The figure here is reproduced with permission from a thesis dissertation by Catherine Barney, entitled "Cross-project learning in project-based organizations"
Catherine was studying knowledge management and lesson-learning in a major European engineering company. As part of her dissertation, Catherine surveyed the company to measure employees' satisfaction with various steps (or "aspects") in the lessons learned cycle (what I have referred to as the knowledge supply chain).
Her lower diagram (opposite) is interesting - there's a reasonable satisfaction with lessons identification and capture, but with every step after that, dissatisfaction grows. This could either be because this company (like many others) thinks the job is done once the lesson is "captured", or because inefficiencies along the chain combine to make each step progressively less satisfactory (in other words, poor verification on top of poor capture leads to even less satisfaction with application).
For me the lesson is clear.
There's no point in capturing lessons, unless you look at the whole supply chain, from supplier to user.