Event app sponsorship inventory is easier to count and easier to oversell
Event app sponsorship inventory covers splash screens, banner rotations, sponsored cards, directory upgrades and session tags, all of which write a server-side impression log. The counting is exact and the audience is small: if 2,100 of 9,000 registrants install the app, every unit prices against 2,100 people.
The app sponsor asks how many impressions their banner got. For once you have an exact answer, because the app wrote a row every time the thing rendered, and you can give it to the unit rather than as a range.
Then you look at the number and understand why nobody asked before.
Event app sponsorship inventory is the only part of the non-booth card where measurement is close to free. Every other asset needs an estimate built from traffic assumptions. The app knows. That precision is a commercial asset and a commercial hazard in the same breath, because the first thing precision does is expose how small the audience is.
What is actually in the app inventory?
The sellable surfaces are more numerous than most rate cards admit, and they behave differently enough to need separate lines.
The splash or interstitial on open. Shown on a cold start, sometimes on resume. Highest attention of anything in the app, lowest volume, and the unit most likely to be a single exclusive sponsor.
Rotating banners. Home screen, agenda, exhibitor list. Sold as a share of a rotation, which means the impression count per sponsor depends on how many sponsors are in the rotation.
Sponsored cards in a feed or list. A card placed inline among organic items. Renders only when the user scrolls to it, which makes its delivery far more variable than a banner pinned to the top of a screen.
Upgraded directory entries. Logo, gallery, documents, higher placement. This behaves like a product listing and earns its own treatment in the directory upgrade economics.
Session and track sponsor tags. A sponsor's name attached to a session detail page, delivered to everyone who opens that session.
Map pins and wayfinding placements. Small volume, high intent, and usually mispriced downward because nobody counts them.
The sponsored push sits outside this list. It has a supply constraint that none of the above share, and it belongs with push pricing.
Why the count is exact and the number is generous
Every unit above writes a server-side record when it renders. There is no traffic model, no pass frequency, no visibility factor. Somebody can open the table.
That record is a served impression, and served is a weaker claim than seen. The Media Rating Council set out what the stronger claim requires in its Viewable Ad Impression Measurement Guidelines, version 1.0, released on 30 June 2014 and prepared with the IAB Emerging Innovations Task Force. A display ad counts as viewable when at least 50 per cent of its pixels are in the viewable space on an in-focus browser tab, and when that condition holds for at least one continuous second after the ad renders. The guidelines put the distinction plainly, saying that an opportunity to see the ad exists with a viewable impression, "which may or may not be the case with a served ad impression".
The same document contains the detail that matters most for an event app. It records that under the Mobile Application Advertising Measurement Guidelines issued by IAB, MMA and MRC in July 2013, ad impressions served in an in-application environment were currently generally assumed to be viewable. An assumption, written down as an assumption, in the foundational standard.
So the app's log is exact about rendering and silent about seeing. That is a much better position than a signage estimate, and it is still a number that needs its own qualifier. Call it a served impression on the rate card and in the report, and you have said something true that survives an audit.
The denominator is the install base, and then smaller
Here is where app inventory usually gets oversold, and the failure is arithmetic rather than intent.
A show has 9,000 registrations. The app has 2,100 installs. That is 23.3 per cent, and the 9,000 has nothing to do with anything the app can deliver. Then take installs down again: suppose 1,540 of the 2,100 opened the app at least twice during the show week, which is 73.3 per cent of installs and 17.1 per cent of registrations.
Every unit on the card prices against a number in that region. A sponsor told they are reaching the show's audience, and later told the app has 2,100 users, will do the division themselves and reach a conclusion about the rest of your card too.
Which of those figures is the right base, and how to define an install against an active user without flattering yourself, is a measurement question in its own right that belongs to app adoption. For the rate card the rule is simpler: print the base you priced against, and print it next to the price.
What a banner rotation actually delivers
Do the arithmetic once and the pricing conversation gets easier.
Take the 2,100 installs. Across the show week they generate 6.2 sessions each, which is 13,020 sessions. Suppose 3.4 screens per session carry the home banner, giving 44,268 banner renders in total. The rotation holds four sponsors, so each sponsor gets 11,067 renders.
Price that slot at 3,200 dollars and the CPM is 3,200 divided by 11.067, which is 289.15 dollars per thousand served impressions.
Now apply a viewability haircut for the sake of honesty. Nothing in an event app measures this by default, so treat it as a scenario: if 62 per cent of renders would have met the 50 per cent pixels for one continuous second test, the sponsor got 6,861 viewable impressions and the effective CPM is 466.38 dollars. That is the range the sponsor is really buying inside, and quoting the lower figure without the qualifier is the part that comes back at renewal.
