The professional who redoes the same step of a quote for the third time, the business owner who learns about the delay from the customer, the department manager who had known about it for weeks with nowhere to say it: the same scene from three angles.
Problem finding is the active search for problems before they show up with a cost: noticing them and naming them while they are still small, recurring facts.
In Italy, the law requires business owners to put in place measures suited to detecting a crisis early, but the signals it lists are debts thirty, sixty or ninety days past due [1]: by the time they arrive, the problem has months behind it.
A problem has two dates: the one when it starts producing effects and the one when someone gives it a name. Problem finding does not eliminate problems, it shortens the distance between the two.
This article separates the three moves that everyday language confuses, shows how to read a weak signal, where to anchor the detection and which mistakes make it useless.
The two dates of a problem: when it starts costing and when someone names it
Are a company that receives its problems and one that goes looking for them dealing with the same problem?
No: the first deals with it months late, with a cost already incurred.
The law itself proves it: the signals listed in Italy's Business Crisis and Insolvency Code are debts thirty, sixty or ninety days past due [1] — they arrive when the problem already has a history.
In a company that only discovers problems when they blow up, the problem arrives with two things attached: a cost that has already been paid and the name of a person to pin it on.
Neither is useful information.
This section defines problem finding starting from what happens when it is missing, distinguishes the two dates of a problem — the one when it starts producing effects and the one when it gets named — and makes clear that the move is not about forecasting: it is about what is already happening and that nobody has added up yet.
It closes by setting the boundaries with the terms most often confused with it: anomaly, risk, management control, reporting.
Problem finding means giving a name to a deviation while it is still a minor fact, on your own initiative and not as a reaction.
The obligation set by the Italian Crisis Code concerns the move, not the outcome [1].
As a subject of study, problem finding dates back to the 1970s: in 1979 Getzels noted that the formulation of a problem had received less attention than its treatment [2].
An anomaly is a single fact: it becomes a problem when it repeats.
A risk has not happened yet; a problem that has been found has.
Management control measures what has already been chosen for measurement, and reporting arrives if someone feels like it, whereas detection looks, on a fixed schedule, for what is not in any indicator.
Problems that are already visible have their own entry point, by type: the page on organizational problems in companies.
Between the fact that gets noticed and the correction, however, everyday language names only one move.
Telling problem finding, problem setting and problem solving apart in less than a page
Between the two dates of a problem there is not one move but three, and skipping one moves the cost further down the line instead of removing it.
Are finding a problem and describing it the same job?
They are not, and the distinction is older than management vocabulary: as early as 1910, Dewey separated the felt difficulty from its location and definition, as two distinct steps of the same line of reasoning [3].
In everyday business talk the three terms are used as synonyms, and the cost of this confusion repeats in the same form: people go from "there's a problem with deliveries" to "let's change the management software" without anyone in between writing down what is not working, how much it is worth and who bears its effects.
This section separates the three moves with an operational criterion: what each one produces, what question guides it, what happens when you skip it.
The order cannot be reversed, but it can be shortened: none of the three takes weeks, as long as it has a place to happen.
| Move | What it produces | The question that guides it | What happens when you skip it |
|---|---|---|---|
| Problem finding | a named problem, which was not an object of work before | "what is not going as it should, and since when?" | problems arrive from outside, with a cost already attached |
| Problem setting | a problem with a scope, a measure and someone affected | "where does it start, how much is it worth, who bears its effects?" | the analysis starts from a generic sentence and produces generic causes |
| Problem solving | a choice between alternatives and a check of the outcome | "which path, with what criteria, with what check?" | the visible effect gets fixed and the issue comes back |
The first move happens while work is flowing, at a shift change or when a job is closed, and takes as long as writing one line.
The second happens at the desk and takes about ten minutes: the time it takes to put in writing where the problem starts, how much it is worth and who bears its effects.
The third happens in a meeting, and starts well only if the two previous moves have left something in writing.
The methods for choosing between alternatives are compared in the guide to business problem solving, which also says when each one is a better fit.
The first move, though, starts before all of this: from a fact that almost nobody records.
