DEV Community

Fathy Shalaby
Fathy Shalaby

Posted on Originally published at specky.space

Problem Space vs Solution Space: How PMs Escape the Solution Trap

Most teams jump from a request to a feature in a day. Here is how to separate the problem space from the solution space, how long to stay there, and a brief you can copy.

Problem space vs solution space is the split between understanding a customer's need and designing an answer to it. The problem space covers who struggles, with what, how often, what it costs them and what they do today. The solution space covers everything you could build in response. The solution trap is jumping to the second before you have finished the first.

Key takeaways

  • Leaders admit the habit. In a survey of 106 C-suite executives from 91 companies in 17 countries, 85% agreed their organisations were bad at problem diagnosis and 87% said the flaw carries significant costs (HBR, 2017).
  • Product people want more problem time than they get. Pragmatic Institute's survey of 2,474 product professionals found they spent 27% of their time on strategic work but believed it should be 53%, and most spent less than 5 hours a month gathering customer feedback (Pragmatic Institute, 2019).
  • Thin problem work shows up as unused features: 80% of features in the average product are rarely or never used (Pendo, 2019).
  • The "Einstein spent 55 minutes on the problem" line has no known source, and the 80/20 ratio is a target, not a measured fact.
  • One question does most of the work: "What is the evidence that this is the problem, and not a symptom of something else?"

What is the difference between problem space and solution space?

Dan Olsen made the split popular with product teams in The Lean Product Playbook (2015). His product-market fit pyramid has five layers. The bottom two, target customer and underserved needs, are the market. The top three, value proposition, feature set and user experience, are the product (Mind the Product, 2017). In a 2018 talk he put the PM's role plainly: "our job as product managers is to focus on the problem space" (Mind the Product, 2018).

Teresa Torres draws the same line in her opportunity solution tree. Under a desired outcome sits the opportunity space, where each opportunity is "an unmet customer need, pain point, or desire", then the solution space, then assumption tests (Product Talk, 2023).

Problem space Solution space
Core question Who struggles, with what, how often, at what cost? What could we build or change?
Typical artefacts Interview notes, support-ticket patterns, problem statements Sketches, prototypes, specs, tickets
What you measure How important the need is, how often it occurs, what the workaround costs Satisfaction, adoption, task success
What customers can tell you A lot: what they did, what it cost, what they tried Less: they can react to options, but rarely design the right one
Typical failure Research that never ends A well-built feature for the wrong problem

What is the solution trap?

A customer complaint comes in, or a sales rep flags a lost deal. Within 24 hours someone has sketched a solution. By the time the PM is in the room, the question has moved from "is this a real problem?" to "how do we build it?" The problem has had 90 seconds of thought. The solution will get six weeks of engineering.

Thomas Wedell-Wedellsborg found the same bias: managers are good at solving problems but switch into solution mode quickly, before checking they understand them (HBR, 2017). Fewer than 1 in 10 executives he surveyed said they were unaffected.

Why do smart teams fall into it?

The trap is a matter of incentives and workflow, not intelligence.

  • Solutions are concrete; problems are ambiguous. "Add a dashboard" can be scoped, estimated and demoed. A problem has to be defined first.
  • Solutions look like progress. Interviews and synthesis look like nothing is happening. Shipping the wrong feature looks like progress until the metric fails to move.
  • Every system rewards solutions. Jira tracks features, roadmaps list deliverables, reviews celebrate launches. The PM who concluded a problem wasn't worth solving has nothing to demo.

The numbers back this up. Product people in Pragmatic Institute's 2019 survey spent 27% of their week on strategic work against the 53% they thought it deserved. The vast majority spent less than 5 hours a month gathering feedback from customers and evaluators (Pragmatic Institute, 2019).

What does the solution trap cost?

The visible cost is wasted engineering. Across 615 product subscriptions, Pendo found 80% of features rarely or never used and 12% driving 80% of daily usage, and estimated public cloud companies had invested up to $29.5 billion in the unused ones (Pendo, 2019). We unpack that number in why 80% of features go unused.

The hidden cost is the team's picture of the customer. A team that lives in solution mode stops mapping the problem space, and the product becomes a patchwork of answers to half-understood problems.

Did Einstein really say he'd spend 55 minutes defining the problem?

