How To Run A Workshop: Guidelines and Checklist
I received feedback that Chapter 9: “Identifying Big Data Use Cases” from my book Big Data: Understanding How Data Powers Big Business needed more details, especially the part on how to run a workshop. So I created a “checklist” of tasks, plus tips and techniques, to ensure a successful workshop.
One thing that this checklist won’t address is having a workshop leader with strong facilitation skills. While it may sound trivial (heck, can’t just anyone lead one of these workshops?), a workshop leader specifically trained in facilitation skills and techniques is important to workshop success. This person needs to have strong interpersonal skills, and the ability to be both empathetic and demanding at the same time. To quote the movie Ghostbusters, less a scientist and more a game show host.
Pre-workshop Prep Checklist
- Identify the targeted business opportunity
- Identify the business stakeholders to participate in the workshop
- Share interview questionnaire with interviewees prior to the interviews. The interviews should address the following:
- What are the key objectives and responsibilities of their role?
- What decisions are they trying to make?
- What questions do they need to answer to support those decisions?
- What are the metrics or KPIs against which success will be measured?
- What are the organization’s value drivers (e.g., the key activities that help the organization make money)
- Collect sample reports, spreadsheets and/or dashboards
- Collaborate with the “data czar” to secure a small, sample data set (5-6 GB) from which the data scientist will create illustrativeanalytic examples
- Confirm your workshop facilitation team, which should include a lead facilitator, an industry or functional subject matter expert, and a data scientist
Workshop Set Up
- Set up the room for the facilitated workshop:
- Arrange chairs in a horseshoe shape
- Create a “Parking Lot” flip chart and tape it to the wall
- Create a “Ground Rules” flip chart and tape it to the wall
- Create a “Prioritization Matrix” chart and tape it to the wall
- Tape 5 to 6 blank flip chart sheets to the walls for brainstorming
- Post the workshop Ground Rules, including:
- Only one conversation at any given moment
- No hierarchy in the room; everybody and their ideas are equal
- Turn off cell phones, tablets and computers (or at least put them into buzz mode)
- Share any and all ideas (the only bad idea is the one that isn’t shared)
- Breaks are planned throughout the workshop, so please stay with the group as much as possible
Workshop: Introductions & Ground Rules
- Open the workshop with a welcome and introductions
- Explain why the participants are there and the objectives of the workshop.
- Share the roles of the workshop team (facilitator, data scientist, subject matter expert, scribe).
- Have everyone share their name, their responsibilities, and their expectations for the workshop.
- (Optional) Have an icebreaker, like “share with the group something about yourself that you don’t think anyone else knows.”
- Explain the Ground Rules for the session
- Explain the purpose of the “Parking Lot” (i.e., captures topics that are outside the scope of the workshop and keeps the workshop moving in the right direction)
Workshop: Targeted Business Initiative Discussion
- Have the executive sponsor led a discussion about the targeted business opportunity (15-20 minutes). This discussion should cover business objectives, financial goals, business drivers, key performance indicators, critical success factors, and timeframe.
Workshop: Interview Feedback
- Share the business and IT interview key takeaways (15-20 minutes). Confirm, clarify, and gain consensus on the interview findings.
Workshop: Envisioning The Art of the Possible
- Share the envisioning and data science work (30-45 minutes). This part of the workshop is designed to stimulate creative thinking and is an important lead into the brainstorming activity
- Share the illustrative analytics that the data science team created from the client’s data to stimulate creative thinking regarding how advanced analytics could energize their business
- Review examples from other industries of advanced analytics applied to different business scenarios
- Share user experience (UEX) mockups that you created from the sample reports and dashboards in order to stimulate creative thinking (PowerPoint works great as your mockup and user experience development tool)
- Have the data scientist integrate external data sources like social media (twitter, Facebook, LinkedIn), app-generated (Zillow, Eventbrite), and public data (gov) in order to help envision the realm of what’s possible
- Lead the participants through a series of big data brainstorming scenarios (30-45 minutes) including:
- Scenario #1: What would you want to know or do if you could get access to ALL the transactional data (not just a small sample); if, for example, you had all data from the last 10 or 20 years?
- Scenario #2: What would you want to know or do if you could get all internal (consumer comments, work orders, physician notes) and external (social media, mobile, blogs, public) unstructured data sources?
- Scenario #3: What would you want to know or do if you could access and analyze data in real time?
- Scenario #4: How would you leverage predictive analytics? What would you want to know or do if you could predict, forecast, score, or correlate data across all historical, real-time, internal, external, structured, semi-structured and unstructured data?
- Tip: have the workshop participants capture one idea or thought per Post-it® note (sticky note) throughout the four scenarios; have the facilitator place the sticky notes on the flip charts on the wall as the ideas or thoughts as they come up
- Tip: very useful to have the facilitator read aloud the idea or thought as they are posting it to the wall; this helps to fuel the creative thinking process
- Optional: Leverage Michael Porter’s value creation model tool to uncover new ideas and insights
- Value Chain Analysis and/or Five Forces Analysis can uncover additional insights the business stakeholders want out of the data
- Tip: never let workshop participants brainstorm in groups. Participants should brainstorm individually; otherwise good ideas can get lost when there are overpowering personalities in the groups
Workshop: Grouping into Common Themes
- Have the workshop participants group the sticky notes into “common themes” or use cases (30-45 minutes)
- Tip: Have participants stand around the flip charts and move the sticky notes into “common themes” on the flip chart sheets
- Once the sticky notes are grouped into common themes, use a marker to draw a circle around each of the themes and give each theme a short name
Workshop: Prioritizing the Common Themes
- Have the workshop participants prioritize the common themes (or use cases) relative using the Prioritization Matrix (30-45 minutes)
- Use the Prioritization Matrix to drive group consensus on the placement of each use case based upon its business impact vis-à-vis its implementation feasibility
- By the end of the prioritization process, the top two to three use cases that settle in the upper right quadrant are good candidates for a Proof-of-Value lab
- Tip: Ask qualifying questions to understand why one use case is more valuable or less feasible than the other use cases; capture the justifications
Workshop: Wrap Up the Next Steps
- Summary the workshop findings and the next steps (15 minutes)
- Review the prioritized list of potential “Analytics Opportunities.” Verify that everyone buys off on the end result.
- Review Parking Lot items and discuss any potential follow up steps
- Discuss next steps
- The final deliverable should include the following:
- Prioritization Matrix with use cases in the agreed upon matrix locations
- The Post-it (sticky) note content grouped by use cases or common themes
- Interview takeaways
- Data scientist illustrative analytics
- User experience mockups
- Documentation of the Parking Lot items (for potential follow up)
- Data Assessment worksheets that assess 1) the business value, and 2) implementation feasibility of each data source
 Dana Barrett (Sigourney Weaver): “You know, you don’t act like a scientist.”
Dr. Peter Venkman (Bill Murray): “They’re usually pretty stiff.”
Dana Barrett: “You’re more like a game show host.”