Passing on what the business learns: people carry practice, documents carry facts

Ask a business how it carries learning from one job to the next and it will usually point to a document: a lessons log, a closure report, a shared folder of notes. Each row records what happened, the impact and an action. It exists, it is complete, and it is rarely read.

The problem is not that the document is badly written. It is that a document carries conclusions, while the next team needs something harder to write down: the reasoning behind the conclusions, the conditions under which they hold, and the points people argued about before settling. A study of post-project reviews by J. S. Busby, published in 1999, observed several review meetings and found one feature that seemed to make the difference to whether learning spread: whether managers of other projects attended. Seeing the discussion carried more than a written summary could.

This article explains what written records can and cannot carry, why learning travels best through people, why some of a business’s most valuable capability lives between particular pairs of people, and how to keep what temporary teams learn after they disband. It is general information for owners and managers who want their business to stop relearning the same lessons.

What a written summary cannot carry

A written lesson has three built-in limits:

• It is written from one point of view. A reader cannot tell whether the finding was obvious to everyone or the surviving side of a long disagreement between two people who still see it differently.
• It records the conclusion, not the conditions. “The coffee machine installation delayed opening” is true and not much use. The useful part is why, and when it applies.
• People do not know what they know. Those who worked on a job often feel the lessons are obvious, so it does not occur to them to pass them on. Some only realise they knew something when they are asked about it directly.

Being able to state a lesson is also not the same as being able to act on it. A business that measures learning by whether people can repeat what the log says is measuring the weaker half of what it needs.

Learning travels with people

If learning travels best through people, a few cheap changes help:

• Invite the next team to the review. The person about to run a similar job should attend the review of the last one, chosen because of what they are about to do, not their seniority. They hear the reasoning and context, not just the headline.
• Let them change something. An attendee who can alter nothing is an expensive audience. Agree beforehand what they may change in their own plan as a result.
• Narrow the written record to what documents do well. Dates, figures, decisions, specifications, supplier details and the points that were disputed, with who held which view.
• Say what the record does not try to carry, so nobody mistakes its silence for completeness.

Three quick tests show where a business stands:

1. The visitor test. At our last three reviews, who attended who was not on the job, and what did they do differently afterwards?
2. The disagreement test. Does our record show what was argued about, or only what was concluded? Most records show no disagreements at all, because a record full of disagreement looks like a badly run job.
3. The point-of-view test. Whose account is this, and who else would describe the same events differently?

Lessons should also change something: a checklist, a template, an estimate, a supplier choice, a standard. A lesson that changes nothing is an archive entry, not capability.

Some capability lives in pairs

Some of a business’s fastest and most valuable judgement is held by no single person. It sits between two people who have worked together for years: a scheduler and a dispatcher who reroute a late order before the customer notices, or a production lead and a quality technician who can settle in one phone call whether a fault comes from the material or the machine. Each knows what the other means by “it sounded different on Tuesday”, and each feels safe saying “it might be us” early, because experience has shown it will not be used against them.

That is not knowledge. It is shared reference points and trust. Two consequences follow:

• Documenting it does not preserve it. A successor who reads every note still has to earn the right to raise a problem early and be believed.
• Reorganising can destroy it without anyone leaving. Move one of the pair to another site or role, and the business still has both people and all their skills, yet a class of decision that took an afternoon now takes weeks. Every headcount report says nothing was lost.

Ties that cross boundaries, such as a supplier’s quality engineer and a customer’s maintenance planner, or a sales lead and a technical specialist in another part of the business, are especially valuable and fragile. They exist because two people once had a reason to talk and kept talking. Remove the reason and nothing formal replaces it.

A simple pair check

Before any reorganisation, ask the people who run important work two questions:

• Who are the three people outside your reporting line you would call first if a decision had to be made today on incomplete information?
• Who would you tell bad news to first?

Pairs that come up repeatedly are load-bearing. For each one affected by a planned change, decide whether the pair stays intact, is stretched but kept alive by a shared meeting or joint sign-off, or is severed. A severed pair should either get a deliberate substitute, such as a standing fortnightly call, or be accepted knowingly as a cost. A small structural reason to keep talking is cheap; rebuilding the trust takes years. The key person dependence article covers risk held by individuals; this is the same risk held between them.

Temporary teams should leave something behind

Project teams are meant to dissolve. People return to their roles, contractors leave, and specialists move on. The risk is that what the team learned leaves with them, often at exactly the moment the business needs it to run, maintain or extend what was built.

Plan the transition from the start, not at closure:

Where outside specialists do important work, pair them with someone from the business who takes part in key decisions, sees the difficult problems being solved and gradually takes ownership. That costs some speed during the job. It avoids a dependence nobody chose.

Closing a team is also a people transition. Staff who were pulled into a project may be unsure what they are going back to, and contractors may leave abruptly. Recognise contributions openly, give people a clear next role, and ask each person what they learned about decisions, customers and handoffs, not just about the technical work. People who see that project work was valued and led somewhere are more willing to join the next one, and their experience can be put to use in mentoring and future planning.

