What Your Daily Coffee Is Actually Solving

What each one is actually for

Collaborative Problem Solving

Collaborative Problem Solving — usually shortened to CPS, and also taught as Collaborative and Proactive Solutions — is Ross Greene's method for handling a child who keeps doing something difficult. Its founding sentence is "kids do well if they can," which is a deliberately provocative reversal of the ordinary assumption that they do well if they want to. On Greene's reading, an outburst is not a motivation problem to be met with more consequence; it is what happens when a specific demand exceeds a specific skill the child does not yet have. The method's core procedure, which Greene calls Plan B, has three steps in a fixed order. First, empathy: you ask about the difficulty and then you keep asking, without rebutting, until you actually know what makes it hard for them — not what you assumed. Second, you state your own concern plainly, once. Third, you invite them to generate solutions with you and you take theirs seriously. The order is load-bearing. Skipping the first step and going straight to solutions is the most common way it fails.

The Latte Factor

The Latte Factor is a phrase from the personal-finance writer David Bach. The idea is that a small recurring discretionary expense you have stopped noticing — the daily coffee is the stock example, but it is really about any autopilot spend — adds up to a substantial amount over years, and that redirecting it into savings or investments compounds into a number most people find surprising. What the method actually asks you to do is narrow: go through the last month of transactions, find the repeated small ones you did not consciously decide on, total them, and see the annual figure. That is the whole mechanic. Its persuasive force comes from making an invisible number visible — you were spending it either way, and the only thing that changed is that now you can see it. The financial claims that get stacked on top of that observation are contested and I will come back to them. The observation itself — that autopilot spending is invisible until you deliberately count it — is not really in dispute.

What you get out of pairing them

What you get is a way to change a stubborn habit that does not depend on being disgusted with yourself. The standard path with an autopilot spend is: notice it, feel bad, resolve to stop, stop for eleven days, resume, feel worse. That cycle is not a character flaw; it is what happens when you remove something without replacing what it was doing. Greene's approach gives you the missing step, and it is a step almost nobody applies to their own behaviour because we reserve that kind of curiosity for other people's children. The specific benefit is that you end up with a substitution rather than a subtraction — which is both more likely to hold and, in most cases, more pleasant. The second thing you get is more interesting. Once you have run the enquiry on one spend, you have a reusable instrument, and it works on things that are not money at all: the phone at 11pm, the third drink, the meeting you keep not cancelling. The move is portable in a way that the coffee arithmetic is not. And there is a quiet third benefit for anyone who is also a parent, which is that practising the empathy step on yourself, where nobody is watching and the stakes are two pounds eighty, makes it considerably easier to run it on a nine-year-old at the point where you are angry.

What Collaborative Problem Solving rests on

• Externalizing makes pattern visible randomised trials
  Putting an internal pattern into an external, structured form makes it visible and available for examination instead of remaining implicit and merely enacted.
• Parts have positive intent clinical practice
  A behavior that looks like misbehavior is usually a mistaken strategy aimed at meeting a legitimate underlying need rather than an expression of badness itself.
• Collaborative solution generation increases buy in observational studies
  Jointly generating a solution with the person who has the problem produces more durable follow-through than an imposed solution.

What The Latte Factor rests on

• Making a cost visible changes spending behavior observational studies
  Identifying and naming a previously unnoticed recurring expense makes it available for conscious decision-making instead of running on autopilot.
• Small changes feel more achievable than large ones practitioner report
  A small recurring change such as redirecting one discretionary purchase feels more achievable to start than a large lump-sum saving goal.
• Small recurring costs compound over time plausible mechanism
  Small recurring expenditures when redirected and compounded over time accumulate into a financially significant sum.

Where they actually meet

Strip the subject matter off both and the same two-part manoeuvre is sitting there. Part one: take something that was running without being looked at and put it into a form you can look at. Greene does this by writing the difficulty down as a specific, observable statement — not "he's defiant" but "he has difficulty starting homework when he's been on the tablet" — because the vague version cannot be worked on and the specific version can. Bach does this by pulling the repeated transactions out of a bank statement and writing a total next to them. In both cases nothing has changed about the world; something has changed about what is available to be examined. Part two, and this is the part the finance version usually skips: assume the behaviour was doing a job. Greene assumes it as a matter of method — the outburst is a strategy, a bad one, aimed at something legitimate. The Latte Factor, as normally taught, does not assume this; it treats the spend as pure leakage, which is precisely why the cut so often fails to survive a hard week. Where they meet is the first part. The second part is the thing one has and the other lacks, and it happens to be the one that determines whether the change holds.

The principle underneath both

A repeated behaviour that looks like a leak is usually meeting a need, and until you know which need, removing the behaviour just relocates the cost. That is the pattern, and it is not owned by parenting or by personal finance — it is closer to a general observation about behaviour that both fields arrived at from different directions, and that one of them then articulated much more carefully than the other. The reason it matters practically is a matter of sequence. If you delete the behaviour first, you get an unmet need with no strategy attached to it, and unmet needs do not sit quietly; they recruit a new strategy, often a worse one, and often one you will not connect back to the original decision. The child who is punished out of the outburst frequently does something less visible instead. The person who cuts the coffee frequently orders something on their phone at 10pm. On this reading, the sequence has to be: find the job, then find a cheaper thing that does the job, then let the original fall away — rather than remove, resist, and hope.

