IFRS 15 by Industry: How Revenue Recognition Actually Works in Telecoms, Software, Retail, Manufacturing, Professional Services, Pharmaceuticals and Construction

"At contract inception, an entity shall assess the goods or services promised in a contract with a customer and shall identify as a performance obligation each promise to transfer to the customer either: (a) a good or service (or a bundle of goods or services) that is distinct; or (b) a series of distinct goods or services that are substantially the same and that have the same pattern of transfer to the customer (see paragraph 23)."

Nothing in that sentence refers to an industry. It refers to promises. The unit of account in IFRS 15 is the promise, and a sector's revenue profile is simply the aggregate of the promises its contracts contain. A mobile operator's contract contains a device promise and a service promise. A construction contract typically contains one integrated promise. A retailer's contract contains one promise and an option. That is the entire explanation for why their revenue patterns look nothing like each other.

The second limb, IFRS 15.22(b), is the one that carries more sector weight than any other single sentence in the standard. It is why an outsourcing contract, a managed service, a cleaning contract and a monthly airtime plan are usually accounted for as one performance obligation rather than 730 daily ones, and it is the gateway to allocating variable amounts to individual periods under IFRS 15.85 rather than smearing them across the contract. Anyone building a revenue policy for a recurring-service business who has not formed a documented view on IFRS 15.22(b) has skipped a step.

"A good or service that is promised to a customer is distinct if both of the following criteria are met: (a) the customer can benefit from the good or service either on its own or together with other resources that are readily available to the customer (ie the good or service is capable of being distinct); and (b) the entity's promise to transfer the good or service to the customer is separately identifiable from other promises in the contract (ie the promise to transfer the good or service is distinct within the context of the contract)."

A handset passes both limbs comfortably. It is capable of being distinct because the customer can use it on another network, resell it or use it on wi-fi, and because the operator sells the same device separately. It is separately identifiable because none of the IFRS 15.29 factors apply: the operator provides no significant integration service, the handset does not modify or customise the airtime, and neither promise is highly interdependent on the other in the sense IFRS 15.29(c) contemplates. The airtime works on any compatible handset and the handset works on any network.

Note that the conclusion does not depend on the handset being paid for separately, on it being priced at zero, or on the customer being locked in for 24 months. Justifying a single performance obligation on the basis that the customer cannot leave is the error catalogued as F-05 in the audit literature and it has no support in IFRS 15.27. Lock-in is a term of the contract, not a test of distinctness.

"The best evidence of a stand-alone selling price is the observable price of a good or service when the entity sells that good or service separately in similar circumstances and to similar customers. A contractually stated price or a list price for a good or service may be (but shall not be presumed to be) the stand-alone selling price of that good or service."

For an operator, both stand-alone selling prices are usually observable, which removes most of the estimation difficulty. The handset is sold outright in the same shop. The airtime is sold as a SIM-only plan on the same tariff sheet. Those two observable prices are the allocation base, and the contractual "£0 upfront, £30 a month" structure is simply the payment schedule.

The bracketed warning in the second sentence is the operative part. An operator whose systems allocate to the zero handset price stated in the contract has assumed exactly what the paragraph says shall not be presumed. The correct sequence is to determine both stand-alone selling prices independently, sum them, and then prorate the total contract consideration across them under IFRS 15.76 to 15.80.

"The nature of an entity's promise in granting a licence is a promise to provide a right to access the entity's intellectual property if all of the following criteria are met: (a) the contract requires, or the customer reasonably expects, that the entity will undertake activities that significantly affect the intellectual property to which the customer has rights (see paragraphs B59 and B59A); (b) the rights granted by the licence directly expose the customer to any positive or negative effects of the entity's activities identified in paragraph B58(a); and (c) those activities do not result in the transfer of a good or a service to the customer as those activities occur (see paragraph 25)."

All three criteria must be met. That word "all" does most of the work in practice, because criterion (c) is the one that fails most often in software. A vendor that ships updates and new versions to the customer is transferring goods or services as those activities occur, which takes it out of B58(c) and into a separate maintenance performance obligation. That is why so many software licences land on the right-to-use side even though the vendor is plainly still developing the product.

Criterion (a) is where the distinction bites for functional intellectual property. Software that runs standalone and delivers its functionality on day one is generally not significantly affected by the vendor's later activities from the customer's perspective. Media, brands, franchise rights and character licences are the classic B58 cases, because their value genuinely does depend on what the licensor keeps doing.

