Once a month the NFIB Research Center puts one question to American small business owners with a fixed answer sheet: what is the single most important problem facing your business today? In August 2026, 476 owners answered out of a 5,000 person sample, a 9.5 percent response rate, and they chose from nine named categories: taxes, inflation, poor sales, financing and interest rates, cost of labor, quality or availability of labor, government regulation, competition from large business, and cost or availability of insurance (NFIB Small Business Economic Trends, August 2026).
There is no box for “the customer could not tell the difference between us and the next quote.”
That problem is real and it is expensive, and because the answer sheet has no line for it, it gets filed every month under some other name. Poor sales took 10 percent of the August responses. Inflation took 16 percent, against a historical average of 7 percent for that category. Competition from large business took 4 percent, down from 5 percent a year earlier, sitting second from the bottom of the list in the same report whose Member Quotes section runs a heading called Large Business Competition, under which a retailer in Arkansas writes that the business “has been damaged by cheap imports and big-box competitors on the internet” and a retailer in Michigan writes that “people do so much more shopping on the internet; it is hard to compete with their prices.” Those quotes are illustrative rather than a measurement, and NFIB presents them that way. That is exactly what makes the pairing worth noticing: the thing owners describe in their own sentences is the thing almost none of them selected on the form (NFIB, August 2026).
A weak value proposition never surfaces under its own name. It surfaces as a slow month, as a price that suddenly feels too high, as a big competitor who seems to be everywhere. And each of those names points at a different fix, which is why owners so often spend money on the wrong one.
What the canvas is, and what the official one costs
The tool people are usually pointed at is the Value Proposition Canvas, from the book Value Proposition Design. Its publisher, Strategyzer, hosts the official page for it, and that page is a sales page. Read live today, it prices the offer twice, at “$299/yr · All playbooks and templates · 14-day money-back guarantee” near the top and “$299/yr · Cancel anytime · Team plans available” at the bottom, with the line “A great tool is just the starting point” in between (strategyzer.com).
Here is the part worth knowing before you decide whether to pay. That same page, above anything you have to buy, describes the entire method in two sentences. The customer half asks you to “identify your customer’s major Jobs-to-be-done, the pains they face when trying to accomplish their Jobs-to-be-done and the gains they perceive by getting their jobs done.” Your half asks you to “define the most important components of your offering, how you relieve pain and create gains for your customers.”
That is six lists. Jobs, pains and gains on the customer side. Products and services, pain relievers and gain creators on yours. A subscription buys facilitation guides, session agendas, digital whiteboards and the rest of the Strategyzer library, which is a real thing to want if you are running a workshop for a team. It does not buy you the six lists. Six boxes on a sheet of printer paper hold them, and the hard part was never the drawing.
Harvest the words before you fill in anything
The single most common way this exercise fails is that the owner fills all six boxes from memory in twenty minutes. Memory is where the current claim came from, so a canvas built from memory returns the claim you already have, dressed up.
Collect raw customer language first, verbatim, before you write a word of your own. Four places hold it already, and none of them cost anything:
- Your Google Business Profile reviews, including the mediocre ones. Read the last 20 and copy out the actual phrases, not your summary of them.
- Your sent folder. Pull the last 20 quote requests or inquiry emails and copy what the customer wrote before you replied.
- Your voicemail and your text threads, where people describe a problem in the words they would use to a neighbor.
- Your lost jobs. The reason a customer gave for going elsewhere is the most valuable sentence in the pile, and it is the one nobody writes down.
Copy phrases, do not translate them. If three people wrote “nobody called me back,” that is your raw material. “Responsive communication” is not; that is your marketing voice quietly replacing theirs. If most of what gets said in your business gets said out loud rather than typed, transcription is now cheap enough to catch it, which we covered in AI Transcription for Small Business: Listening Got Cheap.
The customer side, and why it goes first
Fill these three from the pile you just harvested, and fill them before you touch your own side.
Jobs. What the customer is trying to get done, stated as the outcome they wanted, not the service you sell. A homeowner calling a plumber at 7am is not buying a repair, they are trying to get to work without a flooded kitchen. Write the job in the customer’s timeframe, because the timeframe is usually the job.
Pains. What goes wrong while they try. This is where your lost jobs and your three-star reviews earn their place. Pains are specific and they carry evidence: a quote that took nine days, a technician who did not text ahead, a price that changed after the work started.