What that makes the real goal

Said without either lineage attached: to replace a recurring behaviour with one that meets the same underlying need at a lower cost, arrived at with the cooperation of whoever is doing the behaviour. That last clause carries most of the weight, and it is the part that survives the translation from parenting to money. When the person doing the behaviour is your child, cooperation means an actual conversation in which their proposal is genuinely on the table. When the person doing the behaviour is you, cooperation means something less obvious but quite real — that the part of you that wants the four o'clock walk gets a hearing rather than being overruled by the part of you that read a book about compound interest. Stated this way the goal detaches from both traditions, and you can then ask a much better question than either book asks, which is: what else, from anywhere, reliably produces a lower-cost substitute for a need rather than a suppression of it? Several fields have worked hard on exactly that.

Other practices that serve the same goal

Three others are worth having, and two of them carry considerably more evidence than either practice on this page. The first is functional analysis, from behavioural psychology: before changing a behaviour, you identify what reliably comes before it and what it reliably produces, on the reasoning that the consequence is what maintains it. This is the formal, tested ancestor of what CPS does conversationally, it underpins the functional behaviour assessments used in schools, and it has decades of applied research behind it. The second is habit reversal training, developed by Azrin and Nunn for tics and repetitive behaviours, which pairs awareness training with a deliberately chosen competing response — a specific alternative action, decided in advance, that is incompatible with the old one. It has randomised trial support for tic disorders and trichotillomania, and its central insight is exactly the one the Latte Factor lacks: you do not remove, you substitute, and you decide the substitute before you need it. The third comes from inside personal finance itself. Ramit Sethi's conscious spending approach inverts the standard advice — spend extravagantly on the two or three things you genuinely love, cut hard on everything else — which is the same move as finding the job and paying for it properly rather than paying a little for everything. It has no trial evidence, but it is a lineage that already reached this conclusion without any help from clinical psychology.

A tradition that already does this

The clearest precedent is not in finance and not in parenting: it is motivational interviewing, developed by William Miller and Stephen Rollnick for people with alcohol problems in the early 1980s. Its founding move was to stop confronting the behaviour and start eliciting the person's own reasons and their own account of what the drinking was doing for them, on the finding that direct confrontation predicted worse outcomes rather than better ones. That is the empathy step, arrived at independently, in a field where the behaviour in question was far more costly than a coffee, and it has since accumulated a substantial randomised-trial literature across substance use, medication adherence and health behaviour change. It is worth knowing about because it settles a question people reasonably raise about the approach on this page — whether asking what a behaviour is for is just a soft way of excusing it. In the field where that question had the highest stakes and was studied most carefully, the answer came back the other way.

Running the empathy step on your own spending before you cut anything

Do this once, this week, on paper, and it takes about fifteen minutes. Pull the last thirty days of transactions and pick out the repeats you did not consciously decide on — not the rent, not the one-off flight, the ones that happen because it is Tuesday. Pick the single one you would most like to stop. Now write the specific version of what is happening, in the style Greene uses with children: not "I waste money on coffee," which cannot be worked on, but something like "I buy a coffee at around four on weekdays when I've been at the desk since one." Notice how much more that gives you. Now run the empathy step on yourself, and run it properly, which means asking and then asking again rather than answering with the first plausible thing. What is happening right before? What would be different if you skipped it — not financially, but in the next hour? What is the worst version of the afternoon where you do not go? The answers are frequently not about coffee. Common ones: it is the only time you leave the building; it is a ten-minute gap where nobody can ask you anything; someone behind a counter is briefly warm to you and that is the fourth-most human interaction of your day; it marks the boundary between morning work and afternoon work. Write down what you find, in one sentence, in your own words. Then, and only then, state the other concern — the money, plainly, once, with the actual annual figure. Then generate options against both: something that keeps the job and lowers the cost. If it turns out the job was leaving the building, the answer is a walk, and you may find you stop wanting the coffee. If the job was the warm human contact, a walk will not do it and you should probably keep buying the coffee and cut something else instead, which is a legitimate outcome of this process rather than a failure of it. The common failure here is answering the empathy question with the answer you think you should give — "I'm just being lazy" — which is not a need and closes the enquiry.

Using the visible-total trick to pick which problem to solve with your child

Send it back the other direction and the finance method contributes something CPS practitioners genuinely struggle with, which is triage. Greene's approach is unusually demanding: each conversation takes real time, has to happen when nobody is upset, and cannot be rushed. Parents who take it seriously commonly try to run it on everything at once and abandon it inside a fortnight. What the Latte Factor supplies is the counting instinct — the willingness to write down every small recurring instance and add it up before deciding what to act on, on the reasoning that frequency multiplied by time is where the actual cost lives, and that intuition is bad at frequency. So do the audit version. For one week, keep a running tally of every friction point, one line each, no judgement — bedtime, shoes, the tablet handover, homework start, the bit before swimming. At the end of the week, count. What you will typically find is that the incidents you remember as worst are not the ones that happened most, and that one or two situations are generating a disproportionate share of the total. That is your first Plan B conversation, and now it is chosen by a count rather than by whichever one you are still annoyed about. This also protects the method from its own most common misuse, which is running the empathy step in the middle of a meltdown; a tally gives you a reason to wait, because you are collecting rather than failing to act. And the compounding argument holds here more honestly than it does with money: an unsolved problem that recurs four times a week is being rehearsed roughly two hundred times a year, by both of you, and solving that one is worth more than solving three rare ones.

