On Pivots

Score the struggle, not the solution

If you're staring down a pivot, the postmortem almost always lands on the offering: wrong price, wrong feature set, wrong channel. Almost nobody goes back one layer further, to the struggle you built the offering on top of, and asks whether that struggle was ever strong enough to carry a company through a seed round, let alone a Series A.

John Gusiff · Customer Centric Solutions LLC · Seattle, WA

Here is the pattern I keep seeing in switching interviews, win-loss calls, and the "why didn't this land" retros founders run after a launch underperforms: you can describe your customer's problem in perfect detail. You just can't say why that customer would ever act on it, not to an investor, and not to yourself at midnight.

That's not a positioning problem. It's not a GTM problem. It's a struggle problem, and it usually got baked in before a single line of code was written, back when you or a cofounder said some version of "our customers struggle with X" in an early founding conversation, and everyone nodded and moved straight to solutioning.

Nobody scored the struggle. Everyone just believed it.

Two ways a struggle fails you

The first is the one nobody wants to admit to: the struggle was invented. It came from your last job, a LinkedIn thread, a competitor's marketing copy, an assumption dressed up as a customer insight. You haven't actually watched a target customer live through it, so you can't say what it's really costing them, or whether they've already made peace with it.

The second is subtler, and more common among founders who did talk to customers: the struggle is real, and customers will agree to it in an interview, but it's mild. It's a real problem the way a squeaky door hinge is a real problem. True, mildly annoying, and not worth doing anything about. You can validate that a struggle exists in every discovery call you run and still spend a chunk of your seed round building on ground that can't hold a company.

Both failures produce the same symptom eighteen to twenty months later: a technically sound product, a deck that says you did "customer research," a shrinking runway number, and a pivot memo that blames the solution.

The tell: if your growth plan depends on educating the customer that they have a problem, you're not looking at a struggle. You're looking at an opportunity you're excited about. Those are different companies to build, with very different runway requirements, and only one of them fits inside eighteen months of seed capital.

Not sure yet which one you're building on, a struggle or an opportunity you're excited about?

Book a discovery call

Questions founders ask before they build

How do I know if my customer's problem is a real struggle, or an opportunity I'm excited about?

Check whether your growth plan depends on educating the customer that they have a problem. If it does, you're looking at an opportunity, not a struggle, and those two need very different amounts of runway. The struggle rubric below scores the difference in about ten minutes.

Why did my product validate in every customer interview but still fail to grow?

Validation and strength are different questions. A struggle can be real and still be mild, true the way a squeaky door hinge is true, and not worth anyone acting on. Score its urgency against four signals, disruptive, necessary, hard to postpone, consequential, then separately check what they've actually done about it. Stated urgency with no evidence of action yet isn't proof, it's a flag to look closer, before you build, not after.

What's the difference between a pain point, a job story, and a struggling moment?

A struggling moment is dated: it names when the customer's old way stopped working well enough, and why they started looking now instead of six months earlier. A pain point or a job story usually can't answer that "why now" question, which is the fastest way to tell them apart.

How do I score a customer struggle before I spend a quarter of runway building for it?

Name what's primarily at stake, time, money, risk, or identity. Then score two things separately: how urgent it is, based on what they described, and what they've actually done about it, searched, worked around it, spent money, switched. A 0 or a 1 rarely creates demand on its own, a 2 means the struggle is real but you likely caught it early, and a 3 means you have both the stakes and the proof.

Why does the rubric score urgency and evidence separately instead of one number?

Because a struggle can be severe and brand new at the same time, caught the week it triggered, before the customer has had any chance to search or act. Score urgency and evidence as one blended number and that struggle looks identical to a genuinely mild one. Score them separately and it shows up as what it is: real, and early.

Should I change who I sell to, what I solve for them, or both?

Pick one door. Door One changes the executor and keeps the job the same, Door Two keeps the executor and sharpens the job. Changing both at once turns your pivot into an accidental new company, with no way to tell which change did the work.

What "strong enough" actually means

A struggle worth building on has two independent properties, and most founders only ever check the first one.

The first is what's at stake. Every real struggle threatens something specific: time the customer can't get back, money that hits revenue or budget, risk of failure or exposure, or identity, meaning how they see themselves or how they're seen by people who matter to them. Naming which one is primarily at stake forces a level of precision most discovery calls never reach. "It's inefficient" is not a stake. "It costs the ops lead four hours every Friday she's supposed to be spending on hiring" is.

