The setup
A well-known headliner, one night, a busy spring month with plenty of other things competing for the same evening out. The show was announced roughly two months ahead, and the campaign ran every single day from the announcement to the doors opening.
This one is worth reading precisely because it is an ordinary result rather than a spectacular one. It sat right at the middle of everything we ran that year, which makes it a fair picture of what a normal club night looks like from the inside.
The announcement was the cheapest moment of the whole campaign
In the first two and a half weeks, roughly a sixth of the budget produced a third of everything the campaign would ever sell, at almost four times back on what went in. Those buyers were not persuaded by anything. They were waiting for the announcement and bought as soon as they saw it.
That pool is real, but it is also finite. Once it has been served, every further sale costs more - which is exactly what the rest of the run shows.
Two months, four phases
| Phase | Budget | Sales | Back per euro |
|---|---|---|---|
Announcement wave first 2½ weeks | 16% of budget | 33% of sales | 3.9× |
The flat middle weeks 3 and 4 | 14% of budget | 14% of sales | 1.9× |
Second wind one month out | 18% of budget | 24% of sales | 3.5× |
The final two weeks last two weeks | 50% of budget | 27% of sales | 1.5× |
Read the first and last rows together. Half the money went out in the final two weeks and bought a quarter of the sales. The same money spent in the opening two weeks would have been worth roughly two and a half times more.
The ads were fine. Advertising simply got more expensive.
The usual explanation for a fading campaign is that people got bored of the images. Not here. The ads that ran for more than two weeks performed almost identically in their first week and for the rest of their lives - one of them actually improved with age.
What changed was the price of reaching anyone at all. By the closing week it cost about two and a half times more to put the ad in front of a thousand people than it did at launch, and around six times more to get a click. Nothing about the night, the artist or the ad had changed. The auction around it had.
Where the money went vs where the sales came from
Two thirds of the budget produced nine out of ten sales.
The people who had already looked at the event - visited the page, opened the ticket link, followed along - carried almost the entire campaign.
We ran a test that failed, and we are keeping it in
It sounds obvious that the artist's own fans would be the best people to advertise to, so we built an audience around exactly that and gave it about a fifth of the budget. It returned barely more than it cost, at roughly five times the price per sale of the simpler audience next to it, and the same people kept seeing the ad over and over.
We include it because this is what testing actually looks like. A sensible idea got a fair budget, the numbers said no within a couple of weeks, and now we know not to sell that idea to anyone as a certainty. Without measurement, that same test would have been remembered as a success, because the show still happened and the room still filled.
The mistake we made: we let the budget follow the calendar
As the date approached, we did what most people do - spent more, and opened up to new audiences at the same time. Both instincts pushed in the wrong direction. The extra money arrived exactly when reaching anyone cost the most, and it went disproportionately to the side of the campaign that had never once paid for itself.
Had the same total budget been weighted toward the opening month, with the final weeks held flat unless the cost per sale stayed low, this campaign would have finished noticeably further ahead. Not because of better ads - because of better timing.
What this means for your next event
1. Give the announcement room to prove itself. Most event budgets start cautiously and build toward the date. For this show, the first two weeks were the strongest opportunity, so starting too small left efficient sales on the table. That is not a rule to spend everything early: give the launch enough budget to reveal demand, then let its return decide whether to keep pushing or save money for later buyers.
2. The last two weeks are a harvest, not a push. In this campaign, reaching people became much more expensive near the date. Staying on was still worthwhile because late buyers were real, but spending more simply because the show was close would not have improved the result. Other events can peak late, as the other case studies show. The useful rule is to stay present and let current return, not the countdown alone, decide whether the budget should rise.
3. A flat result usually means the plan, not the poster. When a campaign cools off, the first reflex is to blame the images. Here the ads held their performance for weeks; the conditions around them changed. It is worth knowing which of the two is happening, because refreshing creative and re-timing budget are very different jobs.
4. Give new ideas a budget and a deadline. The fan-targeting test was a reasonable bet that turned out to be wrong, and it only cost us a fifth of the run because we could see it failing. Test things - but decide in advance how much you are willing to spend finding out, and what number would make you stop.
All figures are as measured in the Romanian advertising account for this run. Purchases are checkout events, so one purchase can cover more than one ticket.