Two colleagues comparing a working product on a laptop with a one-page commercial sketch.

Product & Business Models

A useful product is not automatically a viable company.

People can be glad a product exists and still leave the company poorer every time they use it. Usefulness is a product question. Viability is whether the money and the work of delivery can survive contact with the promise.

Marcus Startup Advisor16 min read

Contents

Usefulness and viability are different questions

A product is useful when someone completes a real job with it and would miss it if it disappeared. A company is viable when the pattern of payment covers the cost of delivering that job, the cost of finding the next customer, and enough of the standing costs to continue. Those sentences sound obvious in a quiet room. They get collapsed in a pitch, because a story about happy users is easier to tell than a story about the cost of the next user.

The collapse is dangerous in both directions. A product nobody values is not rescued by a tidy spreadsheet. A product people genuinely value can still be a bad company if every additional use consumes more cash than it brings in. Volume is then not the cure. Volume is the mechanism that makes the loss larger. Founders who have been told to “get to scale” need a prior question: scale of what. Scale of a positive contribution can be a strategy. Scale of a negative contribution is a faster way to need someone else’s money.

This article answers that commercial question for a founder who already sees use, or who is about to price an offer. It is not a claim about search volume. Search Console and Keyword Planner figures for this site were not part of the draft, so none are invented here. The intent is ordinary and durable: unit economics, contribution, and the difference between a product people use and a business that can keep its promise. The reader is deciding whether to spend more on acquisition, delivery or fundraising before that difference is clear.

The case is MoviePass and its parent, Helios and Matheson Analytics Inc., because the company’s own Commission filings describe a popular subscription whose ticket cost sat on the wrong side of the price. The filings also show how much cash the parent advanced, and they contain a going-concern warning. Later, the SEC alleged that statements about profitability at the headline price were false. That sequence is more useful to a founder than another anecdote about a beloved app.

What the filings actually say

On 11 December 2017 Helios bought a majority of MoviePass. The 2017 Form 10-K, filed on 17 April 2018, says the purchase left Helios with 51.7 percent of MoviePass common stock, excluding shares underlying options and warrants. An option agreement of 11 October 2017 had given Helios the right to buy up to a further 20 million dollars of stock. The 10-K says that full exercise of that option took the holding to 62.41 percent, on the same exclusion.

The risk-factor language is the commercial core. MoviePass’s monthly plans, the 10-K says, allowed a subscriber to see a new film each day for the whole month. “In most cases, MoviePass pays the theaters the full cost for each movie ticket that a subscriber uses.” The filing then states the consequence in its own words: increased viewing “results in significant and increasing losses per subscriber and negative cash flow” and adversely affects financial condition and results. That is the company describing a model in which more use worsens the result. It is not a journalist’s gloss.

The same filing describes the exhibitor relationship. MoviePass had historically paid large national chains full price for each ticket bought through the application. In the past it had received as much as a 20 percent discount from a small group of exhibitor partners. It had not historically received a benefit from large chains for sending them customers, “for example, in the form of a portion of concession sales.” The company anticipated negotiating ticket discounts and a share of concessions. If those partnerships did not arrive, the 10-K says, MoviePass would likely have to keep paying full price at the large chains. The hoped-for repair was disclosed as a hope. The present tense of the model was full price, more viewing, larger losses.

A separate filing dated 12 February 2018 describes the consumer offer that had produced the growth. MoviePass, it says, let members see a new film every day for a monthly price of 9.95 dollars. In the four months after the August 2017 announcement of that plan, the company reported growth to more than 1,000,000 total subscribers on monthly or annual plans, and it said the service had passed 1,500,000 subscribers in January 2018. The same passage compares that pace with Spotify, Hulu, ClassPass and Netflix and says the comparison is “estimated based on information available publicly from various news and other sources.” Treat the subscriber counts as company-reported. Treat the peer comparison as the company’s estimate, not as an independent study. This article does not add a February 2018 subscriber figure, because that February filing, as checked, stops at the January count.

The cash, and the chart that is allowed

After the option was fully exercised, the 10-K says Helios advanced cash to MoviePass from 19 December 2017 through 20 February 2018 “to support MoviePass’ working capital and operational requirements, as well as to support the expansion of MoviePass’ business plans.” The amount was 55,525,000 dollars. On 8 March 2018 that advance was converted, in lieu of cash repayment, into additional MoviePass stock. The 10-K describes the agreed pre-money valuation used for that issuance as 240,000,000 dollars as of 31 December 2017.