There is no evidence he did. Quote Investigator found no substantive evidence that Einstein ever said it (Quote Investigator, 2014). The earliest Einstein attribution the etymologist Barry Popik found is a 1995 newspaper article, 40 years after Einstein died in 1955, and it gives no source (Barry Popik). Versions differ on the split: 55 and 5 minutes, 59 and 1, even 40 and 20.

Even Dwayne Spradlin's well-known 2012 HBR article opens with a 59-minute version (HBR, 2012). His argument holds without it: how rigorously you define a problem is the biggest factor in finding a good solution. If you ask stakeholders for evidence, don't open with an unsourced quote; use the survey numbers above.

How much time should a product manager spend on the problem?

No study has measured the ideal ratio, so treat any number, including the 80/20 in this post's original title, as a provocation. Most teams run closer to 20/80. The best invert it, or at least reach 50/50.

A more useful rule: scale problem-space time to the size of the bet and how hard it is to undo. This table is a starting point, not a benchmark.

Size of the bet Problem-space time before building Minimum evidence
Under a day of work, easy to reverse 30-60 minutes One clear ticket pattern or a direct user quote
1-2 weeks of work 2-3 days 3-5 interviews or observed sessions, plus a count of tickets or requests
A quarter or more, or hard to reverse 2-4 weeks A written problem statement, 8+ conversations, a baseline metric and the workaround's cost

For the last row, see one-way and two-way door decisions; reversibility often guides effort better than size.

How do you stay in the problem space longer? Five practices

1. Frame the problem before backlog grooming

Before any feature enters refinement, answer three questions. What job is the user trying to do? What evidence shows the friction exists? Is this the constraint or a symptom? If you can't answer all three in two minutes, the problem isn't ready for a solution.

2. Ask for evidence, not articulation

Describing a problem persuasively is not evidence. Evidence is four interviews where you watched users hit the friction, a ticket pattern with a count, or an experiment. Articulation gets a project started; evidence decides what to build. Our guide to evidence for the four product risks shows what counts for each kind of risk.

3. Separate the request from the need

The request is usually a feature; the need is usually a job to be done. The PM who hears "can we add a CSV export?" and evaluates CSV export is already in solution mode. The PM who asks "why do you move this data, where does it go, what happens if you can't?" is still in the problem space. Stay there longer than feels comfortable. Jobs to be done gives you the vocabulary, and the free JTBD interview guide gives you the questions.

4. Write an explicit problem statement

A problem statement names no feature. This is one: "Customers who have connected more than three integrations can't quickly see which data sources fed a given insight, so they don't trust the output and go back to the raw data, costing them 30 minutes each time." "We need better insight attribution" is not; it's a solution-shaped description of a problem. Our problem statement template walks through the format.

5. Ask: is this a constraint or a symptom?

Some problems are the effect of something upstream. Fix a symptom and users get a workaround that hides the real issue, while the cause stays. One question exposes the hierarchy: "If we solved this, what would still be broken?"

Problem brief template (copy and paste)

Fill this in before anyone names a solution. An empty line is your next piece of research.

PROBLEM BRIEF
Who has it:            [segment; number of accounts or users affected]
Job they're doing:     [in the customer's words]
Where it breaks:       [step in the workflow; how often per week or month]
What it costs them:    [minutes, money or risk, with the source]
What they do today:    [workaround, competitor, or nothing]
Evidence:              [n interviews, n tickets, metric; link each one]
Constraint or symptom: [if we solved this, what would still be broken?]
Biggest unknown:       [the riskiest open question]
Solutions considered:  [at least three, written only after the lines above]
Enter fullscreen mode Exit fullscreen mode

The last line follows Torres's advice to explore three solutions, not one (Product Talk, 2023).

Worked example: the CSV export request (illustrative)

Say a 6-person B2B SaaS team gets four requests for CSV export in a month. Engineering estimates 3 weeks.

The PM spends 3 days in the problem space instead. She reads the 22 support tickets that mention exporting and interviews 5 requesters. Four of them export every Monday to paste the same numbers into a weekly report for their manager. The fifth needs a backup for an annual audit.

For 4 of 5, the need is "show my boss the weekly numbers", not "get my data out". The team compares three options: the full CSV export (3 weeks), a shareable dashboard link (1 week) and a scheduled weekly summary email (4 days). They ship the email first and add a plain raw export for the audit case later.