"If the criteria in paragraph B58 are not met, the nature of an entity's promise is to provide a right to use the entity's intellectual property as that intellectual property exists (in terms of form and functionality) at the point in time at which the licence is granted to the customer. This means that the customer can direct the use of, and obtain substantially all of the remaining benefits from, the licence at the point in time at which the licence transfers. An entity shall account for the promise to provide a right to use the entity's intellectual property as a performance obligation satisfied at a point in time."

Two consequences that surprise preparers. First, a three year term licence with no vendor obligation to change the software is still a point in time transfer, and the full allocated amount is recognised on day one. The term restricts the customer's use; it does not change when control of the licence passes. Second, the same paragraph contains a hard floor: revenue cannot be recognised for a right-to-use licence "before the beginning of the period during which the customer is able to use and benefit from the licence", and IFRS 15.B61 gives the example of a licence period that starts before the vendor has supplied the activation code. Signing early does not accelerate revenue.

"Notwithstanding the requirements in paragraphs 56-59, an entity shall recognise revenue for a sales-based or usage-based royalty promised in exchange for a licence of intellectual property only when (or as) the later of the following events occurs: (a) the subsequent sale or usage occurs; and (b) the performance obligation to which some or all of the sales-based or usage-based royalty has been allocated has been satisfied (or partially satisfied)."

The word "notwithstanding" is the whole point. Where the exception applies, the entity does not estimate the royalty and constrain it under IFRS 15.56. It waits. There is no expected-value forecast of next year's units, no probability weighting and no constraint analysis. Revenue appears as the sales occur.

The scope limit is in the next paragraph. IFRS 15.B63A confines the exception to royalties that relate "only to a licence of intellectual property or when a licence of intellectual property is the predominant item to which the royalty relates". A usage-based fee on a hosted service that is not a licence does not qualify, and neither does a volume-based charge on a manufactured product. Those are ordinary variable consideration under IFRS 15.50 to 15.58, with the constraint fully engaged.

"Factors that indicate that two or more promises to transfer goods or services to a customer are not separately identifiable include, but are not limited to, the following: (a) the entity provides a significant service of integrating the goods or services with other goods or services promised in the contract into a bundle of goods or services that represent the combined output or outputs for which the customer has contracted... (b) one or more of the goods or services significantly modifies or customises, or are significantly modified or customised by, one or more of the other goods or services promised in the contract."

Limb (b) is the software one. Configuration inside the product's designed parameters, using tools the vendor supplies for the purpose, does not significantly modify the software. Writing custom modules, altering core behaviour or building the software into a larger system that the vendor is responsible for delivering does. The distinction is not about the size of the implementation fee or the length of the project. A twelve month deployment of unmodified software is still a distinct licence. A three month project that rewrites the pricing engine may not be.

Note also that the vendor's own identity matters here. Where a third party integrator performs the customisation under a separate contract with the customer, the vendor's licence is distinct even though the customer's overall project involves heavy modification. IFRS 15.29 assesses promises "in the contract", not the customer's programme.

"To account for the transfer of products with a right of return (and for some services that are provided subject to a refund), an entity shall recognise all of the following: (a) revenue for the transferred products in the amount of consideration to which the entity expects to be entitled (therefore, revenue would not be recognised for the products expected to be returned); (b) a refund liability; and (c) an asset (and corresponding adjustment to cost of sales) for its right to recover products from customers on settling the refund liability."

Limb (c) is the one that goes missing. The parenthesis "(and corresponding adjustment to cost of sales)" is not optional colour. It is the instruction that the debit to the return asset comes out of cost of sales, so that the margin reported on the sale reflects only the units the entity expects to keep sold. Omitting it reports a lower gross margin than the transaction actually produced and understates total assets.

The paragraph also settles a question preparers raise regularly. IFRS 15.B22 says the entity's promise "to stand ready to accept a returned product during the return period shall not be accounted for as a performance obligation in addition to the obligation to provide a refund". A return right is a measurement matter, handled through variable consideration, not a separate promise that takes a slice of the transaction price.

"An asset recognised for an entity's right to recover products from a customer on settling a refund liability shall initially be measured by reference to the former carrying amount of the product (for example, inventory) less any expected costs to recover those products (including potential decreases in the value to the entity of returned products)."