A second advance followed. From 1 March 2018 through 12 April 2018 the company advanced a further 35,000,000 dollars. An agreement of 16 April 2018 converted that advance as well. The 10-K describes the pre-money valuation used there as 295,000,000 dollars as of 31 March 2018, for stock equal to 10.6 percent of MoviePass common stock, again excluding options and warrants. Valuations in a related-party issuance are not evidence of what an outside buyer would have paid. They are the figures the company disclosed for those agreements.

Cash Helios advanced to MoviePass, as disclosed in the 2017 10-K
First advance55,525,000Second advance35,000,000
Cash Helios advanced to MoviePass, as disclosed in the 2017 10-K
PeriodUS dollars advanced by the parent
First advance55,525,000
Second advance35,000,000

These are parent cash advances disclosed in the 10-K, not ticket losses and not a contribution-margin calculation. The first advance ran from 19 December 2017 to 20 February 2018. The second ran from 1 March 2018 to 12 April 2018. Source: Helios and Matheson Analytics Inc., Form 10-K filed 17 April 2018. The bars use these disclosed values only. Nothing has been estimated or smoothed.

The bars are those two advances and nothing else. They are not losses per subscriber. They are not the cost of a cinema ticket. They are not a forecast. A chart that divided either bar by a subscriber count and called the result “the unit loss” would invent a number the filing does not give. The filing says increased viewing produces significant and increasing losses per subscriber. It does not, in the passages used here, publish the arithmetic of one average subscriber’s month. Leaving that cell blank is the accurate chart.

The going-concern note in the same 10-K is a second primary record, and it is a table rather than a chart because the rows do not share a job. A net loss, a cash balance and a deficit should not be drawn as bars on one axis and then compared by eye.

Figures in the 2017 going-concern note
Item in the noteAmount reported, US dollars
Net loss for the year150,824,842
Net cash used in operating activities27,286,382
Cash24,949,393
Working-capital deficit107,097,249
Accumulated deficit189,495,185

Copied from the company’s going-concern discussion for the year ended 31 December 2017. The working-capital deficit includes 72,123,262 dollars of warrant and derivative liabilities classified as short-term. These rows are not a unit-economics model, and they should not be added together. Source: Helios and Matheson going-concern note, Form 10-K filed 17 April 2018.

The note says the company had suffered net losses and significant operating-cash outflows. For 2017 it reports a net loss of 150,824,842 dollars and net cash used in operating activities of 27,286,382 dollars. At 31 December 2017 it reports cash of 24,949,393 dollars and a working-capital deficit of 107,097,249 dollars, of which 72,123,262 dollars was warrant and derivative liabilities in short-term liabilities. The accumulated deficit was 189,495,185 dollars. Management concluded that, without additional funding, the company would not have sufficient funds to meet its obligations within a year of issuance. The note states substantial doubt about the ability to continue as a going concern through 16 April 2019 without raising more money. It also says the company secured 60,000,000 dollars of financing on 11 January 2018, and that total cash received from the public offering was 99 million dollars. Those financing lines show that capital was raised. They do not show that the subscription, by itself, funded the tickets.

The 2017 loss and the cash outflow are parent-company figures in a year when MoviePass was acquired on 11 December. They are not a clean twelve-month MoviePass profit-and-loss account. Using them as “MoviePass lost 150 million dollars on tickets in 2017” would be a misreading. The note supports a narrower statement. At the filing, the group’s own disclosure was that losses, cash consumption and the funding need raised substantial doubt, and that more capital was the plan.

What the SEC later alleged

Growth and a low price were not the end of the public record. On 27 September 2022 the SEC announced, in Litigation Release 25525, that it had charged Theodore J. Farnsworth, former chief executive of Helios and Matheson, and J. Mitchell Lowe, former chief executive of MoviePass. The release says the complaint alleged that, between August 2017 and at least March 2019, they intentionally and repeatedly made misstatements in filings, press releases and the press. Among those alleged misstatements was the claim that MoviePass could be profitable at the new 9.95 dollar monthly price, along with claims about data-analytics capability and about the parent’s ability to fund operations. The release also says the complaint alleged tactics to stop subscribers using the service as they expected. A complaint is an allegation. This article does not treat the 2022 release as a finding that every defendant had been held liable.

On 2 January 2025 the SEC published Litigation Release 26207. It says that on 23 December 2024 the Southern District of New York entered a consent judgment against Farnsworth. The release describes that judgment as addressing materially false and misleading statements about MoviePass, including the claim that the service could be profitable at 9.95 dollars a month. Farnsworth consented to injunctive relief. The release says the litigation against the other defendants was continuing. A consent judgment is not a trial verdict, and this article does not invent an admission the release does not quote. It also does not say that Lowe had been found liable. The fact a founder can use is narrower, and it is enough. The company’s own 10-K had already warned that more viewing increased losses per subscriber, and the SEC later treated public claims of profitability at that price as a serious disclosure problem.

