Shifting Traffic Patterns

Early on in the course of ASK, Shelley and I noticed some really interesting patterns related to where people tended to use the app. While most of our visitors come to the Museum for special exhibitions, the majority of app traffic was happening in the permanent collection. This was a pleasant surprise. Like many museums, we are always looking for ways to invigorate the permanent collection and to get people exploring more of the Museum. ASK did this naturally. However, we’ve consistently struggled with how to get more special exhibition visitors (again, the bulk of our foot traffic) to use ASK there as well.

One of the metrics we can pull from the dashboard is the average number of “beacon groups” from which a conversation takes place and is an indication of how much a user is exploring the Museum. Here you can see the conversation from the top “explorer” during this particular time period used 21 beacon groups over 2 hours.

To be clear, we can’t track anyone’s path through the Museum—that functionality does not exist in our system—so we don’t know in what order they visit the galleries (without combing each and every conversation) or where they are when they’re not actively chatting with us (and therefore what galleries they visit, but do not use ASK). But we do know the number of beacon groups a user visits, which roughly equates to number of gallery areas they traverse. We use this number to extrapolate how much and where ASK users are exploring. Since most users averaged 2 beacon areas, we designated users who visit 3 or more beacon areas as “power users.” Below are some metrics about beacons that tell an interesting story:

Two things stand out to me when I look at these numbers. The first is that exploration almost doubled for some reason starting in June of last year. Initially I thought this might be due to the availability of ASK on Android devices, but that happened in April and there was no discernable change in April or May. The major summer shows (The Rise of Sneaker Culture and Faile) didn’t open until July, so I can’t turn to a particular audience shift due to exhibitions. One can, of course, offer credit to the ASK team for continuously improving their engagement practices and encouraging people to explore, but I can’t say if that would account for a such a discernable shift between May and June—a trend that continued until recently. (However, the sustained nature of the trend could be explained by a shift in engagement practices.) Unfortunately, the reason(s) for this shift remain a mystery to me and I am unable to go back and enumerate all the potential variables that might explain this behavior.

The second is that I see a similarly discernable shift happening right now between February and March of this year. Now, I’m usually the first to say that a single month does not make a trend, but last time we had such a shift I missed the opportunity to try to explain it. I’d prefer not to miss it a second time. I’m deep in metrics lately due to the recent launch of our ASK Ambassador program (more on that in a later post) and I’m tracking how this team of “guerrilla marketers” is effecting app traffic on a daily and weekly basis, which is how I caught the shift this time around. Interestingly, this most recent shift coincides with the opening of our stellar Georgia O’Keeffe: Living Modern exhibition.

Last week, Jessica wrote about the surprising kinds of questions we’re getting via ASK within the O’Keeffe exhibition, but that’s not our only surprise. For the month of March, we’ve seen a complete shift in traffic patterns via the app. Not only are users not exploring quite as much, but we are averaging the same number of chats per week with a completely different beacon group makeup. Special exhibitions are starting to make up almost half of the incoming chats, with the O’Keeffe show comprising almost a full third of chats and occasionally more than that.

Since O’Keeffe visitors comprise such a large percentage of app traffic right now, if they aren’t exploring, this would account for the drop in power users. This potential lack of exploration on the part of O’Keeffe visitors can be explained, I think, due to the fact that the show requires a special ticket. I think visitors spend all their mental energy and time in that show and once done, might visit the cafe, restaurant or shop, but otherwise leave the Museum instead of exploring. I don’t have metrics to determine if this is a ticketed show phenomenon or peculiar to this show because the last two ticketed exhibitions took place before ASK existed.

What we are also seeing is that only a small percentage of visitors to O’Keeffe are using the app. While this is in line with previous trends, increasing use by this audience is something we’ll be focusing on in the coming weeks. Well over half of visitors are coming for O’Keeffe and we need a bigger piece of that action. Needless to say, I’ll be keeping a sharp eye on the metrics in the coming weeks to see if I can determine other causes for this traffic pattern shift. If it’s peculiar to O’Keeffe, I would expect the traffic pattern to return to “normal” after this exhibition closes in the summer. Hopefully I won’t have to wait that long for some answers!