Two deductions, and the second is the one that matters commercially. Expected costs to recover covers return shipping, handling and restocking. Potential decreases in value covers the reality that a returned garment is worth less than a new one and a returned electronic item may only be resaleable as refurbished. In fashion and consumer electronics that write-down is often larger than the physical recovery cost.

IFRS 15.B25 also requires the asset to be remeasured at each reporting date for changes in expectations, and IFRS 15.B25 read with the presentation requirement means the asset is shown separately from the refund liability. Netting them off is a presentation error that hides both the gross exposure and the inventory recovery assumption.

"If, in a contract, an entity grants a customer the option to acquire additional goods or services, that option gives rise to a performance obligation in the contract only if the option provides a material right to the customer that it would not receive without entering into that contract (for example, a discount that is incremental to the range of discounts typically given for those goods or services to that class of customer in that geographical area or market). If the option provides a material right to the customer, the customer in effect pays the entity in advance for future goods or services and the entity recognises revenue when those future goods or services are transferred or when the option expires."

The test is incrementality, not generosity. IFRS 15.B41 makes the point from the other direction: an option to buy at a price that reflects the stand-alone selling price is not a material right "even if the option can be exercised only by entering into a previous contract". A discount available to anyone walking in off the street is a marketing offer. A discount earned by having bought before is a material right.

Where a material right exists, IFRS 15.B42 requires the stand-alone selling price of the option to be estimated if it is not directly observable, and that estimate "shall reflect the discount that the customer would obtain when exercising the option", adjusted for the likelihood that the option will be exercised. Breakage on loyalty points is therefore built into the initial allocation rather than recognised as a later windfall.

"If an entity expects to be entitled to a breakage amount in a contract liability, the entity shall recognise the expected breakage amount as revenue in proportion to the pattern of rights exercised by the customer. If an entity does not expect to be entitled to a breakage amount, the entity shall recognise the expected breakage amount as revenue when the likelihood of the customer exercising its remaining rights becomes remote."

Two regimes, and the entity does not choose between them. Where there is sufficient evidence to expect breakage, the proportional method is mandatory and revenue starts appearing from the first redemption. Where there is not, nothing is recognised until the residual rights become remote. IFRS 15.B46 further requires the entity to apply the constraint in IFRS 15.56 to 15.58 in determining whether it expects to be entitled to breakage at all, which sets a genuinely high evidential bar for a new card programme with no redemption history.

IFRS 15.B47 adds a carve-out that catches out retailers operating across jurisdictions: any consideration attributable to unexercised rights that the entity is required to remit to another party, for example under unclaimed property or escheat laws, is a liability and not revenue. A single group card programme can therefore have a breakage policy that produces revenue in one country and a payable in another.

"Indicators that an entity controls the specified good or service before it is transferred to the customer (and is therefore a principal (see paragraph B35)) include, but are not limited to, the following: (a) the entity is primarily responsible for fulfilling the promise to provide the specified good or service... (b) the entity has inventory risk before the specified good or service has been transferred to a customer or after transfer of control to the customer..."

They are indicators, not a scorecard. IFRS 15.B35 sets the principle: the entity is a principal if it controls the specified good or service before that good or service is transferred to the customer. The indicators exist to help evaluate control, and a marketplace that scores two out of three on a checklist has not answered the question. Discretion in establishing the price, the third indicator, is the weakest of the three and is routinely over-weighted by platforms because it is the one they most obviously have.

The specified-good analysis in IFRS 15.B34A comes first and is often skipped. Before asking whether the entity controls the good, identify what the good actually is. On a marketplace the specified good is usually the third party's product, and the platform's own promise is the service of arranging the sale. Two different specified goods, two different answers, in the same transaction.

"An entity transfers control of a good or service over time and, therefore, satisfies a performance obligation and recognises revenue over time, if one of the following criteria is met: (a) the customer simultaneously receives and consumes the benefits provided by the entity's performance as the entity performs (see paragraphs B3-B4); (b) the entity's performance creates or enhances an asset (for example, work in progress) that the customer controls as the asset is created or enhanced (see paragraph B5); or (c) the entity's performance does not create an asset with an alternative use to the entity (see paragraph 36) and the entity has an enforceable right to payment for performance completed to date (see paragraph 37)."

