I’m leading a Blockchain integration project, which is new territory for our team. We have no historical data to go on. How do you create an accurate Project Schedule when you’re dealing with "unknown unknowns" and technologies that your team is still learning?
3 answers
When historical data is missing, I always switch to Three-Point Estimation (PERT). You ask your lead developers for the Optimistic, Pessimistic, and Most Likely durations. The formula $[(O + 4M + P) / 6]$ gives you a weighted average that accounts for the high risk involved in new tech like Blockchain. I also strongly suggest running a "Spike" or a short research phase first. This allows the team to experiment for a fixed amount of time (e.g., 3 days) before they give you a final estimate. It’s much better than just guessing and missing your Milestones later.
Are you building in enough time for the "learning curve"? New tech usually takes 3x longer than the team thinks because of the initial setup and troubleshooting.
Use Analogous Estimating by reaching out to peers in the industry who have done similar integrations. External benchmarks are better than no data at all.
Great advice, Nancy. I often check community forums and GitHub to see average resolution times for similar Blockchain modules before finalizing our internal Baseline Schedule.
Brian is right on the money. I usually add a "Learning Buffer" to the first two sprints of any new tech project. During this time, the Velocity is expected to be low. I communicate this clearly to stakeholders from day one so they don't panic when they don't see massive features in the first week. It’s all about managing expectations while the team builds their core competency in the new domain.