Reading weak signals: what separates an anomaly that matters from background noise
Does the same anomaly carry the same information when noticed once as when noticed a second time?
No, and the second time is the moment a case becomes data.
The expression "weak signals" comes from a 1975 article by Igor Ansoff in the California Management Review [4].
A company is not short of signals: it produces them continuously.
What it lacks is a criterion for separating the ones that matter from the rest, and without a criterion the choice is made by the volume of whoever is speaking.
This section proposes four reading criteria, applicable to a single fact in a few seconds, and the practical consequence of the third: almost no signal is weak for the person who sees it up close — it is weak only at the point where decisions are made.
That article deals with strategic market surprises: translating it to the internal scope of a fifteen-person company is the editorial team's work, not the author's [4].
There are four criteria, and they apply to one fact at a time.
- It is below the threshold — it has not yet produced a complaint, a dispute or a line in an indicator, which is why it does not show up in month-end reports.
- It has repeated — the second occurrence turns an episode into information, and it is the cheapest criterion of all, because it only requires having noted down the first.
- Someone already knows about it — whoever sees it up close considers it normal, and for that reason the signal does not need to be searched for: it needs to be brought in.
- It has no recipient — nobody is tasked with receiving it, so it stays where it is.
The third criterion is where the difference lies between an informed company and one that finds things out from its customers: a channel in which what someone has already noticed finds a destination is described in the page on business improvement suggestions.
The signals the law recognizes, on the other hand, arrive when all four criteria were passed months earlier [1].
The second occurrence, in short, carries information the first did not have: what remains is to give it a place to land.
The detection checklist: five moments that already exist and the question to ask
Is looking for problems an activity to do when there is spare time, or an appointment that already exists?
The first assumption is the most common and the one that does not hold: when an activity has no set moment and no owner, it stays the thing you do if there is time left over.
Official statistics do not track the search for problems, but they do track its closest relative: in Italy, only 20.8% of companies with innovation activities have a unit or a person responsible for those projects [5].
In a fifteen-person family business that works to order, there is no time for a weekly inspection round, and it is unlikely anyone will do one.
But five moments already exist: the handover between shifts, the closing of a job, a customer phone call, someone's return from an absence, the month-end meeting.
This section adds a single question to each one and a place where the answer ends up.
This is the problem detection checklist: not one more activity, but five questions anchored to moments that happen anyway.
| Moment (already exists) and frequency | Who does it | The question to add | Where the answer ends up |
|---|---|---|---|
| Shift handover — every day | whoever hands over | "what is left half-done, and who picks it up?" | one line in the log, with date and moment |
| Closing a job — for every job | whoever closes it | "what cost more than expected, and in which phase?" | one line under that job's entry |
| Customer phone call — every call | whoever answers | "was this the first time they asked?" | one line under the customer's name |
| Return from an absence — every return | whoever comes back | "what stalled while I was away?" | one line at the point where work stalled |
| Month-end meeting — once a month | whoever leads it | "what appears more than once in these lines?" | the list of problems to set up |
There is only one log, in a place everyone can reach: one line per fact, with the date and the moment it came up.
Without a destination, what comes up remains a conversation, and a conversation cannot be reread at the end of the month.
In many family businesses a channel already exists: the business owner's son or daughter who joined two years ago, to whom team members tell things that never reach the parent.
The kind of issue a line records can be recognized in the page on organizational problems in companies.
Five lines in a log, however, are not yet five problems to work on.
From found problem to framed problem: scope, measure and who bears the effects
Does the same problem get the same attention when described in one sentence as when described with three coordinates?
It meets two different fates: the first gets discussed, the second gets worked on.
In Italian, the same word, problema, refers both to a drawback suffered and to an issue to be tackled — the Treccani Italian dictionary records "problema" as any situation that presents difficulties, obstacles, drawbacks to be faced [6] — and framing is the move that goes from the one to the other.
"Deliveries are going badly" is not a problem: it is a shared impression.
It becomes a problem when it acquires three coordinates: a scope (where the piece of work involved starts and ends), a measure (how many times, how much time, how much it costs) and someone affected (who bears its effects, inside or outside the company).