Thursday, 4:05pm, standing up before you have decided to

You are already on your feet. You did not decide to be — you registered the time somewhere around ten past three and the rest happened without you. This is the fourteenth of these this month; you counted on Sunday, which is why you know, and the annual number you wrote next to it was large enough that you have been avoiding the note ever since. So today you sit back down and give it two minutes with a pen, because you have decided that cutting it outright has failed twice and you would like to know why. What happens if you don't go? You write: the afternoon has no seam in it. What is the actual moment you like? You write: the walk back, holding it, not drinking it yet. Third question, the one that gets the real answer: what would have to be true for you not to want it? You write: someone would have to have spoken to me since one o'clock. Then you sit with that for a second, because it is not what you expected to write and it is not a spending problem. What you do with it is the ordinary, unheroic thing. You keep Tuesday and Thursday, because the walk is real and you are not going to pretend otherwise. On the other three days you put a fifteen-minute walk in the calendar at four, and you message a colleague on the way out about nothing in particular. The saving is about forty per cent of what an aggressive cut would have produced on paper, and unlike the aggressive cut it is still running in March. Two weeks later you notice something you were not looking for: on the walk days you come back and start the afternoon block immediately, whereas on the coffee days you used to come back and read email for twenty minutes. That was never in the spreadsheet either.

When this is worth reaching for

Three situations call for this, and any one of them is enough. You've already tried cutting a small recurring spend, more than once, and it keeps coming back — a repeated failed cut is strong evidence the spend is doing a job worth naming. Or you're a parent also trying to change something in your own behaviour, where running the curiosity step on yourself at low stakes is real rehearsal for running it on a child at higher stakes — the transfer goes more easily in that direction than the reverse. Or the spend is small but oddly loaded: you get defensive about it, or it's caused an actual argument with a partner, and defensiveness reliably means the thing is doing something unnamed. It's also worth an hour if you've simply never once looked at your recurring small transactions — the counting step is cheap, and the total is usually not what people expect. None of this is for a household that can't cover its bills. There, the coffee was never the real problem, and the next section explains why.

Where this does not apply

The main limit is about what the Latte Factor can and cannot reach, and it is a serious one. If your budget genuinely does not close, the shortfall is very unlikely to live in small discretionary purchases — it is far more often housing, transport, childcare, debt servicing or income, and those are large, structural and mostly not solved by attention. Spending your effort auditing coffee while a rent-to-income ratio is the actual issue is not neutral; it costs you the weeks you could have spent on the thing that would move. Bach's framing has drawn sustained criticism on exactly this ground, and the criticism seems to me largely fair: the arithmetic tends to assume a long uninterrupted investment horizon and a rate of return that is not guaranteed, and it can quietly imply that a structural problem is a discipline problem. Take the noticing move and leave the moralising. The second limit is on the parenting side and is more consequential. Adapting the empathy step for use on yourself is fine; adapting it for use on a child in the middle of a meltdown is not, because the method explicitly requires that nobody is dysregulated when the conversation happens. And where a child's difficulty involves safety — self-harm, violence, anything that frightens you — a collaborative conversation is not the whole answer and should sit alongside a clinician rather than instead of one. Finally, if the honest answer to "what is this spend doing for me" turns out to be that it is the only pleasant thing in a fifty-hour week, treat that as information about the week rather than as a budgeting result.

What the evidence does and does not support

These two are not evidentially comparable and it would be misleading to present them as a pair of equals. The clinical side has real trial data: Greene's approach has been tested in randomised comparisons for children with oppositional behaviour — most notably against standard parent-management training, where it performed comparably — and in inpatient and school settings where restraint and seclusion use dropped after implementation, though several of those are pre-post studies without controls rather than randomised designs. That is a genuine evidence base with known limits. The Latte Factor has no such literature. It is a popular-finance heuristic; the part of it that is well supported is the ordinary and uncontroversial observation that people underestimate the annual total of small repeated purchases until they count them, and the part that is contested is the investment arithmetic layered on top, which depends on assumed returns and an uninterrupted horizon that many people do not have. Nobody has studied applying a clinical conversational method to one's own spending, and it is not obvious how you would: the outcome would have to be measured over a year or more, with a control group that made the same cut without the enquiry, and the interesting effect would be persistence rather than size. So treat the transfer here as an argument from a shared mechanism, made openly as an argument. What you can do is run it as a single-person experiment where the readout is unusually clean, since your bank statement in ninety days will tell you plainly whether the substitution held.