The second, and the one that actually predicts demand, is strength, and it has two parts that need to be scored separately. The first is urgency: is it disruptive rather than just annoying, does it have to be resolved rather than sitting on a someday list, is it hard to postpone past the next planning cycle, does ignoring it have a consequence the customer can point to. Urgency is what the customer told you, and you can score it the moment you hear it. The second is evidence: what they've actually done about it, searched, worked around it, spent money, switched. A struggle can score high on urgency and still be new enough that no evidence has shown up yet. That's not a false alarm. It just means you caught it early, and the rubric below is built to say so instead of quietly scoring it the same as a struggle nobody will ever act on.

Score all three, stake, urgency, and evidence, and you get something a lot more honest than a persona doc: a number that tells you whether you're looking at a real problem, an interesting one, or a company worth spending your runway on.

The rubric

This is the version I now ask founders to fill out for every struggle they're considering building the next twelve to eighteen months on, before it becomes a roadmap item, or worse, the thesis of a pitch deck. It takes about ten minutes if the struggle came out of a real interview, and considerably longer, tellingly, if it didn't.

Working Tool

Score One Real Customer Struggle

A specific moment when a target customer said their current way of getting something done was no longer working well enough.

Write the sentence first: when , I realized was no longer working well enough, because .

The test: why did they start looking now, and not six months earlier? If you can't point to the "when," you're probably holding a pain point or a job story, not a struggling moment.

1What is primarily at stake?

Time
Hours they cannot recover
Money
Revenue, margin, or budget
Risk
Failure, blame, or exposure
Identity
How they see themselves, or are seen

2How strong is it?

Urgency · score this the moment you hear it

No action required from the customer yet, this is about what they described, not what they've done. 0–2 checked: low urgency. 3–4 checked: high urgency.

Evidence · the highest thing they've actually done

0 weak signal, unlikely to create demand  ·  1 mixed signal, real but survivable  ·  2 early signal, worth investigating now  ·  3 strong signal, demonstrated demand  ·  urgency and evidence are scored separately, then combined: high urgency with no evidence yet still lands on a real 2, it just means you caught the struggle early Score: 0 · Weak signal
The short version: ask whether it's bad enough to act on, disruptive, necessary, hard to postpone, consequential. Then ask what they've actually done about it, searched, worked around it, spent money, switched. A strong yes to the first question with nothing yet on the second isn't a false alarm. It just means you caught the struggle early, and that's still worth a second look.

Notice what the rubric does not ask: whether you can build a solution for it, whether it's a big enough market, whether a competitor already serves it. Those are all real questions. They're just the second question, not the first. Score the struggle before you let your team start solutioning, and you'll kill a lot of ideas in the first ten minutes instead of eighteen months of runway.

Evidence of a workaround or spend, the top of that Evidence axis, is also the fastest way to find what the Should We Pivot diagnostic calls the fingerprints of struggle: the spreadsheet with eleven tabs nobody wants to touch, the contractor hired just to duct-tape two tools together, the manual process someone is already paying to avoid. If you can point to one of those, you're not stuck at emerging, you already have demonstrated evidence.

This is a hypothesis until you reconstruct the customer's timeline.A Customer Progress Interview, a Switching Interview paired with a Core Jobs Interview, reveals the trigger, consequences, search behavior, workarounds, and post-purchase outcome needed to score a struggle from evidence rather than team assumptions.
The score is not the research. It is the structure for the research.

You can identify a plausible struggle in a workshop. You need a Customer Progress Interview to determine whether it actually created demand. A team can usually hypothesize what's at stake. What it cannot reliably reconstruct from internal knowledge alone is the specific event that made the old way stop feeling acceptable, why the customer acted then and not six months earlier, whether the problem was truly hard to postpone, and what searching, workarounds, spending, or switching actually followed.

Rubric questionWhat the interview reconstructs
What was at stake?The functional, financial, social, or identity consequence
Was it disruptive?What stopped working, and how daily behavior changed
Was resolution necessary?Why tolerating the old way was no longer acceptable
Was it hard to postpone?The "why now" trigger and any deadline behind it
What did they do?Searching, workarounds, spending, trial, and switching behavior

Switching Interview

Reconstructs the trigger, the "why now" timeline, and the switching behavior itself, what they searched for, worked around, spent on, or replaced. This is where the Evidence axis comes from.

Core Jobs Interview

Establishes the full functional, social, and emotional job the customer is hiring a solution for, including what happened after they switched. This is where the Urgency axis, and the check on whether progress actually followed, comes from.