The lesson is not “subscriptions are bad”, and it is not “regulators will come for your seed-stage pitch”. MoviePass was a public-company story with a debit card, a cinema ticket and a national chain on the other side of each visit. Most early software companies do not have that cost curve. They do have the same temptation. A usage chart is easy to celebrate. A sentence that says “we lose money on each additional customer, and we have not yet secured the discount that would change that” is harder, and it is the sentence a buyer of the company’s story needed.

Contribution is the test, and it is not a 10-K line

The 10-K does not present a contribution-margin statement, and this article will not pretend that it does. The idea is older and simpler than the filing. In managerial accounting, contribution is the price of a unit minus the costs that rise because you sold and delivered that unit. What remains is what you have left to put toward standing costs and, only after those are covered, profit. A standard explanation of that split is enough. Gross margin, as a financial-statement idea, follows accounting classifications. Contribution follows the economic question of what moves when volume moves. They are not always the same number. Using the word “margin” without saying which one is how teams talk past each other.

The contribution test used in this article
TermMeaning in this article
PriceWhat the customer pays for the unit you are judging
Variable costCost that rises when you deliver one more of that unit
ContributionPrice minus that variable cost, before fixed costs
Fixed costCost that does not rise with the unit, inside the current range of volume

This is a definition, not a MoviePass disclosure and not a chart of market data. Contribution here means price minus the cost that rises when one more customer is served. Fixed costs sit outside that line. A conventional income statement does not have to present the split this way. Source: Contribution margin, as explained by Wikipedia.

Apply the definition to the mechanism the 10-K describes, without inventing a ticket price. The subscriber paid a monthly fee. The filing’s February 2018 description puts that fee as low as 9.95 dollars on the plan that drove the growth. Each film the subscriber actually attended caused MoviePass, in most cases, to pay the cinema the full ticket price. The variable cost of a viewing was therefore a whole ticket, not a fraction of one, unless a particular exhibitor had agreed otherwise. The 10-K says the large chains historically had not, and that even the smaller discounts it had sometimes received were not the general case. If one ticket already costs more than the monthly fee, a second viewing in the same month cannot repair the month. That is an inference from the structure the company disclosed. It is not a calculated average loss, because the filing passages used here do not state the average ticket paid or the average number of films watched. Anyone who fills those cells with a remembered industry price is producing a different document.

Fixed costs still matter, and they are a separate problem. Support, the card, the fraud controls, the corporate overhead and the cost of acquiring the subscriber do not disappear because contribution on a viewing is negative. They make the hole deeper. They are not required to prove the point. A negative contribution does not become acceptable because the team hopes to “make it up on scale”. There is nothing to make up. Each extra viewing adds a loss. Scale multiplies it. The only structural repairs are a higher price, a lower ticket cost, a limit on viewing, or a new payment from someone else, such as the concession share the 10-K says the company hoped to negotiate and did not historically receive from the large chains. Those are different companies. They have to be tested as different companies.

A worked illustration, labelled as fiction

The clinic below is not MoviePass and not a research result. The numbers are chosen so the arithmetic is visible. Suppose a software service charges 30 pounds a month. Hosting, payment fees and the support that genuinely rises with each extra account come to 8 pounds. Contribution is 22 pounds before rent, salaries and acquisition. You can then ask a separate question: how many accounts at 22 pounds cover the standing costs you actually have. That question can have an honest answer, including “not at this price”.

Now change one fact. The promise is not software. It is that a person will do four hours of work for each customer every month, and that work costs 40 pounds. Contribution is minus 10 pounds before you have paid the rent. No volume target repairs it. The tenth customer is not cheaper in this illustration, because the cost was defined as varying with the customer. A founder in this position has four real levers, and they are the same levers as in the filing, translated out of the cinema. Raise the price. Cut the variable work. Cap the promise. Or get a third party to pay. Hoping that “brand” or “data” will one day pay the 10 pounds is an inference. It belongs in the inference column of the decision memo, not in the cash column.

If you do not know which of your costs vary, you are not ready to claim a margin. The first commercial proof is often a list, built from the last ten deliveries, of what actually consumed money or founder hours. Founder hours count if the model assumes they will later be replaced by paid staff. Leaving them out creates a margin that only exists while the founder is the unpaid cost centre. That is a legitimate way to learn. It is not the margin of the company you are describing to an investor.

