at least three brands and five business lines: KinderCare Learning Centers, Children's Creative Learning Centers, Champions, Grove Schools, and Knowledge Beginnings, among others. Many acquisitions predate the existence of the parent company, Knowledge Universe, by many decades. Furthermore, ownership of the parent company has changed hands numerous times. In 2015, shortly after I began this project, the original parent company, Knowledge Universe, was purchased by The Partners Group, a global private markets management firm based in Switzerland. As a part of the purchase, Knowledge Universe's name changed to KinderCare Education, to take advantage of the size and brand recognition of the flagship business line, KinderCare Learning Centers. 

As a consequence of this complex history, there was a laundry list of monolithic technology issues, all of which profoundly impacted the work described in these case studies:

1. The lack of continuity with corporate ownership and many acquisitions meant there had consequently been no progressive or consistent development of IT infrastructure for years or decades.
    
2. IT infrastructure was not centralized, but existed as an amalgamation of inherited, legacy technologies from the original enterprise, as well as many acquisitions and disparate business lines.
    
3. Some acquired companies were almost 50 years old. Most key software systems were at least 20 years old. The company's most important system, an ERP application, was command-line based, and nearly impossible to enhance or maintain. 
    
4. As an enterprise with core expertise in child-care education, the company did not have the talent necessary to successfully execute IT projects. Instead of building internal teams with a long-term vision of knowledge retention and iterative development, year after year the company turned to different external vendors, agencies, and consultants. As a result, projects that should have been critical, internal initiatives were usually managed piecemeal by various short-term partners. 

The company resolved to address these problems in a multi-year enterprise IT program designed to modernize and unify IT infrastructure. But the nearly unassailable technology landscape that had evolved made for daunting obstacles. Leadership turnover had been extremely high, both within the IT department and the program itself, even at the highest executive levels. The program had actually been started and stalled several times. These factors set the major stage for all work within the program. 

I assumed various, concurrent responsibilities while consulting at KinderCare Education, but my overarching role was project lead for the development of a custom ERP financial software product designed to manage and process tuition subsidy. 

Tuition subsidy is reimbursement of private-pay child-care tuition, a financial benefit often offered to families and students by government agencies and other similar institutions. Approximately 40% of all KinderCare Education students receive this benefit, and at least 33% of the company's $1.6B in revenues come directly from tuition subsidy, making it one of the company's top priorities. The Tuition Subsidy Group that manages this work comprises nearly 200 employees, both at the company's National Support Center (headquarters) in Portland, Oregon, as well as at numerous field sites (child daycare centers) throughout the United States. 

problem

This v1.0 MVP release was limited to contract-management functionality: contracts between agencies and KinderCare Education, and contracts between agencies and students. While contracts might sound superficially simple, these legal documents are extraordinarily long and complex, and the information in them serves as a sort of rules engine to determine all financial calculations, making for a byzantine web of relationships, inter-dependencies, and workflows throughout the entire financial product. Moreover, this complexity is not limited to the tuition subsidy software alone. The new software had to fully integrate with two existing, homegrown ERP systems. Finally, it also had to support two major business lines in one system, where two separate software products had previously existed. 

Despite its importance, the group dedicated to tuition subsidy was built atop a set of decaying and disconnected legacy software systems. Two core applications were failing regularly, and the most important of the two had been developed in a language so obsolete that maintenance and updates were impossible. In some cases, the software was so inadequate users depended almost exclusively on Excel-based manual processes they had invented as workarounds. As a result, sensitive enterprise data was being stored on individual computers, creating further risk with disaster recovery and regulatory compliance. Staff was pushed to the limits and the situation for the group was devolving rapidly.

The company was desperate to move off of these legacy systems and the situation soon escalated to the C-level and executive boardroom. Soon after, among all enterprise projects inside and outside of the program, tuition subsidy was designated the number one company priority and risk, both by the CFO and CEO.