Any one of the three is sufficient, and they are not interchangeable. Criterion (a) fits services consumed as delivered and almost never fits manufacturing, because a half-built machine delivers no benefit. Criterion (b) fits work performed on an asset the customer already owns, for example refurbishment of the customer's own equipment on the customer's own site. Criterion (c) is the manufacturing criterion, and it is the only one with two cumulative limbs.

Naming the criterion is not a formality. An over time conclusion recorded as "the customer benefits as production progresses" is not an IFRS 15 conclusion, because that sentence describes criterion (a) while the facts almost always engage criterion (c). The file has to say which one, because the evidence needed for each is different: (b) needs evidence of customer control of work in progress, (c) needs a contractual restriction or practical limitation plus a termination clause.

"An asset created by an entity's performance does not have an alternative use to an entity if the entity is either restricted contractually from readily directing the asset for another use during the creation or enhancement of that asset or limited practically from readily directing the asset in its completed state for another use. The assessment of whether an asset has an alternative use to the entity is made at contract inception. After contract inception, an entity shall not update the assessment of the alternative use of an asset unless the parties to the contract approve a contract modification that substantively changes the performance obligation."

Two routes to no alternative use, and both are narrower than they first appear. The contractual route requires a genuine restriction on redirecting the asset, not merely a customer's expectation of exclusivity. The practical route is explained in IFRS 15.B6 to B8: a practical limitation exists where the entity would incur significant economic losses to redirect the asset, either through significant rework cost or by selling it at a significant loss. Highly customised equipment usually qualifies. A product built to a customer's order but from a standard specification usually does not, because it could be sold to the next buyer without loss.

The freeze in the last two sentences is easily missed and has real consequences. The assessment is made once, at inception, and is not revisited as the build progresses. A manufacturer cannot start on a point in time basis and switch to over time when the asset becomes visibly customer-specific halfway through.

"In accordance with paragraph 37, an entity has a right to payment for performance completed to date if the entity would be entitled to an amount that at least compensates the entity for its performance completed to date in the event that the customer or another party terminates the contract for reasons other than the entity's failure to perform as promised. An amount that would compensate an entity for performance completed to date would be an amount that approximates the selling price of the goods or services transferred to date (for example, recovery of the costs incurred by an entity in satisfying the performance obligation plus a reasonable profit margin) rather than compensation for only the entity's potential loss of profit if the contract were to be terminated."

This is the limb that most often fails. A termination clause that reimburses costs incurred but no margin does not carry criterion (c), because the standard asks for an amount approximating the selling price of what has been transferred, and costs alone do not approximate a selling price. Equally, a clause that pays a fixed cancellation fee unrelated to progress does not qualify, because it is not compensation for performance completed to date.

IFRS 15.B12 adds a warning that catches out preparers who look at the milestone schedule instead of the termination clause: "the payment schedule specified in a contract does not necessarily indicate whether an entity has an enforceable right to payment for performance completed to date." A generous milestone schedule proves nothing about what happens if the customer walks away in month seven. Look at the termination clause and, where relevant, at whether it is enforceable in the governing jurisdiction.

"In addition to applying the requirements in paragraph 38, for a customer to have obtained control of a product in a bill-and-hold arrangement, all of the following criteria must be met: (a) the reason for the bill-and-hold arrangement must be substantive (for example, the customer has requested the arrangement); (b) the product must be identified separately as belonging to the customer; (c) the product currently must be ready for physical transfer to the customer; and (d) the entity cannot have the ability to use the product or to direct it to another customer."

"In addition to" is the operative phrase. The general control test in IFRS 15.38 must be satisfied first, and then all four of these on top. Bill and hold is therefore harder than an ordinary sale, not easier, which is the opposite of how it is often used at a period end.

Criterion (a) is the one that fails most often. An arrangement entered into because the seller wants the revenue in the current quarter is not substantive. The standard's own examples of substantive reasons in IFRS 15.B79 are the customer's lack of storage space and delays in the customer's production schedule, both of which are the customer's reasons, not the seller's. Criterion (b) rules out goods sitting in undifferentiated finished stock, and criterion (d) rules out anything the manufacturer could reallocate if a better order arrived.

IFRS 15.B82 adds the consequence that is usually forgotten: where revenue is recognised on a bill and hold basis, the entity must consider whether it has remaining performance obligations, for example custodial services, to which part of the transaction price has to be allocated. Storing the customer's goods for six months is a service, and it is not free.

