Hi folks
We're asking the question (like a lot of you will be no doubt, post-pandemic) - are audiences generally booking later than they used to? We're pretty sure the answer is yes, from watching sales curves etc, but I'd like to be able to put some numbers on this.
So I'm trying to build a widget which will show the average number of days before a performance that people booked per season.
I thought I got what looked like a reasonable result (that people were tending to book a little later in 18-19 and 19-20 than they had in 17-18, then they had to book a lot later during 21-22 as we split our season in two, and put each half on sale very late, but then I get to the 22-23 value and my chart doesn't make any sense, so I'd question whether the earlier season results were correct as well.
We've put the 22-23 season on sale in May 22, and that is for concerts from Sept 22 - June 23. We saw a big rush for onsale, then much slower sales from Jun - Aug 22 onwards, so I would expect this value of average days before performance for orders to be pretty high at the moment - well into the 100s days, but I'm getting a value much lower than that.
No doubt I'm not using the formulas/ breaks correctly and need to group each season's orders together before working out the average for each? - can anyone on here who is great with formulae help me out?
Best wishes
Melanie
I'd come up with a widget to break down the numbers by performance in specific seasons. The distance between May and September is a lot smaller than September and June of the next year: if you're correct and people are not booking as far out, then most of your bookings will be for September, and that will skew your overall average _shorter_.
Thanks Gawain - yes I see what you mean. I also realised I don't fully understand how the Order Days prior to performance works - a lot of our bookers for the 22-23 season will have booked for the whole Sep 22- Jun 23 period in one order, perhaps booking one or two concerts in each month. Does the Order Days prior to performance mean prior to the first performance in that Order I wonder?
I may have cracked it - but would appreciate more experienced Analytics folks correcting me if I've got this wrong!
Certainly this is giving the results I expected to see, as the orders that are in for 22-23 so far are only advance orders, as the season doesn't start until 24 September 22. So the orange line will reduce as we go through the season, and more tickets are bought nearer to the day of each performance. 21-22 was our weird pandemic split season, where we announced each half-season at very short notice each time.
By putting in Ticket Paid Count, it will look at each performance ticket separately, so takes into account the fact that some orders contain multiple performances.
I then had to filter on Ticket Paid Count = 1 to ensure it was only looking at paid tickets (I think...?).
Thanks for reading/ any comments, in advance!
Each ticket has its own value for Order Days Prior to Performance. Multiple tickets with different performance dates in the same order will not skew. Put another way, "Order Days Prior to Performance" is a property of the ticket, not of the order.
We're rapidly leaving my area of expertise, but I'm concerned that this is just filtering out any order where Ticket Paid Count > 1.
looks like you got it. The distribution for a full season is pretty large. From herre you can get pretty funky when you look at segments like super-engaged or subscribers which I find to book really early, covid-hesitant or nervous patrons which book later, Millenials (which are not one size fits all I agree). The distribution of the segmentation can get nice and small.
It can be interesting to craft a different message to those folks that have had their behaviouur changed in 2020+ to late buyers lettin them know of your great refund policies etc to alleviate stress.
Now every time that I answer on the forum I'll be singing "ManSplaining Men" by the Tessiturians
ha, I know the feeling! All the time with Analytics. Every day's a school day, and this forum is so helpful!
The only values that came up to filter on were 0 and 1 so I think the 0 is either to do where the order was refunded so the tix are no longer there, or to do with comps (as I'd specified Paid).
Thanks - good to know!
Thanks Heath - I didn't think you were ManSplaining don't worry - just reassuring a novice Analytics bod :-) My Analytics knowledge comes from nicking things off this forum and trial and error!
Yes it would be interesting to break this down to see which segments have changed their behaviour the most. I'm going to have to wait until the 22-23 season has matured (or a good amount of it) until we can truly tell if booking patterns have changed a lot, as the 21-22 season was such an anomaly.
But perhaps focussing in on similar concerts in different seasons that happened at similar time of year might also give us a good feel for that, as Gawain suggested.
Cheers!
Oh sorry I was pretty high on post conference feels and lack of sleep, typing at a diner in the airport, waiting for my plane. The Tessiturian's played an exceptional song at Conference the previous night - a rendition of Raining Men that Chuck reworded to Mansplaining Men which was hilarious.