Gains. What a good outcome looks like to them, including the parts that are not about the work. Some gains are things they will never say to your face, such as not looking foolish to a spouse or a business partner for having chosen you.
Order matters here for one reason. Your side of the canvas is the only side you can write without evidence, so if you start there, you will spend the rest of the exercise finding customer language that flatters what you already wrote. Doing the customer side first makes that harder, which is the point.
Your side, and the line that saves money
Products and services. Everything you actually sell, listed plainly, including the things you offer because you always have.
Pain relievers. For each pain on the customer side, the specific thing you do that removes it. Not “great service.” The named practice: a same-day callback, a fixed quote in writing, a text when the van leaves the previous job.
Gain creators. The same discipline applied to the gains column.
Now the step that most write-ups skip, and the one that pays for the afternoon. Draw a line from every item on your side to the specific customer-side item it addresses. Some will not reach anything. Those items are not a marketing failure, they are a cost you are carrying for nobody: a service tier nobody asked for, an accreditation you renew annually that never once appeared in a customer’s sentence, a feature you pay a subscription to support. This exercise is usually sold as a way to make more money. Its first honest result is often a list of things to stop paying for.
The same care applies to reading pains correctly. A price objection is frequently a pain about certainty rather than about money, which changes the fix entirely, and we worked through how to sort the kinds of no in Sales Objection Handling: A Small Business Playbook.
Where the claim goes, and what it replaces
A value proposition that lives in a document is not installed. Take the strongest connected line on your canvas, write it as one sentence, and put it in the three places where a stranger actually decides:
- The description on your Google Business Profile, which many customers read before your website.
- The top of your quote or estimate, where the reader is comparing you against another number on another piece of paper.
- The first fifteen seconds of the phone call, spoken by whoever answers it, which means it has to be a sentence a human can say without wincing.
Each one of those already contains a sentence. Note what you deleted. If nothing had to be deleted, the new claim is probably not saying anything the old one did not. And if the strongest support you have for the claim is what customers said about you, be careful how you publish it: the FTC’s Rule on the Use of Consumer Reviews and Testimonials has been in effect since October 21, 2024 (ftc.gov) and it treats a sentence on your homepage differently from the same sentence on Google, which we covered in Customer Testimonials: The FTC Rule Changed the Ask.
What AI is actually good for here
In the Federal Reserve’s 2026 Report on Employer Firms, fielded from September 3 to November 14, 2025 with 6,525 responses from firms with 1 to 499 employees, “nearly half of firms (46%) reported that their business or its employees currently use AI, while an additional 15% planned to begin using it in the next 12 months,” and the most common task by a distance is writing or marketing at 83 percent (fedsmallbusiness.org). Which is to say most owners are already pointing these tools at exactly this job.
They are good at two parts of it. Paste your harvested phrases in raw and ask the model to sort them into jobs, pains and gains, keeping the customer’s wording and flagging anything ambiguous rather than guessing. Then ask it for ten candidate claims drawn only from lines that appear in the pile. Both tasks are sorting and drafting against evidence you supplied.
They are not good at supplying the evidence. Ask a model what your customers care about and it will answer, fluently, from a general picture of your industry, and you will have automated the memory problem instead of fixing it. The same Fed report has current AI users naming accuracy as their top challenge at 46 percent (fedsmallbusiness.org), which is the polite version of this.
One number in that report deserves more attention than it gets. Among firms using AI, 71 percent said it led to increased productivity while “the vast majority of firms that use AI experienced no change in their labor costs” (fedsmallbusiness.org). The gain showed up without anyone being removed from the payroll to collect it. That is the right shape for this work in particular, because the people who can actually fill the customer side of your canvas are the ones already answering your phone.
The tools, and what they cost
You may find your existing reviews and emails are enough, in which case skip this section and keep the money. If you want to ask customers directly, three real options, at prices read on the vendors’ own pricing pages today:
- Jotform. Starter is $0 with 5 active forms and 100 monthly submissions. Bronze is $39 per month, or $408 per year, with 25 forms and 1,000 monthly submissions (jotform.com).
- Typeform. The free plan is $0 with 100 responses per month. Basic is $28 per month billed annually, or $39 billed monthly, and also carries 100 responses per month; Plus is $56 per month billed annually with 1,000 (typeform.com).