Run both together and you have what a Customer Progress Interview actually is: not a separate, seventh method, but a Switching Interview paired with a Core Jobs Interview. The Switching Interview supplies the evidence this rubric asks for, not what the customer said they'd do, but what they actually did. The Core Jobs Interview supplies the stakes, and closes the loop after the fact: did real progress follow, or did the workaround just move the problem somewhere else.

Use this tool to choose which struggles deserve investigation, not to declare them validated. Then run a Switching Interview to reconstruct the "why now" moment and the behavior that proves whether real demand exists, paired with a Core Jobs Interview to confirm the stakes are real and see what actually happened after. Together, that's a Customer Progress Interview.

Scored a 1 or a 2 just now? That's exactly the range a Customer Progress Interview is built for, real enough to take seriously, not yet proven. Let's look at what one would surface.

Book a discovery call

This doesn't require a Systrom-style rebuild

Everything above can sound like it only applies pre-seed, when you're deciding what to build in the first place. It applies just as often to companies with a working product and paying customers, where the fix isn't a pivot at all, just a sharper version of the same struggle, found without changing the customer, the workflow, or the product underneath it.

Run each one through the same rubric from above. Nobody at either company filled out a card like this, but the shape holds.

Gong

Records and analyzes sales calls, originally sold as a coaching tool for sales managers. Founded in 2016 by Amit Bendov and Eilon Reshef, who raised a $6M seed round led by Norwest Venture Partners and Shlomo Kramer to get started.
Before · 2016 to 2018
1· what's at stake
Time

When headcount grew past what one manager could shadow, I realized sitting in on live calls was no longer working well enough, because most reps went uncoached for weeks at a time.

in their own words

"The sales manager is struggling to coach reps because they can't listen to every call."

2· how urgent is it
Disruptive, not just annoyingfrustrating for managers, but reps kept hitting quota anyway
Necessary to resolve, not optionallistening to more calls was a nice-to-have, nobody was budgeting for it
Hard to postponemanagers had gotten by on spot-checks for years, no forcing event
Significant consequence if ignoredreps went uncoached for weeks and it showed up in lost deals
3· what did they do
Nothing yetno coaching workaround in place beyond occasional spot-checks, nobody had committed budget or time to fixing it
Score: 0 · weak signal, unlikely to create demand
+3
After · 2019 onward
1· what's at stake
Risk

When forecast calls started running on rep-reported updates, I realized trusting what a rep said about a deal was no longer working well enough, because the revenue leader was committing a number to the board with no way to verify it.

in their own words

"I'm committing a forecast without knowing what's actually happening in our deals."

2· how urgent is it
Disruptive, not just annoyingleadership stopped trusting the forecast number itself
Necessary to resolve, not optionala board-facing forecast can't run on guesswork
Hard to postponethe next forecast call was already on the calendar
Significant consequence if ignoredmissing the number meant answering for it to the board
3· what did they do
Built a workaround, or committed additional time or moneyrevenue leaders were already building manual deal-risk spreadsheets and paying analysts to reconstruct pipeline health by hand
Score: 3 · strong signal, demonstrated demand
What followed: conversation intelligence became revenue intelligence; a $200M Series D at a $2.2B valuation, ARR since past $500M and growing 55%+ year over year.

Chili Piper

Scheduling and lead-routing software that connects inbound leads with a sales rep to book a meeting. Founded in 2016 by husband-and-wife team Nicolas and Alina Vandenberghe, who bootstrapped past $2M in ARR before raising a $3M seed round led by Flashpoint Venture Capital.
Before · 2016 to 2019
1· what's at stake
Time

When inbound volume outpaced manual scheduling, I realized emailing back and forth to find a time was no longer working well enough, because reps were losing hours a week just landing on a meeting slot.

in their own words

"The sales team is struggling to schedule meetings because coordinating calendars takes too long."

2· how urgent is it
Disruptive, not just annoyingannoying back-and-forth, but reps still got meetings booked eventually
Necessary to resolve, not optionaltedious, but nobody was escalating it
Hard to postponeteams had lived with the scheduling shuffle for years
Significant consequence if ignoredreps were burning hours a week just landing on a time
3· what did they do
Nothing yetreps just lived with the back-and-forth, nobody had built a fix or paid to avoid it
Score: 0 · weak signal, unlikely to create demand
+3
After · 2020 onward
1· what's at stake
Money

When a form-fill sat in a queue waiting for a rep to notice it, I realized following up whenever someone got to it was no longer working well enough, because every hour of delay was pipeline quietly going to a faster competitor.

in their own words

"A high-intent prospect just filled out our form, and we might lose them while sales is still getting to it."