I had originally started working at the company on another customer-facing software product, the Parent Portal. Because of my high-profile success with that work, when the tuition subsidy became a C-level priority, this new tuition subsidy project was thrust upon me. 

solution

All my work began with the subject matter expert (SME) assigned to collaborate with me, a senior manager from the Tuition Subsidy Group. She had been with the company for 23 years and was largely responsible for architecting both the group's business processes and existing legacy systems many years earlier. She became the lifeline that made it possible for me to succeed with this project. Furthermore, she had participated in other failed attempts to fix the existing software over the years, and was hungry for success.

Longer term, part of my success was founded in educating and training this SME to assume the role of product owner. But more immediately, she helped me form an advisory group of 20 core users. With this internal group, I began some of the most intense and extensive research and seemingly never-ending whiteboarding sessions of my career. I spent three months, from the beginning of September until late November 2015, locked in a conference room absorbing knowledge and insight. The SME worked on the East Coast and would fly to company headquarters in Portland every other week. While she was there, I would spend 10 to 12 hours a day brainstorming, collecting, and reviewing information. In the subsequent weeks when she was off-site, I would hunker down, distill, and document the knowledge I had gleaned into artifacts for review. Wash, rinse, repeat. Wash, rinse, repeat. This went on for at least 12 focused weeks.

By the end of the third month, I had my head wrapped around the full product vision, and it allowed everyone on the team to fall into a more reasonable cadence with development work. At that point, I began finalizing the design specifications for the initial part of the application, and handing it off piece by piece to engineering. Thereafter, I managed and supported engineering work in regular scrums, sprint planning, UX reviews, and retrospectives, all while regularly dedicating the rest of my time to the continuation of product design. 

Challenges

As an enterprise company with no domain expertise in software development, there were many modern, software-development roles and processes that were missing. The larger program had begun and failed several times, and other program managers had attempted to introduce change management in these areas, only to be met with intense opposition. Putting mechanisms for success into place was slow and difficult, requiring a significant cultural shift. 

It was a frantic game, trying to stay ahead of the engineering wave while I designed, ensuring that subsequent pieces of the system were fully specked and waiting for the engineers as they progressed through the first release. From the project's outset, an entire development team had been hired to build the software. But no one had considered the lead time required to research, design, and scope a new software product. It was too late to try a staggered approach to onboarding. So I had to divest another portion of my attention from product design, to ensure meaningful technical gains were being made on the engineering side. Fortunately, development of the new front-end architecture and UX style guidelines allowed for valuable contributions at the outset, but juggling this additional work was difficult.  

Probably the most difficult challenge of all the work, however, was the exquisite complexity of the system and business processes. This difficulty was compounded by a lack of understanding from upper management, who were disconnected from daily work and use of the legacy software. Workarounds like Excel spreadsheets had essentially become a stealth system that masked the severity of issues with existing software, as well as the scope of work in the new software product.

Finally, from the perspective of architecture and integration, the new software product had to unify many disparate pieces. There were two business lines, each with its own core tuition subsidy system and an ancillary, homegrown ERP system. The new product had to offer one application to manage both business lines. Each system had different workflows, even when they were meant to accomplish identical tasks. Four separate systems in the project each used different database technologies, and data was being replicated at regular intervals between databases to support the use of multiple applications. All data needed to be migrated and consolidated to a single database as a part of the project. I had to untangle this web of relationships and distill it into a single software product, including a a plan to seamlessly cut over to the new software without interruption of daily operations. 

Outcome 

A case study by nature, with its problem-solution-benefit format, Iends itself to exaggeration or dramatization for narrative effect. But frankly, I believe this case study falls short of properly painting the size, complexity, or challenges of this work. This was probably the most difficult project of my career, and I have had a more than a few humdingers in my time.  

