Skip to content

Valuing a media partnership when no money changes hands

Acquisition and attributionUpdated 2026-08-188 min read

In short

Value a barter partnership as two separate prices. Price what you give away at realised rates multiplied by the chance the space would have sold, and price what you receive at what you would genuinely have paid for the same inventory, with a zero against anything you would never have bought.

The trade title wants nine square metres in hall two, four delegate passes and a logo on the visitor site. In return you get six newsletter slots, a quarter of a million banner impressions and two posts to their audience. Nobody invoices anybody, everybody shakes hands, and the deal never appears in a cost per registration calculation.

Valuing a media partnership is the work of putting a price on both halves of that trade separately, and it matters because the alternative is a channel that produces registrations at an apparent cost of zero, which will beat every paid line in your reporting forever.

Why does a barter deal need two prices?

A barter deal is two transactions that happen to net to no cash. You sell something. You buy something. The fact that the two settle against each other does not make either one free.

Finance has understood this for a long time. The International Accounting Standards Board issued SIC-31, Revenue: Barter Transactions Involving Advertising Services, in December 2001, specifically to determine when a seller of advertising can reliably measure revenue at the fair value of the advertising received in a barter transaction. IFRS 15, effective for annual reporting periods beginning on or after 1 January 2018, is the current revenue standard that treatment sits under. The accounting question of what may be recognised as revenue is separate from the marketing question this post answers, and it is a useful reminder that a professional discipline has already decided you cannot call a barter zero on both sides.

So run two valuations. What did the outbound half cost you, and what would the inbound half have cost if you had bought it.

What does the space you gave away actually cost you?

Start with the booth. Nine square metres at a 700 rate card is 6,300, and that number is fiction, because nobody pays rate card.

Take the realised rate instead. If the average realised rate in that hall last edition was 560 per square metre, the nine metres are worth 5,040 in revenue terms.

Then apply the question that decides everything: would that space have sold? If the hall closed at 96 per cent occupancy with a waiting list, the space had a buyer and you gave away 5,040 of revenue. If it closed at 82 per cent, the marginal nine metres had no buyer and giving it away cost you the servicing, not the revenue. Most halls sit between those cases, so use a probability. At a three in four chance of selling, the expected cost is 0.75 multiplied by 5,040, which is 3,780.

The passes need the same treatment and usually a different answer. Four delegate passes at a 495 rate card is 1,980, but the realised average paid rate is 340, and two of the four are going to journalists your press policy would have comped anyway. So the revenue forgone is two passes at 340, which is 680. Add the marginal servicing cost of four attendees, say 28 each for badge, catering and materials, which is 112. The passes cost 792.

Outbound total: 3,780 plus 792, which is 4,572. The rate card version of the same package is 8,280, so the honest number is 55 per cent of the number a partner would quote back at you.

Valuing the inbound side at what you would have paid

The temptation on the inbound half is to accept the partner's rate card, which for six newsletter slots at 1,400, 250,000 impressions at a 32 CPM and two social posts at 900 comes to 18,200. That number belongs in no calculation you will ever defend.

Price it instead at what you would genuinely have paid, with a hard rule that anything you would never have bought is worth zero.

Six newsletter slots you would have bought, and you would have negotiated. At 35 per cent off the card, that is 5,460. The 250,000 impressions you would have bought programmatically at your achieved 12 CPM, which is 3,000. The two social posts you would not have bought at all, so they are worth nothing, however nice they are to have. Inbound total: 8,460.

Cross-check that against the outcome the inventory should produce. If comparable placements last edition delivered around 343 registrations for this volume of exposure, and your marginal paid cost per registration is 31, the outcome route values the same package at 10,633. The inventory route and the outcome route disagreeing by a fifth is normal. The two disagreeing by a factor of three means one of them is wrong, and it is worth finding out which before you renew.

Use your own measured cost per registration for that cross-check, never a published average. Sopro's 2025 update to its B2B cost per lead benchmarks puts referrals at an average of 25 dollars per lead, affiliate marketing at 73, cold email at 225, paid search at 463, and trade shows and events at an average of 840 with a range running from 180 to over 1,500. A spread of that size across channels, and inside a single channel, is the argument for measuring your own number rather than importing one.

