I go to a lot of events. I cover them. I organize them. I speak at the em. I attend them. For any given event, easily the most common question is how many people attend. It’s how we get a sense of how popular (which is a clumsy shorthand for how valuable something is) the event was. But it’s the wrong question and, I’ve found, almost always a lie.
Because it’s so damn hard. Think about the challenge of estimating attendance at large-scale public events. We always have our reporters estimate attendee counts and often have organizers challenge us. Once an event stretches beyond even just a few dozen people, there’s no sure thing that anyone there will have a good sense of the attendee count. People will have a perceived sense of the crowd — was the event well attended or not — but that has very little to do with actual account and more to do with how full an event location is, among other biases and perspectives. Give me the right number of chairs, and I’ll make your 20-person event crowded.
It’s become second nature for me to hand count attendance at smaller events and do batch counting for larger ones (gauge what a group of 100 looks like and then estimate from there). So I read other event estimates with heavy skepticism.
A friend of mine organizes a popular monthly tech demo breakfast event series. I’ve been to a dozen of these, and they have always had great attendance, ranging between 100-120 people. But somehow a news report gives a 700-person estimate. It’s just not true, but it sure changes how 100 people sounds to you. Now 100-people would seem like a poor showing.
I got this lesson a few years ago.
Leading into the second Philly Tech Week, an annual event series I help organize, back in April 2012, my team and I caught an interesting turn of phrase in an Inquirer story that compared our open calendar of tech community events with the Philadelphia Science Festival, organized by the Franklin Institute museum along with university and other institutional ties. The premise of the story was whether it was good that both of our events — with some cross-over at the institutional level but not necessarily with our core audience — were happening at the same time.
Both of our efforts started in 2011. When we were asked by the Inquirer how many people attended the first Philly Tech Week, we counted up the attendees who came to our events and got a fine 4,000 figure. That seemed impressive to us. Even still, bigger isn’t better, we thought, so the number didn’t much matter. It just helped with a sense of scale.
But then in the same story we saw the figure our friends at the Science Festival shared: that in their first year they had “connected” with 100,000 people, which is just an astounding total. The Science Festival had an outdoor carnival that could have had 10,000 people come out — if given wild discretion on the total. They also branded a Penn robot, which threw the first pitch out at a Phillies game (now that’s cool). The museum itself also had its doors open during the festival, and we heard that the number of people who visited the century-old museum was included too.
So it became increasingly clear to me that those numbers are endlessly mushy.
And though those counts are helpful for context, they don’t actually say much more than that, particularly when a given event series is meant to do a different thing than another. Philly Tech Week is aimed to be a community tool for a sector to celebrate and welcome others. The Philadelphia Science Festival wants to get science in the hands of a more general interest user. Both are valuable. Both do good work.
But reporters are better at comparing big overall numbers than on-the-ground goals. 100,000 is sure a lot bigger than 4,000, even though it’s clear those numbers represent very different people. Everyone in that 4,000 count likely knew they were in that count. I’m not sure if I’d say the same about what I assume is in that 100,000 number.
Because we’ve been turned off by the marketing aims around these types of events, we’ve tried to be as true as we can whenever we’re asked about attendance at something like Philly Tech Week. No bloat from our end.
To do so, we still do what we did that first year. We are conservative on the attendee counts for all of the individual events on the calendar, and, even still, we’re careful about quoting the overall number of people who attended the series, as opposed to event visits — knowing some people went to more events than even our conservative numbers could cover for. We also use estimates from actual attendees and event organizers, in addition to our own staff counts to vet accuracy.
So we’ll say more than 20,000 people attended Philly Tech Week 2014 (that’s in fact lower than what we had estimated might show up, having said 25,000) but are open that despite our conservative numbers, anyone who attended more than 2-3 events was probably double counted. It’s the best number we can come up with.
The most complicated events are those free, public ones. Our two largest events in 2014 were the pair around Cira Centre Tetris and our closing Signature Event. The latter was an easier figure to arrive at, between tickets sold and sponsors who checked in. But for our kickoff night, which included our Arcade at the Oval event, we coupled two conservative numbers: the 2,500 attendees of the actual event by counts from the city Parks and Rec department and early, plus the viewing groups we saw pictures of at the Art Museum, the Parkway, near Drexel and other locations.
Could we have included every driver of I-76 who honked at the Tetris game? Yes, after all the media coverage of the night was epic. But we didn’t. We estimated 5,000 people opted in to watch and engage, not just glimpse (or walk by).
We could all do ourselves a favor by knowing this. The numbers don’t actually matter. What matters is the value provided, the experience had and, in truth, the energy felt. We had more valuable, “crowded” 75-person events than we did events that brought more than 1,000 people together. So in the end, what matters more is what comes from them.