Before accepting the result, check capability, not just technical completion: can the people receiving it operate it safely and reliably, with the knowledge, supplier access and decision rights they need? The what outlasts handover article covers records and obligations after a project ends.

Mentoring: who gets access

Mentoring speeds up judgement by helping someone interpret their experience. But the informal relationships that change careers, where an experienced person chooses to invest time in someone, tend to form between people who already resemble each other: the same background, the same route into the business, the same shift. Nobody intends that. It is simply how unmanaged selection works.

Formal mentoring reaches more people but usually achieves less per relationship, because an assigned obligation is not the same as a chosen investment. Neither is wrong. A small business can use both, perhaps a structured buddy arrangement for every new starter and informal mentoring for a few key roles, as long as the owner notices who is getting the owner’s own attention and who is not. A mix of relationships across functions and outside the business also gives people a wider view than a single mentor can.

A worked example

This is an illustration. A café group opens two or three new sites a year, each fit-out run by a different site supervisor. After each opening, the supervisor fills in a lessons log. The log from the last opening reads: “Coffee machine installation delayed opening by eight days.”

The next fit-out is in a shopping centre. The new supervisor reads the log, notes the lesson and books the coffee machine technician early. Opening is delayed again. The real cause, which nobody wrote down, was that in shopping centres the centre’s own contractor must make the water connection and needs two weeks’ notice from the centre manager. At street-front sites, the group’s own plumber can do it. The two previous supervisors had argued about when to book the connection, and the one who was right had moved to another role.

The operations manager changes three things:

• The next supervisor attends the review. Before each fit-out starts, the incoming supervisor sits in on the review of the previous one and may adjust their own plan as a result.
• The log records conditions and disagreements. Entries now say when a lesson applies (“shopping centre sites only”) and note where people disagreed and why.
• A working pair is kept alive. The supervisor who understood shopping-centre fit-outs moves into an operations role, but attends the first site meeting of each new centre fit-out. A one-hour meeting keeps a valuable working relationship with the group’s shopfitter alive.

At handover, the shopfitter walks the store manager through the point-of-sale and network cabling, and leaves a marked-up plan in the store, so the store team can deal with routine problems without calling the shopfitter. The next shopping-centre opening is on time, and the checklist now includes “confirm centre water connection process and notice period at lease signing”.

How this applies to a small Australian business

• Invite the next person to run similar work to the review of the last job.
• Record conditions and disagreements, not just conclusions.
• Turn lessons into changes in checklists, templates or estimates.
• Identify working pairs before reorganising, and keep important ones talking.
• Plan knowledge transfer from the start of projects using outside specialists.
• Check the receiving team’s capability before accepting a handover.
• Notice who gets mentoring attention, and widen it deliberately.
• Encourage relationships across functions and outside the business.

Signals worth watching

• The same problems recurring on similar jobs.
• Lessons logs that nobody reads before starting work.
• Records with no disagreements in them.
• Decisions slowing after a reorganisation in which nobody left.
• Suppliers holding knowledge the business needs to run what they built.
• Development attention going to the same kinds of people.

Common mistakes

• Treating a lessons log as the whole of learning.
• Recording conclusions without conditions.
• Reorganising without noticing working relationships.
• Waiting until project closure to think about knowledge transfer.
• Assuming documentation can replace experience.
• Leaving mentoring entirely to chance.

Frequently asked questions

Should we stop keeping a lessons log? No. Keep it, but narrow it to facts, decisions, conditions and disagreements, and pair it with attendance by the people who need it.

How do we get people to share what they think is obvious? Ask specific questions in the review, such as “what would you do differently on a site like the next one?”. People often find out what they know only when asked directly.

Can small businesses really protect working pairs? Yes, usually cheaply. A standing meeting, a shared review or a joint sign-off keeps two people with a reason to talk.

Is formal mentoring worth it? It is useful for making sure everyone gets some access, especially new starters. Expect less from each relationship than from the best informal ones.

What if a key supplier holds all the knowledge? Agree knowledge transfer in the contract, pair a staff member with the supplier during the work, and check capability before handover.

Questions to ask

• Who will run similar work next, and will they attend this review?
• Does our record show when a lesson applies, and what was disputed?
• Which working relationships would a planned change break?
• What knowledge does a supplier hold that we will need?
• Can the people receiving this result actually run it?
• Who in the business is getting development attention, and who is not?

Bringing it together

Documents carry facts and conclusions; people carry practice, reasoning and context. Invite the next team to hear how the last one reasoned, and narrow written records to what they carry well, including conditions and disagreements. Recognise that some of the business’s best judgement lives between particular people, and keep important pairs talking through change. Plan what temporary teams and outside specialists will leave behind, and check that the people receiving the result can actually use it. Learning that does not change what someone does next is not yet learning.

---

Source: KEVOS notes, drawing on J. S. Busby (1999) on post-project reviews in the Project Management Journal, teaching material on project closure, knowledge transfer, mentoring and networks, D. Day’s review of leadership development, and M. Granovetter and R. Burt on weak ties and network brokerage. Examples and figures in this article are illustrations. This article is general information.