The splash screen behaves differently. If the 2,100 users produce 3.1 cold starts each, that is 6,510 splash impressions, and a 7,500 dollar exclusive works out at 1,152.07 dollars per thousand. Far higher than the banner, and arguably correct, because a full-screen unit before any content loads is closer to a page than to a strip. Whether a sponsor will pay it depends on whether you can show them the count, which you can.
What does honest app inventory pricing look like?
Three changes, none of which need a new system.
Price the unit against active app users for the show week and say so on the card. A line reading "home banner, one of four in rotation, priced against 2,100 active app users" is a stronger sales document than one reading "app banner, reach the full audience", because the first one cannot be dismantled by a sponsor with a calculator.
Guarantee a floor and reconcile against it. Software makes this cheap. Promise 8,000 served impressions, report the actual, and make up any shortfall with an extra rotation week or a credit. A guarantee you can measure is the closest thing to proof this category offers, and it converts the precision from a liability into the reason to buy.
Report served impressions, unique users reached and taps separately. The unique figure is the one that keeps the whole card honest, because 11,067 renders across 1,900 unique users is a frequency of 5.8, and a sponsor who reads only the first number believes they reached eleven thousand people.
There is demand-side evidence that this inventory is worth the effort of getting right. CEIR published the first report in its Exhibit and Sponsorship Sales Approaches series on 9 October 2023, asking organisers which offerings they found effective at generating revenue, and mobile app advertising was rated effective by 72 per cent of them, with exhibition website advertising at 74 per cent. Both sit below email to the registration list at 83 per cent and lanyards at 84 per cent, which is roughly where a well-run app line should be: solid, unspectacular, and steady.
The oversupply trap that software makes easy
Physical inventory limits itself. There are only so many aisle banners before the hall looks like a bus station, and somebody will say so.
An app rotation has no such governor. Adding a fifth sponsor to a four-slot rotation costs nothing to build, takes a morning, and instantly cuts every existing sponsor's delivery by 20 per cent. Do it twice and the sponsors who bought first, at the price justified by the original count, receive 11,067 renders in year one and 7,378 in year two for the same money, with nothing in the contract about it.
The rule I would hold is that any rotation is sold with a stated maximum number of positions, printed on the card, and that number is treated as hard inventory in the same way a lanyard is. If you want more revenue from the surface, add a different surface or raise the price. Diluting a rotation quietly is the fastest way to teach your best sponsors that your app numbers move for commercial reasons.
Where this stops
The precision is real and it is narrow.
A render log tells you the pixels were sent to a device. It cannot tell you the phone was out of a pocket, the screen was on, or the person was looking at it instead of at the queue for coffee. Everything in this post is a served count, and if a sponsor wants attention rather than delivery, they need a survey and a different conversation.
The second limit is comparability. An app impression and a hall banner impression are different units of the same word, and putting them in one column invites a comparison neither survives. If your card carries both, say what each count is derived from.
The third is seasonality of the base. App installs concentrate in the four days before the show and decay fast afterwards, so an annual figure and a show-week figure differ by a lot, and the temptation is to quote whichever is larger. Fix the window, use it every edition, and keep it on the same exhibitor analytics record as everything else you report to sponsors.
Pull last edition's app logs and count three things: installs, users who opened the app during show week, and renders per sponsored unit. Divide the price each sponsor paid by the renders they got. One of those numbers will be indefensible, and it is better to find it before the sponsor does.
Questions people ask about event app sponsorship inventory
- What units make up event app sponsorship inventory?
- A splash or interstitial on app open, a rotating banner on the home and agenda screens, sponsored cards placed in a feed or listing, upgraded exhibitor directory entries, sponsor tags on session detail pages, and map pins. Each writes a server-side record when it renders, so each can be reported as a count instead of an estimate.
- Should app inventory be priced against registrations or app installs?
- Against people who actually opened the app. An install count is closer than a registration count and still generous, because a share of installs never open the app during the show. Price against active users for the show week, state that base on the rate card, and reconcile the delivered figure afterwards.
- Is a served in-app impression the same as a viewed one?
- No. A served impression means the server rendered the unit. The MRC viewability standard for display requires at least half the pixels in view for at least one continuous second, and the 2014 guidelines note that in-app impressions had generally been assumed viewable rather than tested. Report served counts and say that is what they are.
Related reading
- Push notification sponsorship pricing has to price the annoyance as well
- Enhanced directory listing pricing works when you can show the click difference