2· how urgent is it
Disruptive, not just annoyingsales stopped trusting that inbound leads got followed up on at all
Necessary to resolve, not optionala lead that goes cold is pipeline that doesn't come back
Hard to postponeevery hour of delay favored whichever competitor responded first
Significant consequence if ignoredlost meetings meant lost revenue leadership could point to directly
3· what did they do
Built a workaround, or committed additional time or moneysales ops were already cobbling together manual round-robin sheets and paying for basic scheduling links just to keep leads from going cold
Score: 3 · strong signal, demonstrated demand
What followed: scheduling got reframed around converting inbound demand; reported near $43M ARR at a roughly $625M valuation, about 14x revenue, priced like pipeline infrastructure rather than a calendar tool.

Sources: Gong Series D announcement, Gong ARR update, Chili Piper revenue estimate, Latka

Same customer base in both cases. Same call data, same workflow, same product underneath. Neither company needed a new customer or a rebuild to do this, they needed the exercise on this page, run honestly, against a struggle they'd already convinced themselves they'd validated. The growth followed the stake, not the other way around.

That's one of exactly two doors. This is a Door Two move: same executor, a sharper job. Door One runs the same exercise on the other variable, same job, a different executor, the way ConvertKit stalled for over a year building generic small-business email software and only moved once Nathan Barry narrowed it hard to professional bloggers, whose actual struggle was tagging and segmenting an audience to sell to later, not "send a newsletter." Either door can hold a company. What the rubric on this page won't forgive is walking through both at once: change who you sell to and what you solve for them in the same swing, and you've made a new company by accident, with no way to tell which change did the work. Score the struggle for whichever side of the door you're on before you commit to it, that's the full framework, and it lives on the Should We Pivot diagnostic. Door One and Door Two are the two moves that show up most often, but they're not the only ones; if neither fits what you're looking at, the other eight are mapped out in Every Pivot Has a Required Input.

Sometimes the fix is what you build, not what you say

The Gong and Chili Piper moves didn't require rebuilding the product around a new customer or workflow, they began by recognizing the product already resolved a more consequential struggle than the one it was being sold on. Sometimes the honest answer to the same exercise is different: a company builds something real on one strong struggle, then later runs into an adjacent struggle, just as consequential, that the product hasn't answered yet. That's not a correction of a mistake, the first struggle was never weak, it's expansion: same customer, same core product, but this time the fix is something you build.

Run each one through the same rubric below. Neither struggle needed a new customer to be real, just a company far enough along to run into it.

Stripe

Payments infrastructure that lets businesses accept and manage online payments without building their own systems. Founded in 2010 by Patrick and John Collison, who raised a $2M seed round in 2011 from investors including Sequoia Capital, Andreessen Horowitz, Peter Thiel, and Elon Musk.
Before · 2010 to 2015
1· what's at stake
Time

When integrating payments required weeks of engineering work, I realized building and maintaining our own payment stack was no longer working well enough, because every week spent on it was a week the business couldn't launch or collect a dollar of revenue.

in their own words

"We need to start accepting payments without building the infrastructure ourselves."

2· how urgent is it
Disruptive, not just annoyingblocking rather than disruptive, there's no ongoing operation yet for it to disrupt
Necessary to resolve, not optionala business cannot accept payments at all without solving this
Hard to postponeit blocks launch and revenue entirely until it's built
Significant consequence if ignoredno payments infrastructure means no revenue, full stop
3· what did they do
Actively searching for a better wayfounders kept hearing the same complaint from other developers who'd spent weeks evaluating banks and legacy processors, with no workaround yet in place
Score: 2 · early signal, worth investigating now
+1
After · 2016 onward
1· what's at stake
Risk

When fraudulent transactions began rising with payment volume, I realized manually reviewing suspicious payments was no longer working well enough, because every decision either exposed us to losses or rejected a legitimate customer.

in their own words

"We're losing money to fraud, but tightening the rules is also blocking good customers."