"If a customer has the option to purchase a warranty separately (for example, because the warranty is priced or negotiated separately), the warranty is a distinct service because the entity promises to provide the service to the customer in addition to the product that has the functionality described in the contract. In those circumstances, an entity shall account for the promised warranty as a performance obligation in accordance with paragraphs 22-30 and allocate a portion of the transaction price to that performance obligation in accordance with paragraphs 73-86."

Separately purchasable settles it immediately. If the customer could have bought the warranty on its own, it is a service-type warranty and part of the price belongs to it. No further analysis is required, and the separately observable price is a ready-made stand-alone selling price under IFRS 15.76.

"In assessing whether a warranty provides a customer with a service in addition to the assurance that the product complies with agreed-upon specifications, an entity shall consider factors such as: (a) Whether the warranty is required by law-if the entity is required by law to provide a warranty, the existence of that law indicates that the promised warranty is not a performance obligation because such requirements typically exist to protect customers from the risk of purchasing defective products."

The remaining factors in IFRS 15.B31 are the length of the coverage period, where a longer period indicates a service element because the assurance function does not require an extended term, and the nature of the tasks the entity promises to perform, where tasks such as a return shipping service for a defective product do not add a service element on their own.

Where a warranty contains both elements and they cannot be reasonably separated, IFRS 15.B33 requires them to be accounted for together as a single performance obligation. That is the reverse of the intuitive answer, and it moves the whole warranty into IFRS 15 rather than leaving it in IAS 37.

"For each performance obligation satisfied over time in accordance with paragraphs 35-37, an entity shall recognise revenue over time by measuring the progress towards complete satisfaction of that performance obligation. The objective when measuring progress is to depict an entity's performance in transferring control of goods or services promised to a customer (ie the satisfaction of an entity's performance obligation)."

Note what the objective is not. It is not to match revenue to cost, and it is not to smooth margin. It is to depict performance. IFRS 15.40 then adds a constraint that catches out project accounting systems: "an entity shall apply a single method of measuring progress for each performance obligation satisfied over time and the entity shall apply that method consistently to similar performance obligations and in similar circumstances." One method per obligation. A firm that measures a project by hours in the first phase and by milestones in the second has either identified two performance obligations, or has breached IFRS 15.40.

IFRS 15.44 sets a hard floor that is genuinely useful in a services firm: if an entity cannot reasonably measure the outcome of a performance obligation but expects to recover the costs incurred, revenue is recognised only to the extent of the costs incurred until the outcome can be reasonably measured. That is the correct treatment for the opening weeks of a genuinely novel engagement, and it is a zero-margin answer, not a nil-revenue one.

"Input methods recognise revenue on the basis of the entity's efforts or inputs to the satisfaction of a performance obligation (for example, resources consumed, labour hours expended, costs incurred, time elapsed or machine hours used) relative to the total expected inputs to the satisfaction of that performance obligation. If the entity's efforts or inputs are expended evenly throughout the performance period, it may be appropriate for the entity to recognise revenue on a straight-line basis."

IFRS 15.B15 defines the alternative: output methods "recognise revenue on the basis of direct measurements of the value to the customer of the goods or services transferred to date relative to the remaining goods or services promised under the contract", with examples including "surveys of performance completed to date, appraisals of results achieved, milestones reached, time elapsed and units produced or units delivered".

Neither is preferred by the standard. IFRS 15.B19 records the shortcoming of input methods, that there may not be a direct relationship between inputs and the transfer of control, and IFRS 15.B17 records the shortcoming of output methods, that the outputs may not be directly observable without undue cost. In professional services the practical position is usually the reverse of construction: hours are precisely measured and outputs are not, so input methods dominate.

"As a practical expedient, if an entity has a right to consideration from a customer in an amount that corresponds directly with the value to the customer of the entity's performance completed to date (for example, a service contract in which an entity bills a fixed amount for each hour of service provided), the entity may recognise revenue in the amount to which the entity has a right to invoice."

This is the paragraph that makes time and materials accounting straightforward, and it is conditional in a way that is regularly ignored. The condition is that the amount invoiced corresponds directly to the value transferred. A flat hourly rate satisfies it. A rate card that steps down after 500 hours does not, because the later hours are billed at a price that no longer corresponds to their value. Nor does a contract with a significant upfront mobilisation fee, a back-loaded success element, or an annual rate escalation unconnected to delivery.

