As a product management consultant, I am sometimes hired by a company to be a product manager for a particular feature. Most recently, that was cross-platform holiday messaging for a major retailer. What was new to me, was that I was taking on a feature that I didn't design, but needed to revisit and improve. Here are my biggest takeaways.

Don't assume anything. Designs might not reflect the functionality that was built previously. And what was built might not be the best design. 

Here are a few steps for revamping an existing feature or product:

  1. Bridge the past and current teams*
    1. Get a roster of all key players from last iteration. And then the list of players to be part of this iteration. Include business stakeholders.
  2. Research the current state.
    1. Get this requirements from last iteration. 
    2. Get the screenshots from last iteration.
    3. Get a list of all outstanding bugs from the previous iteration.* (see note below)
      1. This will show you possible improvements.
      2. For example, some of the bugs I encountered came up last holiday but they would require a month of work to fix. It was too late...again.
  3. Confirm the product works as you understand it.
    1. Temporarily throw away the old requirements. Rewrite the requirements as if they are new, at least at a high level. 
    2. Hold a meeting/screenshare with all designers, builders and stakeholders to review screens and outstanding bugs. Jog their memory. Tell them how you think it works and get confirmation. 
    3. Find out what's missing - Ask your stakeholders: what might come up this year that will hurt us? What do you wish we could improve?
  4. Re-examine your strategy.
    1. Find all the data you can from the previous iteration. How did it do? Where is there opportunity for improvement? Was it a failure? Success? Why?
    2. Review all existing user research from the previous iteration. What new insights can you glean when you couple it with research?
    3. Decide if you think that the existing strategy is still relevant and useful.
      1. For example, if existing messaging focuses on shipping, but there is a now a strategic focus on in-store visits does the messaging still work?
  5. Prioritize enhancements. 
    1. Make a list of all new work including stakeholder requests, open issues, open bugs.
    2. Consider new product ideas that arise from new broader strategies, or your own research and brainstorming with the team.
    3. Work with your product team to prioritize based on impact (get numbers for every piece of the feature).
      1. Make sure you reexamine the product mix between iterations. For example, this year we had 5x more 3rd party merchandise on the site than the previous year. Therefore issues we classified as minor for 3rd parties in the last iteration were now major. 
    4. Prioritize list of bugs and issues - must fix, should fix, could fix. 
    5. Get t-shirt sizes (S, M, L) from the team. 
    6. Finalize your scope based on rough ROI (impact / effort).
  6. Communicate broadly. Be Captain Obvious.
    1. Make sure all product owners and engineers know what you're planning to build. Some will assume that because something was broken during launch last time, it's OK that it stays broken. Correct them!
  7. Pay it forward and create a knowledge base for the next product manager.
    1. Screenshot the entire live experience and save it to a nice, labeled zip or deck.
    2. Create a deck with what you would fix in the next iteration (next steps)
    3. Tag all open issues and stories with a label (e.g. "holiday_message_v2")

* If you have a project manager, ask them to help with these items.

If you're at a startup and all of your features are new, this probably isn't relevant to you. But if your job is to enter a large organization and revamp a feature, this advice might help! Please add your own lessons to the comments.

Happy revamping! :D