Before Argentina won the 2022 World Cup, Messi had been on the losing side of four major international finals: the 2014 World Cup final and the Copa América finals of 2007, 2015 and 2016. After the 2016 defeat he publicly announced his retirement from international football, then returned.
That is the documented record. What it is not is a record of failure — reaching four finals is itself extraordinary, and most players never reach one. But for years it was discussed as failure, because the only outcome anyone counted was the trophy.
The people question this raises
How does your organisation read a run of near-misses? Someone who repeatedly gets close to a hard outcome and does not land it looks identical, on a dashboard, to someone who cannot do the job at all. Distinguishing between them is a judgment call — and most companies make it by mood, by whoever spoke last, or by whichever quarter the conversation happens to fall in.
The stakes are asymmetric. Persisting with someone who genuinely cannot do the work costs you a year and damages the team's belief in standards. Cutting someone who was one supported attempt away costs you the person, their accumulated context, the cost of replacement, and the year the replacement needs to reach the point the first person had already reached.
Where this shows up at work
- The salesperson who reaches final-round with enterprise accounts and loses three in a row on price.
- The product lead whose last two launches were well built and badly timed.
- The internal candidate who has been runner-up for promotion twice and has stopped applying.
- The team that has fixed the same class of incident three times, each time faster than the last, and is still described as unreliable.
Separate the signal from the outcome
| Signal | Likely 'not yet' | Likely 'not here' |
|---|---|---|
| Pattern of losses | Losing later, in harder contests, to stronger competition | Losing at the same stage, the same way, every time |
| Learning | Can explain what they changed and why | Explains only what happened to them |
| Controllables | Executed their part; external factors decided it | Repeated avoidable errors inside their control |
| Support | Brief, staffing or tooling never changed | Had everything asked for and still stalled |
| Effect on others | Team still trusts and follows them | Colleagues have started routing around them |
A framework for persist-or-part decisions
- 1Is the process improving?Losing later in harder contests is progress. Losing the same way each time is not. Write down what improved between attempt one and attempt three — if you cannot, that is your answer.
- 2What was inside their control?Separate the person's decisions from market, timing, staffing and luck. Hold them to the first, and stop billing them for the rest.
- 3Have we changed anything?If the support, team, brief or resourcing never changed across three attempts, the result is partly ours and a fourth identical attempt is our error, not theirs.
- 4What would we regret?Name the full replacement cost, including the twelve months a new hire needs to reach the same point, and the message the team takes from the decision.
Tell the person plainly: we are backing you for two more attempts, here is what we are changing to support you, and here is what we will look at in March. Unspoken patience feels exactly the same as being quietly written off — and people leave on that feeling.
A worked scenario
An enterprise seller has lost three large deals in nine months. The reflex is a performance plan. Instead the manager works the four questions and finds: losses are moving later in the cycle, two failed on a missing security certification the company had not prioritised, and legal turnaround averaged three weeks. Controllable errors: one, a weak stakeholder map, now corrected. The decision is explicit — two more supported attempts, with legal turnaround cut to five days and a solutions engineer joining discovery, reviewed at the end of the quarter. That is a defensible decision either way. If the next two also fail with the constraints removed, the conversation becomes clear and fair rather than emotional.
What readers often get wrong about this
- This is not permission for indefinite patience. Persistence without changed conditions is not resilience; it is a habit. Put a date on the decision.
- Do not confuse effort with progress. Working harder at an unchanged approach is the clearest signal of 'not here'.
- Resilience is not purely individual. Argentina's 2022 squad, coaching and circumstances differed from 2014's — a person's ability to come back is heavily shaped by the system around them.
- Be careful whose near-misses get patience. Audit who in your organisation historically received a second chance and who did not; the pattern is rarely random.
Interpretation, not fact: attributing the 2022 outcome to resilience alone is a story imposed afterwards. Squad quality, coaching and circumstance all changed too. The reliable lesson is narrower — the difference between "not yet" and "never" is rarely visible in the moment, which is why the decision needs evidence rather than instinct.
The conversation, in five sentences
- Here is what I have seen improve across your last three attempts.
- Here is the part I hold you accountable for, specifically.
- Here is what we got wrong and are changing on our side.
- Here is what we will look at, and on what date.
- Here is what happens if it goes well, and what happens if it does not.
Common questions
How many attempts is fair?
Enough attempts for the changed conditions to be tested at least twice. One attempt after a change proves little; three attempts under identical conditions prove nothing at all.
Does this conflict with a performance improvement plan?
No. It should precede one. The four questions tell you whether a plan is even the right instrument, and if you do open a plan, this evidence makes it specific, fair and defensible.
What if the person has lost confidence rather than capability?
Treat it as a real and separate problem. Shrink the next attempt so a win is achievable, make the support visible, and stop discussing the streak publicly. Confidence recovers through completed work, not through encouragement.
Over to you
Who on your team is one supported attempt away from something good — and what would you have to change to give it to them?
- Copa América and World Cup final records — FIFA
- Messi's 2016 international retirement announcement and return, as reported — BBC Sport
- Why failing well matters — learning from intelligent failure — Amy Edmondson, Harvard Business Review