This section shows the three coordinates on a single case, rewritten before and after, and makes clear that this is the point where the work changes hands: from whoever finds to whoever analyzes.
The same impression, rewritten after three weeks of lines in the log, reads like this: jobs that go through the paint shop leave on average four days late, it has happened nine times in three months, and the ones who bear the effects are the two customers who pick up on a fixed schedule.
The scope tells you where to look, the measure tells you whether it is worth looking, and the person affected tells you whom you will answer to once something has changed.
Read next: how to write a problem statement, where the three coordinates become five elements and one verifiable line.
The log lines that rewrite starts from have a form of their own, described in the problem log.
How large the gap is compared with what was expected is the subject of a method of its own, gap analysis, which is only mentioned here.
From here on the work no longer belongs to finding: causes are searched for with their own methods, from the 5 whys technique to root cause analysis, and the guide to business problem solving compares them.
The move itself costs little: what makes it useless are five habits people pick up without noticing.
The most common mistakes when looking for problems, and how to avoid them
Does a company that finds more problems become more attentive, or just slower?
It depends on what it does with the second problem it finds.
Searching without a destination produces lists, and a forty-item list stalls decisions just as much as having no list at all.
Five mistakes recur among those who start looking for problems instead of receiving them, and four stem from the same misunderstanding: mistaking detection for an inventory.
Searching with no set frequency; counting problems instead of choosing one; asking "is everything okay?"; finding without framing; jumping straight to the cause.
This section describes how each one shows up and the move that prevents it, with particular attention to the family business, where a generic question gets the most polite answer available.
| Mistake | How it shows up | Correction (the move that prevents it) |
|---|---|---|
| Looking for problems when there is spare time | the round happens in quiet weeks, that is, when it is not needed | anchor the questions to moments that happen anyway |
| Counting problems instead of choosing one | a forty-item list and none of them framed | the round of questions is a filter, not an inventory: from the second month on, you choose |
| Asking "is everything okay?" | polite answers and no information | questions about facts that have already happened: what was left half-done, what came back |
| Finding and not framing | the problem stays an impression and dies in the meeting | three coordinates written down before calling anyone in |
| Jumping from discovery to cause | people discuss the why of a problem whose how-much nobody has measured | scope and measure first, then cause analysis |
The first mistake is reflected in official statistics, which do not track the search for problems but do track its closest relative: in Italy, only 18.3% of companies with innovation activities plan those projects with a predefined budget [5].
The second and fifth mistakes have names of their own: choosing which lines deserve to be framed is problem triage, and the why of an issue is looked for after measuring its how-much, with the methods of root cause analysis.
A company that finds more problems becomes more attentive to the extent that it frames at least one of them.
Limitations and conditions of applicability
The two dates of a problem are an interpretation, not a measurement: the date when the effects begin is reconstructed after the fact and approximately, and without a log it cannot be reconstructed at all.
The four criteria for reading signals derive from literature on strategic market surprises [4], translated by the editorial team to the internal scope of a small company: they are an operational proposal, not a research finding.
The ISTAT figure concerns innovation projects in Italian companies and does not measure the search for problems [5]: it supports a stated analogy, not a quantification of the phenomenon.
The case rewritten with the three coordinates — the four days, the nine occurrences, the two customers — is an example built by the editorial team and serves to show the form, not the frequency.
The five moments apply where there is a handover, a job and a recurring meeting: in an organization with other rhythms the appointments change, not the criterion.
FAQ
What does problem finding mean?
It means looking for problems instead of receiving them: giving a name to a deviation while it is still a minor, recurring fact, before it shows up with a cost.
It is the move that shortens the distance between the date when a problem starts producing effects and the date when someone names it.
What is the difference between problem setting and problem solving?
Problem setting gives the problem that has been found a scope, a measure and someone affected, which turns it into an object of work.
Problem solving is the next step: it chooses between alternatives using a criterion and checks the outcome of the choice.
Skipping the first leads to discussing the causes of a problem that nobody has defined yet.
What are weak signals in a company?