2· how urgent is it
Disruptive, not just annoyingfraud review turned into a constant operational drag on the team
Necessary to resolve, not optionalpayments can't scale safely without catching fraud automatically
Hard to postponeevery transaction processed was live exposure, no waiting
Significant consequence if ignoredchargebacks, lost revenue, and legitimate customers blocked
3· what did they do
Paid for, switched to, or seriously attempted a replacementbusinesses were already paying separately for bolt-on fraud tools and eating chargebacks as the cost of staying live
Score: 3 · strong signal, demonstrated demand
What followed: the payments struggle was strong enough to build a company on immediately, no correction needed. As volume scaled, fraud became an equally consequential struggle, and Stripe built Radar directly into the payment flow; it reported preventing $4B in attempted fraud in 2017 alone.

Figma

Browser-based design tool for creating and prototyping interfaces. Founded in 2012 by Dylan Field and Evan Wallace, who raised roughly $4M in seed funding in 2013 to build it.
Before · 2013 to 2016
1· what's at stake
Time

When designing meant installing desktop software and passing files back and forth, I realized save-and-send editing was no longer working well enough, because every handoff risked opening the wrong version or overwriting someone else's work.

in their own words

"I need a design tool that just works in the browser, without me juggling which file is the current one."

2· how urgent is it
Disruptive, not just annoyingswitching tools and passing files around broke workflow constantly
Necessary to resolve, not optionaldesigners needed real access without installs or platform lock-in
Hard to postponeteams had lived with file-based handoffs for years, it was an accepted cost
Significant consequence if ignoredversion conflicts meant real work got lost or overwritten
3· what did they do
Actively searching for a better wayteams were already asking around for a better way to hand off files, without landing on a fix that actually stuck
Score: 2 · early signal, worth investigating now
+1
After · 2017 onward
1· what's at stake
Risk

When multiple designers began editing separate copies of the same file, I realized passing files back and forth was no longer working well enough, because nobody could tell which version was current before a release went out.

in their own words

"I can't tell which file is the real one anymore, or whether the feedback even made it in before this ships."

2· how urgent is it
Disruptive, not just annoyingdesigners stopped trusting that the file open in front of them was the real one
Necessary to resolve, not optionala shared source of truth wasn't optional once more than one person touched a file
Hard to postponethe next handoff or release was already coming, with no time to reconcile versions
Significant consequence if ignoredshipping the wrong version meant reverted work or a broken release
3· what did they do
Paid for, switched to, or seriously attempted a replacementteams were already paying for separate version-control tools bolted onto their design software just to stop overwriting each other's work
Score: 3 · strong signal, demonstrated demand
What followed: Figma shipped multiplayer editing directly into the same tool in 2017, along with shared components, commenting, and version history; it became the product's defining difference from desktop tools, drawing a $20B acquisition offer from Adobe in 2022 (later terminated) and going public in 2025 above an $18.8B valuation.

Sources: Stripe Radar 2.0 announcement, Figma, Multiplayer Editing in Figma, Figma/Adobe deal coverage, Crunchbase News

Notice what didn't change in either case: the customer stayed the same, and so did the core product category. What changed was which struggle had earned a roadmap slot instead of a release note, and in both cases that struggle was already strong before the company built anything for it, the rubric just confirmed it was strong enough to justify the engineering time.

Gong, Chili Piper, Stripe, and Figma didn't guess their way to a 3. If you want a second set of eyes on your own struggle sentence, bring it to a call.

Book a discovery call

Where the sentence has to come from

The rubric only works if the opening sentence comes out of a real customer's actual words during a switching interview, not out of a whiteboard session:

When ___, I realized ___ was no longer working well enough, because ___.

All three blanks have to trace back to something specific: a moment you can point to, an old way you can name, a stake you can defend. If you're filling this out from memory of a sales call, or worse, from what you assume must be true of "customers like this," you're scoring your own belief and it will score however you want it to.

That's the whole reason the Customer Progress Interview (a Switching Interview paired with a Core Jobs Interview) exists as an interview method: it walks a customer back through the First Thought and Passive Looking stages in chronological order, before the Hiring Moment, specifically so the sentence, all three parts of it, gets built from what they actually said, not from what you asked them to confirm.

You don't pivot the offering. You pivot the struggle, and the offering follows.

Once you've scored a struggle a 2 or a 3, the interesting part of the work is still ahead of you: what pushed them to start looking, what almost stopped them from switching, what they were afraid of getting wrong. But you're now spending your effort, and your runway, on ground that can hold a company. That's the difference between a pivot and a rewrite.

Next

Not sure if you have a struggle problem or a solution problem?

The Should We Pivot diagnostic walks through both, and the Switching Interview field guide has the full question set for pulling a struggle sentence out of a real customer instead of a whiteboard, before you spend another quarter of runway on the wrong one. Or skip straight to talking it through.