What went well

• Our in-team communication was great, which sped up the process a lot.
• The final product is quite close to what we initially envisioned it to be.
• The client is satisfied with the result.
• Our documentation on GitHub and in code was quite decent and helped us a lot in the long run.

What should have been done differently

• Continuous Integration testing, because our solution relies heavily on the demo server.
• We should have pushed for different communication means with the client as they usually took a long time to reply to our inquiries, which hampered our progress.
• We should have pushed more for access to STACC’s Client Dashboard for web store owners, so we could see what the users of our extension will see.

What decisions should have been made at which stages

• Getting access to the real API should have been done in the third iteration instead of the last, because it unveiled some problems that are currently out of our control and take quite some time to investigate by STACC.

Changes

• Instead of creating our own customizable containers for displaying related products, we decided to use WooCommerce’s default related products template. This was done to ensure compatibility with WooCommerce themes.
• Switching from user IDs (which stay the same across multiple sessions) to session IDs (which change with every new session), because users that aren’t logged in should still get personalized recommendations.

Customer's opinion

The project is the foundation, which they'll use to enter a new market. They plan to start using the extension in the coming months.