Good questions.
Useful work.

I start with the task people need to perform. I check whether they need practice, clearer expectations, or better tools before deciding what to build.

Leaders needed to understand what employee feedback could tell them. The approach connects survey design to data-quality rules and reporting, with clear limits on how small groups are represented.

The tools and policies around a task affect whether people can use what they know. I use research and UX methods to understand those conditions before recommending support.

Training is one possible response, so I start with discovery. What’s the business problem? What are people doing now, and what should they be doing? I investigate whether the gap involves skills, a process, tools, or incentives. I talk with the requester, managers, and people doing the work, and review the data already available.

At Nationwide, I built the team’s first intake process for e-learning requests so we asked those questions before development began. The response could be a course, a job aid, or a process change. I want us to agree on what would help before we commit to a build.

At Nationwide, my team had more requests than we could handle. I built a three-tier build menu and an hours estimation guide. We sized requests so stakeholders could compare the effort involved in a high-fidelity build with a lighter approach.

We considered business impact, deadlines, and how many people the work would reach. When two leaders both viewed their requests as urgent, I brought them the estimates and the implications for delivery. That gave them a concrete tradeoff to resolve together.

At Nationwide Academy, evaluation was focused on learner reactions. I made the case for Kirkpatrick’s New World Model, standardized our Level 1 and Level 3 surveys on a five-point scale, and rebuilt evaluation for more than 100 courses.

I presented the approach to the enterprise Nationwide Learning Leaders Council, which adopted the standard across the enterprise. The decision expanded the evaluation approach to include questions about behavior after training. My synthetic-persona example shows another way I worked with leaders to clarify what the evidence could support.

Completions show participation. Satisfaction surveys capture the learning experience. To assess application, I investigate behavior on the job through follow-up surveys with learners and managers, existing performance data, and conversations about what helped or got in the way.

At Allstate Claims, I created the Levels 1–3 measurement strategy, and the team revised and adopted its approach across Levels 1–4. I conducted a Level 3 study of workplace application. The mixed-method study included learner and leader interviews, surveys, and stratified sampling. We examined the frequency of job tasks after onboarding, and I worked to make performance objectives measurable so application could be assessed in the field.

When learning isn’t being applied, I investigate manager support, tools, opportunities to practice, and the learning itself. I keep the learning experience and workplace transfer distinct so we can decide where support needs to change.

At Nationwide, e-learning builds took six to eight weeks. We were dealing with late stakeholder feedback and rework. I moved the team to a SAM-based agile process so stakeholders could review rough versions early. I also introduced rubrics describing what good work looked like and a tiered build menu to make scope clear at the start. Development time dropped to about three weeks.

During the SkillPort-to-Meridian migration, I moved us to a full QA pass on every course so we could identify what was breaking. I led a team of about 10, and we fixed roughly 100 broken courses in about two weeks.