Who benefits is not always who pays

Usage is ambiguous until you name the person who pays. A practitioner can love a scheduling tool while the clinic owner, who holds the budget, sees no reason to switch. A subscriber can love unlimited cinema while the cinema, which holds the ticket inventory, has no reason to discount. MoviePass’s 10-K is unusually clear on that second gap. The company was sending customers through the door and, by its own account, not receiving a share of what those customers then spent on concessions. The party that benefited from the extra visit was not the party funding the ticket.

Map four roles before you read a usage chart as demand. The user, who gets the job done. The buyer, who can say yes to a price. The approver, who can veto the buyer. The party that bears the cost if the promise is kept, which may be you. If two of those roles are the same person, the test is simpler. If they are not, a successful user test can coexist with a failed company. The next experiment should be aimed at the missing role, not at more of the usage you already have.

Price is part of that map, not a coat of paint at the end. A discount can be a deliberate way to buy a first observation, if you write down what it changed. A free pilot does not demonstrate the price you hope to charge later. A deposit does more. A paid renewal does more than a deposit. A renewal at the undiscounted price, without the founder managing the account, is the observation most teams are actually relying on when they talk about viability. If you do not have it, say you do not have it. Then decide whether the next month is for getting it, or for changing the offer until it is gettable.

The companion question, what counts as a real commitment rather than a warm conversation, is the subject of Compliments are not a purchase order. The prior question, which proof the company should chase before it spends, is What is the company trying to prove next?. Viability is often the proof that should come before a sales hire or a raise, not after.

What to do before you buy more volume

Take the last ten customers, or the last ten uses if you do not yet have customers, and write the variable cost beside the price. Include the delivery work, the refunds, the payment fees and the founder time you would have to replace. If you cannot get the list to a reasonable standard, that gap is the experiment. Do not fill it with an industry average you found in a slide. An average that is not your cost is someone else’s company.

Then ask which single change would flip the sign of contribution, and whether you have any evidence that the change is available. For MoviePass, as the 10-K describes it, the candidates were a higher subscription price, a limit on films, a real ticket discount, or a concession share. The filing shows the company hoping for the partnership terms and warning about the consequences if they did not come. A founder in a similar structure should not pitch the hope as if it were the present. Pitch the present, and describe the hope as a hypothesis with a date.

Set a threshold that can stop spend. An example, and it is only an example of a rule rather than a benchmark: do not increase acquisition spend while contribution on the next customer is negative, unless an identified third party has already contracted to cover the gap. “We will be fine at ten times the volume” is not a threshold. It is a wish that volume will change a cost that, by definition, rides with volume. Acquisition cost and retention still matter once contribution is positive. They cannot rescue a negative contribution. Confusing the two is how a growth plan becomes a cash plan written by the advertising budget.

Capital has a place. The 10-K’s financing lines show a company raising money while the operating model consumed cash. Raising can be rational for a milestone that changes the model, such as a contracted discount or a rebuilt promise that removes the variable loss. Raising to buy more of the same negative contribution is a bet that someone else will keep funding the gap. The going-concern note is what that bet looks like when it is disclosed properly: substantial doubt, an explicit funding need, and no assurance that the plan will work. Early-stage companies do not file 10-Ks. They can still write the equivalent paragraph for themselves before they ask anyone else to fund the gap.

If you want a second person in the room for that paragraph, the commercial stage of the startup journey is the checklist, and a Focused Session is the place to bring the price, the last deliveries, and the spend you are about to approve. The useful result is not a slogan that the product is promising. It is a contribution you can recompute, a missing contract you can name, and a decision to pause volume if the next customer still makes the company poorer.

Sources

  1. Helios and Matheson Analytics Inc., Form 10-K for the year ended 31 December 2017, filed 17 April 2018 Risk factors on full ticket cost, and the company’s account of the MoviePass acquisition and later cash advances.
  2. Helios and Matheson Analytics Inc., going-concern note in the same 10-K Reported loss, operating cash outflow, cash, working-capital deficit and the substantial-doubt statement.
  3. Helios and Matheson Analytics Inc., filing of 12 February 2018 Company description of the 9.95 dollar plan and company-reported subscriber counts. The peer comparison in that filing is the company’s own estimate.
  4. US Securities and Exchange Commission, Litigation Release No. 25525, 27 September 2022 The SEC’s complaint, including the allegation that profitability at 9.95 dollars a month was misstated. Allegations are not a trial verdict.
  5. US Securities and Exchange Commission, Litigation Release No. 26207, 2 January 2025 Consent judgment against Theodore J. Farnsworth. The release says litigation against the other defendants was continuing.
  6. Contribution margin, Wikipedia A plain explanation of the managerial definition used in this article: price minus the costs that vary with the unit. It is not a line the 10-K was required to present.

pricing · business models