- SurveyMonkey. The Basic plan is free and capped at 10 questions per survey with limited responses per survey. Advantage Annual is $39 per month billed annually, which is $468 a year, for 15,000 responses per year, with additional responses at $0.15 each (surveymonkey.com).
Read those response limits against your actual need. An owner trying to understand why quotes are not converting needs a few dozen honest answers, not 15,000, so the free tiers are not a trial here, they are the correct plan. The paid tiers buy volume and reporting that this particular job does not use.
The prompt page, and a mismatch worth knowing about
Our sibling site BusinessPrompter.com publishes a Value Proposition Designer prompt, filed under Strategy and Planning, summarized as “craft compelling value propositions that resonate with target customers.” It is a Pro prompt, behind the upgrade wall, alongside 255 other premium ones out of the 335 plus on the site.
Read the page carefully before you use it, because its three About This Prompt panels are describing a different job than its title. “What This Prompt Does” is entirely about pricing: setting prices that attract customers while remaining profitable, experimenting with pricing models, identifying optimal price points. “Who It’s Best For” names early-stage startups and product managers “seeking to validate pricing strategies,” and “What You’ll Have When You’re Done” promises “a data-driven strategy outlining optimal price points and testing methodologies.” Pricing is a genuine and adjacent problem. It is not the job named in the title, and if what you need is the six lists, the panels will send you somewhere else. Take the title and the summary as the description of the prompt, and treat the panels as belonging to a pricing exercise you may or may not want next.
Our read, and it is a read
A judgment rather than a reported finding, so treat it as one. The reason the missing checkbox matters is that the substitute names each carry a reflex, and the reflex attached to poor sales is to cut the price.
At the level of the whole economy that reflex is not what is happening. In the same August 2026 NFIB report, a seasonally adjusted net 31 percent of owners raised their average selling prices, against a historical average of net 14 percent, and among owners reporting lower earnings only 3 percent named cutting selling prices as the reason, the smallest category on that list (NFIB, August 2026). So this is not a discounting economy. The temptation is individual, it arrives on a slow Tuesday, and it is nearly irreversible, because a price cut made to fix a claim problem teaches your existing customers what you are worth while leaving the claim exactly as unclear as it was.
The canvas is not magic and a good sentence will not save a business that is genuinely more expensive and no better. What the exercise does is give the failure somewhere to show up under its own name, which the survey form and the accounting software both fail to do.
Frequently asked questions
Do I have to pay for the Value Proposition Canvas?
No. Strategyzer’s official page prices its playbook subscription at $299 a year, and that buys facilitation guides, session agendas and the wider toolkit. The method itself is described on that same page above the paywall as six components: the customer’s jobs, pains and gains, and your products and services, pain relievers and gain creators. Six boxes on paper will hold them.
How many customers do I need to hear from before the customer side is usable?
There is no published threshold, and anyone quoting one is guessing. The practical answer is to start with evidence you already own, which is your existing reviews, quote requests and lost-job reasons, and only run a survey if that pile leaves a specific question unanswered. Note that the free tiers of Jotform and Typeform both cap at 100 responses per month, which is more than most owners will collect for this.
Is a value proposition the same thing as a tagline?
No. A tagline is written to be memorable; a value proposition is written to be checkable, and the six lists exist so that every claim on your side points at something a customer actually said. It is also a different thing from an employee value proposition, which is the offer you make to people you are hiring rather than to people who are buying, covered separately in Employee Value Proposition on a Small Hiring Budget.
Can I just ask ChatGPT or Claude to write my value proposition?
You can ask a model to sort customer language you supply and to draft candidate claims from it, and both work well. Asking it what your customers want, with nothing supplied, returns a plausible industry average in your market’s general voice. In the Federal Reserve’s 2026 Report on Employer Firms, current AI users named accuracy as their top challenge at 46 percent (fedsmallbusiness.org), and this is the shape that problem takes here.
Give it a date
Pick the one number the claim should move. Not revenue, which absorbs too much noise to tell you anything. Something narrow, such as the share of quotes that turn into jobs, or how many callers mention the specific thing your new sentence promises. Write today’s figure down, put the claim in the three places, and write a date eight weeks out next to it.
If the number has not moved by then, you have learned something real, which is more than a slow quarter ever tells you. The whole reason this problem hides inside other categories is that nobody ever gave it a deadline of its own.
