Customer pain points: find them in behavior, then rank them by cost

Customer pain points are the problems your customers already pay for in money, time or hassle, and the useful ones are the ones you can see in their behavior. Many guides sort customer pain points into four types (financial, productivity, process and support), which is a fine checklist and a poor substitute for evidence. Find real pains by watching what people do and asking about specific past events, then rank them by how often they happen and what each one costs.

The short version: a pain point counts when someone can tell you the last time it happened, what it cost, and what they did about it.

1. Define a pain point by what it costs the customer today

Jenny Keohane at Yesware defines pain points as problems causing prospective customers "pain" that require a solution. Make that definition stricter: a pain point costs the customer something today. If nothing is being lost, you have found a preference.

This framing keeps you honest. "Our reporting is clunky" is a complaint. "Every Monday, Sam spends two hours copying numbers into a slide deck" is a pain point, because you can put a price on it.

It also tells you what to write down in your notes: the cost, in the customer's own units (hours, dollars, missed deadlines, angry clients). Adjectives like "frustrating" go in the margin.

2. Use the four common types as a checklist, then go looking for evidence

A common version of the four types comes from the sales software company Yesware, which describes them like this:

Treat the table as a list of places to look. Before an interview, read down the rows and ask yourself which ones you have no evidence about yet. That gap is your interview plan.

Do not treat the type as the finding. Most real pains cross rows. Front gives the example of a customer who has to explain the same problem to several support agents because nobody can see the earlier conversation. That is a support pain, a productivity pain for the customer and a process pain for the vendor, all at once. Labelling it is less useful than measuring it.

Short version: the four types tell you where to ask. They do not tell you what is true.

3. Watch what customers do before you trust what they say

Behavior is better evidence than opinion. Jakob Nielsen made this case in 2001: across 113 pairwise comparisons of interfaces, users' stated preference and their measured performance correlated at 0.44. In a later update to the article he calls a similar figure low, because you can predict only about a quarter of how well a design works from how much users say they like it. He also notes that people bend the truth toward what they think you want to hear and rationalize what they did.

People are not lying when they do this. They are just bad at reporting their own habits. (Try guessing how many times you checked your phone yesterday.)

So start with the traces customers leave behind. Good places to look:

• Workarounds: the spreadsheet next to your product, the shared doc of copy-paste replies, the reminder someone sets because a notification never arrives
• Support tickets and chat logs, grouped by what the customer was trying to do
• Cancellation reasons and the accounts that shrank before they left (see churn analysis)
• Reviews of your product and of competitors, read for the task that failed rather than the star rating
• Sales call notes, especially the moment a prospect says "we currently handle that by..."

If you already have a pile of this, sort it before you collect more. The voice of customer post covers how to do that.

A workaround is the strongest signal on this list. Somebody cared enough to build a fix out of spreadsheets and sticky notes, on their own time, with no budget. That is a pain point waving at you.

4. In interviews, ask about the last time it happened

When you talk to customers, ask about a specific past event. "Tell me about the last time you closed the month" gets you steps, tools, people and delays. "What are your biggest challenges?" gets you a tidy list that sounds like a conference talk.

This idea has a long history. The critical incident technique asks people to recall and describe a specific time something affected an outcome. John Flanagan formally introduced it in a 1954 paper, after developing it with other psychologists in the US Aviation Psychology Program during World War II, as Maria Rosala of Nielsen Norman Group explains.

Rosala also names its weakness: participants often recall extreme events, and small issues and typical usage rarely come up. That matters for pain points, because the boring weekly annoyance often costs more in total than the memorable disaster. Pair incident questions with observation. Ask to see the spreadsheet. Ask them to share their screen and do the task.

Follow-ups that work well:

1. When did that last happen?
2. What did you do next?
3. Who else was involved?
4. How long did it take?
5. What did it cost you?

Avoid asking whether they would use your idea. Compliments are free, and The Mom Test explains why they tell you almost nothing. For a fuller question bank, see customer interview questions.

5. Ask what they have already tried, because real pain gets attention

A customer who has a real pain point has usually done something about it. They searched for a tool, asked a colleague, built a spreadsheet, hired a contractor, or complained to their boss about it. Ask "what have you tried so far?" in every interview.

If the answer is "nothing, really", take note. The problem may be real and still too small to act on. People live with plenty of mild annoyances; most of us have a kitchen drawer that sticks and no plan to fix it.

What they tried also tells you what you compete with. The jobs to be done post covers this idea in more depth: the rival for a new tool is often a spreadsheet and a tired person.

6. Rank pain points by frequency times cost, in dollars per month

Once you have a list, rank it with one number: how often the pain happens, multiplied by what it costs each time, expressed per month. This turns a pile of quotes into a priority list that a team can argue about productively.

For time costs, you need an hourly rate. The US Bureau of Labor Statistics puts the median annual wage for business and financial occupations at $82,660 in May 2025. Divided by 2,080 working hours (40 hours times 52 weeks), that is about $40 an hour before benefits. Use the rate for the actual role you are studying when you have it.

Here is a made-up example. Priya runs a 12-person bookkeeping firm. In interviews about her invoicing software, three pains come up:

The price increase was the pain Priya talked about most, with real feeling. The re-keying barely came up until she shared her screen, and it costs more than ten times as much. That gap between what people complain about and what costs them the most is what this ranking is built to catch.

Two adjustments keep the ranking honest:

• Mark rare, severe pains separately. A once-a-year compliance failure that could lose a client will look small on a monthly average and still deserve attention.
• Count how many customers share each pain. A $2,000 pain that only Priya has is a consulting project. The same pain across most of your interviews is a product.

Short version: rank by frequency times cost, then sanity-check the top of the list against how many customers have it.

7. Know when this approach does not fit

Frequency times cost works best for business customers with measurable work: hours, invoices, tickets, deals. It fits less well in three cases.

Consumer products often sell on feelings that do not convert to dollars, like embarrassment or the wish to look competent in front of friends. You can still ask about the last time it happened, but drop the cost column.

Brand-new categories may have no workarounds yet, because people do not know the problem can be solved. There, watch for effort people spend avoiding the task altogether.

Very early ideas sometimes have no customers to watch. In that case, talk to people who do the job today before you build anything, and treat every pain you hear as a hypothesis until you see it in someone's actual work.

Your next step

Pick the three pains you believe matter most and, for each one, write down when you last saw evidence of it in a customer's behavior. Any pain with no date next to it goes to the top of your next interview guide.

When the people with the pain are outside your customer list, Instant Expert can find people who match a description such as "operations managers at freight brokerages" and send them paid invitations for a short call or a written or voice answer, and you pay only if someone books or answers. The directory pages for customer experience consultants and heads of customer success are a good place to find people who see customer pain every day.