Ultimately, the intensity and focus bore fruit. By the time I departed in July, 2016, I had fully designed the v1.0 MVP product and delivered all specifications to engineering, just as the company's primary tuition subsidy application was beginning to fail without remediation. 

addendum

Almost a year after delivering this project, I ran into the department lead who was the subject matter expert for this project; she had since officially become the product owner because of the mentoring during the project. She shared the ongoing success of this software. At the time of this project, I managed this Tuition Subsidy software product, while another agency was building or upgrading two other major applications. Of these three applications, only Tuition Subsidy was successfully delivered; the second application was thrown away and the third was still waiting for its first successful release. Furthermore, users continued to be enthusiastic about the Tuition Subsidy v1.0 user experience. The following are unedited testimonials written by end users to communicate the project outcome to executive management: 

Addie Klassen
The new system is easy to use. I like being able to just type in names or even part of names to search for children. The training and rollout were good.  It is saving me time. There have been no real challenges except for a few minor bugs that were fixed already.

Jami Hiner
I like the new system. I feel its user friendly and so much easier to use. I like that you can narrow down searches and uploading authorizations is a breeze.

LeAnna Inglis
I love it! For me I think it is a lot faster, less clicking since it’s all on one page! I love how you can clone and use the same image for siblings. It of course took a second to get used to, but I think it is very user friendly.

Kara Muller
Overall, I do like the new Tuition Subsidy system better than ACCA (legacy system). I think that it is easier to navigate and I really like that it is all on one page. There are less headaches when processing authorizations. 

Alison McCarley
I have thoroughly enjoyed working with the new Tuition Subsidy v1.0 system.  It has drastically reduced my time processing authorizations and made this part of the job much more enjoyable. 

Lisa Cluff
We love it! Even with learning a new system it is so much faster than using ACCA (legacy system). It is much easier to use and more intuitive. We are so glad that they got feedback from coordinators as they designed it.

Kathy Misner
I love the fact that we can use the same image scan for all children in a family.  Everything is on one page, you just scroll down.

Christy Finton
Tuition Subsidy v1.0 was a critical tool needed for coordinators to efficiently update authorizations. The system was created with coordinators in mind and was made extremely user friendly. You can move from student to student, family to family, and center to center with ease.   
 

product owner and advisory group

A senior manager working in the software's business group was assigned to work with me as a subject matter expert. She helped me organize an advisory group of 20 core users that served as the knowledge base for all my work. Over the course of months, I taught the subject matter expert to take on the role of product owner and make increasingly larger contributions to our work. 

research and usability 

My first three months were almost nothing but research and usability work: studying the legacy applications, conducting interviews, leading working sessions, marathon whiteboarding sets. This work integrated directly with the advisory group, and the intense effort set the foundation and pace for the entire project. 

pre-development development

Upper management had pulled the trigger too early and hired an entire development team, which created an initial period of doldrums for engineering. In the first months while I was primarily dedicated to scoping and conceptualizing the software product, I had to divide my time between product design and tasking and managing engineering work to advance the project. The development of UX style guidelines and a new front-end architecture helped prepare us for pending development work on the v1.0 MVP product.

interaction design and specifications

In whiteboarding sessions, I always started with flows and proceeded to sketching initial designs. Then I moved to Balsamiq and quickly created artifacts that could be used in a succession of UX reviews to move towards a final specification. We initially used Axure. But because of tight resources, limited labor and roles, and unrealistic timelines, I opted for a Lean Agile approach. After some initial high-fidelity mockups, completion of the UX style guidelines, and the implementation of the first design patterns, the front-end developer was able to accurately develop interfaces from just sketchy Balsamiq wireframes and my detailed written specifications. 

agile iterations: wash, rinse, repeat.

Once I built a platform of knowledge for the product in my initial months of research and design, and slowly began to ease into change management with a set of standard Agile concepts and processes, the project slowly brook free from its moorings. We then began to fall into the regular rhythm of modern software development lifecycle, ideating on features within sprints and course-correcting through out development cycles, so we landed on the right solutions. 

Ux reviews: small meetings, big picture

Beyond general design and Agile processes, I am a fervent advocate of running UX review meetings early and often. These iterative sessions allow everyone to avoid time-consuming and expensive rework, or even worse, being forced to commit to wrong solutions. Those unfamiliar with this cultural practice, especially some engineering staff, were surprised by its value and the voice it gave them in building the product. Everyone wins when the team is fully engaged.