Three days of problem work replaced roughly two weeks of build, and the shipped feature matched how people used the data. Real trades are rarely this clean, but they rarely run the other way.

Doesn't problem work slow down shipping?

No. The point is to front-load the work so what you ship is more likely to matter. Teams that invest in the problem are usually faster once they start building: the wrong solutions are ruled out before anyone writes code, and the build doesn't reverse halfway through when someone asks, "wait, is this what users need?" Teams that skip it ship faster at first, then spend months asking why the metric didn't move. Skipping problem understanding is not speed. It is debt, and the habit that pays it down is steady customer contact, covered in continuous discovery.

When does staying in the problem space go wrong?

The opposite failure is real too.

  • Research as avoidance. If no evidence could change the decision, more interviews only delay it.
  • Over-studying cheap changes. A one-hour copy change doesn't need a week of interviews. Ship it and watch.
  • Refusing to prototype. People react to something concrete. Use a rough prototype as a probe, not a commitment.
  • Solution-shaped problem statements. "Users need a dashboard" smuggles in the answer. Rewrite until no feature is named.
  • Treating volume as severity. Ten requests from people who would never churn can matter less than one blocker for your best-fit customer.

Where Specky fits

The slowest part of problem-space work is collecting the evidence behind a request. Specky reads customer calls, feedback, Slack, Jira and analytics into a Product Graph, so the evidence for a problem is linked rather than scattered. The Opportunity Tree keeps problems and solutions on separate levels, and specs drafted in the PRD Editor carry inline citations back to the customer evidence. If you want the full method first, start with our product discovery process guide.

Frequently asked questions

What is the difference between problem space and solution space?

The problem space is the customer's world: who has a need, how often, at what cost and how they cope today. The solution space is what you could build in response. Dan Olsen's product-market fit pyramid puts target customer and underserved needs in the problem space, and feature set and user experience in the solution space.

What is the solution trap in product management?

The solution trap is jumping from a request or complaint straight to building a feature, before anyone has checked who has the problem, how often and at what cost. It feels productive because solutions are concrete and demoable. It produces features that miss the real need, which is one reason so many features go largely unused.

Did Einstein say he would spend 55 minutes defining the problem?

There is no evidence he did. Quote Investigator found no substantive evidence for the attribution, and the earliest Einstein version Barry Popik found dates from a 1995 newspaper, 40 years after Einstein's death, with no source. The idea behind it, that defining the problem well matters more than solving it fast, is sound. The quote is not.

How much time should a PM spend on the problem versus the solution?

No study defines an ideal ratio. Scale problem time to the size and reversibility of the bet: an hour for a small reversible change, a few days for a two-week feature, weeks for a quarter-long bet. As a measure of the gap, product people in Pragmatic Institute's 2019 survey spent 27% of their time on strategic work but wanted 53%.

How do you know you understand a problem well enough to build?

You can fill in every line of a problem brief with evidence: who has it, how often, what it costs, what they do today and whether it is a constraint or a symptom. You can name at least three possible solutions. And you can say what result after launch would prove you wrong. If any of those is missing, keep researching.

Is it ever right to start in the solution space?

Yes, when the change is cheap and easy to reverse, or when a rough prototype is the fastest way to learn about the problem. A prototype shown to five users often surfaces the real need faster than abstract questions. The risk is treating that prototype as the answer rather than as a probe that sends you back to the problem space.

Sources

  1. Harvard Business Review — Are You Solving the Right Problems? (Thomas Wedell-Wedellsborg, 2017)
  2. Harvard Business Review — Are You Solving the Right Problem? (Dwayne Spradlin, 2012)
  3. Quote Investigator — I Would Spend 55 Minutes Defining the Problem and then Five Minutes Solving It (2014)
  4. Barry Popik — If I had an hour to solve a problem
  5. Mind the Product — The playbook for achieving product-market fit (Dan Olsen, 2017)
  6. Mind the Product — Mastering the Problem Space for Product/Market Fit by Dan Olsen (2018)
  7. Product Talk — Opportunity Solution Trees (Teresa Torres, 2023)
  8. Pragmatic Institute — 2019 Annual Product Management and Product Marketing Survey
  9. Pendo — The 2019 Feature Adoption Report

This story was originally published on 13 July 2026 at https://www.specky.space/blog/pm-solution-trap-spend-time-on-problem.

Top comments (0)