The expedient also has a disclosure consequence that firms miss. IFRS 15.121 permits an entity that applies the IFRS 15.B16 expedient not to disclose the transaction price allocated to remaining performance obligations, but IFRS 15.122 requires the entity to disclose that it has applied the expedient. Using it silently is a disclosure omission.

"An entity shall estimate an amount of variable consideration by using either of the following methods, depending on which method the entity expects to better predict the amount of consideration to which it will be entitled: (a) The expected value-the expected value is the sum of probability-weighted amounts in a range of possible consideration amounts. An expected value may be an appropriate estimate of the amount of variable consideration if an entity has a large number of contracts with similar characteristics. (b) The most likely amount-the most likely amount is the single most likely amount in a range of possible consideration amounts (ie the single most likely outcome of the contract). The most likely amount may be an appropriate estimate of the amount of variable consideration if the contract has only two possible outcomes (for example, an entity either achieves a performance bonus or does not)."

The choice is not free. IFRS 15.53 selects on predictive quality, and the standard's own signposts point clearly: expected value for large homogeneous populations, most likely amount for binary contracts. A single completion fee on a transaction that either closes or does not is the textbook IFRS 15.53(b) case, and probability-weighting it produces a number that cannot occur.

IFRS 15.54 then requires one method to be applied consistently throughout the contract. Switching from most likely amount to expected value as the deal nears completion is not an update of an estimate. It is a change of method inside a contract, which IFRS 15.54 does not permit.

"An entity shall include in the transaction price some or all of an amount of variable consideration estimated in accordance with paragraph 53 only to the extent that it is highly probable that a significant reversal in the amount of cumulative revenue recognised will not occur when the uncertainty associated with the variable consideration is subsequently resolved."

The constraint applies after the estimate, not instead of it. IFRS 15.57 then lists the factors that increase the likelihood or magnitude of a reversal, including that "the amount of consideration is highly susceptible to factors outside the entity's influence", that the uncertainty is not expected to be resolved for a long period, and that the contract has a large number and broad range of possible consideration amounts. A success fee dependent on a regulator's approval, a third party's financing or a market window is highly susceptible to factors outside the firm's influence, which usually constrains it to nil until the event occurs. The mechanics are set out in the note on variable consideration and the constraint.

"If the consideration promised in a contract includes a variable amount, an entity shall estimate the amount of consideration to which the entity will be entitled in exchange for transferring the promised goods or services to a customer. An amount of consideration can vary because of discounts, rebates, refunds, credits, price concessions, incentives, performance bonuses, penalties or other similar items."

The list is not exhaustive but it covers the entire pharmaceutical gross-to-net bridge. IFRS 15.52 extends it further, making consideration variable where "the customer has a valid expectation arising from an entity's customary business practices, published policies or specific statements that the entity will accept an amount of consideration that is less than the price stated in the contract", which captures price concessions that have never been written down anywhere.

The measurement consequence is that revenue is never the invoiced amount. It is the estimate of what will be kept, made at the point of transfer, and updated at every reporting date under IFRS 15.59.

"The most likely amount-the most likely amount is the single most likely amount in a range of possible consideration amounts (ie the single most likely outcome of the contract). The most likely amount may be an appropriate estimate of the amount of variable consideration if the contract has only two possible outcomes (for example, an entity either achieves a performance bonus or does not)."

A development milestone is binary. The trial reads out or it does not, the regulator approves or it does not, and the payment is a fixed amount either way. That is the IFRS 15.53(b) case, and the estimate before the event is either the full milestone or nil, never a fraction of it. The fraction appears at the next step, not this one.

Then IFRS 15.56 applies. A milestone contingent on a regulatory approval is highly susceptible to the judgement or actions of third parties, which IFRS 15.57(a) names expressly as a factor increasing the likelihood of a reversal. In practice that constrains development and regulatory milestones to nil until the event occurs, and the correct disclosure explains that reasoning rather than describing the milestone as merely uncertain.

Commercial milestones behave differently. A payment triggered by cumulative sales exceeding a threshold is a sales-based amount, and where it relates predominantly to a licence of intellectual property, IFRS 15.B63 takes it out of the constraint machinery entirely and defers recognition until the sales occur.