Working both halves of one deal

Put the two sides together and the deal reads like this.

Outbound cost: 4,572. Inbound value: 8,460. Net position in your favour: 3,888.

Now convert it into the figure your reporting can use. The partnership delivered 343 registrations, and it cost you 4,572 in forgone revenue and servicing, so the cost per registration is 13.33. Against a marginal paid cost per registration of 31, this partnership is a strong channel and should be renewed and enlarged.

Two adjustments before that goes in a deck. The first is the new to file share, because registrations from a partner audience that already sits in your database are worth less than the arithmetic suggests, and the method for splitting that out is measuring overlap against your existing file. The second is that the 343 has to be a union of code redemptions and tagged registrations rather than either alone, which needs one code per placement and a matching source tag.

The partner should be running the mirror image of this calculation. Their outbound is the advertising inventory they gave you, priced at what they would really have sold it for, which for unsold newsletter inventory in a quiet month is well below card. Their inbound is your booth and passes, priced at what those are worth to them as a sales tool. When both sides do the work, the negotiation gets shorter, because most disputes in barter renewals come from each side valuing its own contribution at card and the other's at cost.

Recording the deal so it appears in reporting

A valuation nobody can find is a valuation that will be redone from memory next year, differently.

Raise the barter as a zero-cash order in whatever system holds your media plan, with the agreed inbound value in the amount field and a note naming the outbound package. Two things follow. The partnership appears in every channel report next to the paid lines instead of hiding underneath them, and the finance lead who has to explain the marketing spend has a document rather than an anecdote.

Give the outbound half the same treatment on the sales side. Nine square metres booked to the title at a stated value, flagged as barter, keeps hall occupancy honest and stops the same space being sold twice in a busy week. It also means that when somebody asks why net square feet sold is up while revenue is flat, the answer is available in a query instead of a conversation.

Keep both entries at the values you calculated, not at rate card. The moment a barter is recorded at card on either side, the numbers stop being usable for anything except negotiation, which is the job they were doing before you started.

Where this stops

The probability of selling the space is a judgement, and it carries the whole outbound valuation. A show director who believes every square metre would have sold produces a number nearly twice that of one who thinks three quarters would, from identical facts. Write the assumption next to the number, use the same figure for every barter deal in an edition, and revisit it once occupancy is known.

The second limit is that the inbound valuation only counts registrations. A trade title's endorsement also moves exhibitors, speakers and the general sense that the show is the one to be at, none of which appear in a registration count, and some partnerships are carried for exactly those reasons. That is a defensible commercial choice as long as somebody says out loud that it is being made on judgement.

The third is timing. Space given away in November is priced against an occupancy outcome you will not know until the following autumn, so every barter valuation made during the selling year is provisional. Recalculate at close and keep the two versions.

There is also a definitional trap. Whether the forgone revenue from barter belongs in the numerator of your cost per registration at all depends on how that metric is defined in your organisation, and getting a consistent answer across cash and non-cash channels is part of deciding what enters a blended cost per registration.

The first step this week is to list every barter arrangement running for the current edition and put two columns next to each one: the realised value of what you gave, and the price you would have paid for what you got. Most portfolios discover between four and a dozen live deals that nobody has ever priced, and the total sitting in that column is usually larger than any single line in the paid media plan, which makes it the fastest correction available to your acquisition and attribution reporting.

Questions people ask about valuing a media partnership

How do you value booth space given to a media partner?
Take the realised rate per square metre from the last edition rather than the rate card, then multiply by the probability that space would have sold. Nine square metres at a realised 560 in a hall that closed at 96 per cent occupancy is worth about 3,780 once you apply a three in four chance of selling it.
Should a barter deal be valued at rate card?
No, and both sides will inflate if it is. Rate card is a starting price nobody pays, so a rate card comparison usually shows both partners winning, which is the sign that neither valuation is real. Use realised prices on what you give and the price you would actually have negotiated on what you receive.
What cost per registration should be used to price inbound advertising?
Your own measured figure for the channel that would have delivered the same registrations, never a published benchmark. Cost per lead varies enormously by channel, with Sopro's 2025 update putting referrals at 25 dollars and trade shows and events at an average of 840, so an external average carries no information about your show.

Related reading

All attribution articles