They are facts that four criteria make recognizable: they are below the complaint threshold, they have repeated, they are already known to those working close to them, and they have no recipient.
The four criteria help you decide in a few seconds whether a fact should be noted down or not.
Where do you start looking for problems?
From five moments that already exist in any organization: the handover, the closing of a job, a customer phone call, the return from an absence and the month-end meeting.
To each one you add a single question, and every answer ends up as a line in a single log.
Key takeaways
A problem has two dates, and all the work that can be done lies in bringing the second one forward.
Between the two there are three distinct moves — finding, framing, choosing — and skipping one moves the cost further down the line instead of removing it.
A fact deserves to be noted down when it is below the threshold, has repeated, is already known to someone and has no recipient.
Detection does not require extra time: it requires five questions anchored to moments that happen anyway and a log where the answers end up.
What has been found becomes workable when it acquires a scope, a measure and someone affected, and from there it moves on to analysis methods.
There are five mistakes that make the round useless, and four of them stem from the same confusion between a filter and an inventory.
Conclusion
Finding a problem is not the same as tackling it, nor the same as describing it.
A problem has two dates, and the only one you can act on is the second: the moment when someone gives it a name.
The criteria for reading a signal, the questions anchored to moments that already exist and the three scope coordinates serve to bring that date forward by a few weeks, not to make the company immune.
To choose the method for tackling a problem once it has been framed, the guide to business problem solving compares the tools and says when each one is a better fit.
To start from problems that are already visible and recognize them by type, the entry point is the page on organizational problems in companies.
What has not happened yet is a different matter and has its own register: how to build one is in the guide to risk management.
A company that finds its problems before they show up with a cost changes the tone of its meetings: people discuss this week's anomalies instead of last month's shutdowns, and those on the shop floor have a place to put what they notice.
It is not a company without problems: it is a company that meets them while they are still small.
Sources and references
[1] Decreto legislativo 12 gennaio 2019, n. 14, "Codice della crisi d'impresa e dell'insolvenza", art. 3 "Adeguatezza delle misure e degli assetti in funzione della rilevazione tempestiva della crisi d'impresa" (Italian Legislative Decree no. 14 of January 12, 2019) — Gazzetta Ufficiale, Serie Generale no. 38 of February 14, 2019, ordinary supplement; current text. Available at: https://www.gazzettaufficiale.it/atto/serie_generale/caricaArticolo?art.versione=3&art.idGruppo=2&art.flagTipoArticolo=0&art.codiceRedazionale=19G00007&art.idArticolo=3&art.idSottoArticolo=1&art.idSottoArticolo1=10&art.dataPubblicazioneGazzetta=2019-02-14&art.progressivo=0
[2] Getzels, J. W., "Problem Finding: a Theoretical Note", Cognitive Science, vol. 3, no. 2, 1979, pp. 167-172. Available at: https://doi.org/10.1207/s15516709cog0302_4
[3] Dewey, J., How We Think, D. C. Heath & Co., Boston, 1910, ch. VI "The Analysis of a Complete Act of Thought", pp. 68-78. Public domain text: https://www.gutenberg.org/ebooks/37423 — chapter VI in a university archive (Brock University, Mead Project): https://brocku.ca/MeadProject/Dewey/Dewey_1910a/Dewey_1910_f.html
[4] Ansoff, H. I., "Managing Strategic Surprise by Response to Weak Signals", California Management Review, vol. 18, no. 2, 1975, pp. 21-33. Available at: https://doi.org/10.2307/41164635
[5] ISTAT, "Censimento permanente delle imprese 2023: primi risultati", press release, November 14, 2023 (reference year 2022; about 280,000 responding companies, representative of 1,021,618 companies with at least 3 employees). Available at: https://www.istat.it/comunicato-stampa/censimento-permanente-delle-imprese-2023-primi-risultati/ — PDF: https://www.istat.it/it/files/2023/11/REPORTCensimprese.pdf
[6] Treccani, "problema", Vocabolario on line, Istituto della Enciclopedia Italiana. Available at: https://www.treccani.it/vocabolario/problema/