"An entity is an agent if the entity's performance obligation is to arrange for the provision of the specified good or service by another party. An entity that is an agent does not control the specified good or service provided by another party before that good or service is transferred to the customer. When (or as) an entity that is an agent satisfies a performance obligation, the entity recognises revenue in the amount of any fee or commission to which it expects to be entitled in exchange for arranging for the specified goods or services to be provided by the other party."

The genuine gross or net question in life sciences arises in collaboration structures, not distribution. Where two parties co-promote a product and one books all the sales while paying the other a profit share, the recording party must decide whether it controls the specified good before transfer under IFRS 15.B34A and B35, and the non-recording party must decide whether its profit share is revenue from a contract with a customer at all or an outcome of a collaborative arrangement outside the scope of IFRS 15.6. Those two questions have different answers and both need to be documented.

"(c) the entity's performance does not create an asset with an alternative use to the entity (see paragraph 36) and the entity has an enforceable right to payment for performance completed to date (see paragraph 37)."

For a building constructed on the customer's own land, this is not the criterion that applies. Where the customer owns the land, the customer controls the work in progress as it is created, and IFRS 15.35(b) is met directly and more simply. Criterion 35(c) is the route for assets being built on the contractor's premises or on land the contractor controls, and for fabrication that will later be installed.

The point is not academic. The evidence needed for 35(b) is evidence of the customer's control of work in progress, typically title and land ownership. The evidence needed for 35(c) is a contractual restriction or practical limitation under IFRS 15.36 and B6 to B8, plus a termination clause satisfying IFRS 15.37 and B9. A file that concludes "over time" without naming which criterion, and then produces the wrong evidence, has not supported the conclusion at all.

"When a cost incurred is not proportionate to the entity's progress in satisfying the performance obligation. In those circumstances, the best depiction of the entity's performance may be to adjust the input method to recognise revenue only to the extent of that cost incurred. For example, a faithful depiction of an entity's performance might be to recognise revenue at an amount equal to the cost of a good used to satisfy a performance obligation if the entity expects at contract inception that all of the following conditions would be met: (i) the good is not distinct; (ii) the customer is expected to obtain control of the good significantly before receiving services related to the good; (iii) the cost of the transferred good is significant relative to the total expected costs to completely satisfy the performance obligation; and (iv) the entity procures the good from a third party and is not significantly involved in designing and manufacturing the good (but the entity is acting as a principal in accordance with paragraphs B34-B38)."

All four conditions, assessed at contract inception. Where they are met, the answer is zero-margin revenue equal to the cost of the good, and the good is stripped out of both the numerator and the denominator of the progress calculation so that the remaining margin is earned across the remaining work. Condition (iv) is the one that excludes a contractor's own fabricated components: if the entity designed and manufactured the item, the cost does represent its performance.

The same paragraph's opening limb, IFRS 15.B19(a), deals with the other adjustment: costs attributable to significant inefficiencies not reflected in the contract price, such as unexpected wasted materials or rework, are excluded from the measure of progress altogether. Note the difference. Uninstalled materials are recognised at zero margin. Wasted materials are excluded from progress entirely and simply reduce the contract margin.

"An entity shall account for the contract modification as if it were a part of the existing contract if the remaining goods or services are not distinct and, therefore, form part of a single performance obligation that is partially satisfied at the date of the contract modification. The effect that the contract modification has on the transaction price, and on the entity's measure of progress towards complete satisfaction of the performance obligation, is recognised as an adjustment to revenue (either as an increase in or a reduction of revenue) at the date of the contract modification (ie the adjustment to revenue is made on a cumulative catch-up basis)."

This is the paragraph that governs almost every construction variation, because a variation to an integrated construction obligation is by definition not distinct from the remaining work. The consequence is a cumulative catch-up at the date of modification, not a prospective reset, and the catch-up hits revenue in the period the variation is approved even though the work relates to the whole contract.

IFRS 15.18 also settles a question that generates a great deal of argument on site. A modification exists when the parties approve a change that creates or changes enforceable rights and obligations, and a modification "could be approved in writing, by oral agreement or implied by customary business practices". A verbally instructed variation is a modification. What it is not, without more, is a determined price: IFRS 15.19 requires the entity to estimate the change to the transaction price under the variable consideration paragraphs where the scope has been approved but the price has not.