Blog

On Siggraph and other conferences

Resiliency due to relevance Siggraph and other shows are right-sized for the audience they serve. Allow me to elaborate. We at-tend, as visitors, and occasionally as speakers, several conferences a year (maybe as many as 40). Some are company events like IDF, DellWorld, GTC, etc., and most are industry events such as Siggraph, GDC, VRLA, etc. They all have their ...

Robert Dow

Resiliency due to relevance

Siggraph and other shows are right-sized for the audience they serve. Allow me to elaborate. We at-tend, as visitors, and occasionally as speakers, several conferences a year (maybe as many as 40). Some are company events like IDF, DellWorld, GTC, etc., and most are industry events such as Siggraph, GDC, VRLA, etc. They all have their merit, rewards, workload, and stress. 

The cost-effectiveness of conferences is an on-going discussion in every company we work with, as it should be. The parameters used in the evaluation of trade shows and conferences is, however, subjective. What weighting, for example, do you give to “Need to be seen” or “Customer meetings?” So it’s not an exact science and thus requires the wisdom and experience of top management, especially when such large sums of money are involved (it’s estimated the average cost for a position at CES is north of $2 million). 

One of the metrics companies look at in making such evaluations is attendance. It’s a relatively easy number to get because the conferences use it for marketing. However, its validity or verisimilitude is questionable sometimes, and not all (few, actually) are audited, but they are improving due to things like bar codes, which prevent double counting if someone goes out a side door and comes back in the front door. 

We keep track of reported attendance for a few shows that we care about, such as FMX, GDC, GTC, IDF, Laval, and Siggraph. And even though we are successfully recovering CES attendees, we even track its numbers. 

This year at our fifteenth annual Siggraph Press Luncheon, we reported, as we always do, on the attendance of Siggraph and give our estimate, forecast, for the attendance for the current year. Our estimate for this year was 10 to 11K, which would be less than the 14,045 that attended Siggraph LA last year. Looking at the overall trend, the conference has been declining in attendance since 1995, when more than 49,000 people came to Los Angeles for the show. The low point was 11,250 in 2009 in New Orleans, and if our forecast for this year was correct, that would mark a new low point for the venerable conference. However, I am of the opinion that it is an appropriate size for the conference. It is, after all, an academic conference, and it addresses a somewhat esoteric and eclectic group including artists, scientists, animators, and mathematicians, plus some very imaginative students and experimenters. But it’s not a place the suits, big deal makers, and stock market hot shots go (and say a little prayer for that). Therefore, the companies that want to appeal to those folks don’t pay much attention to quirky little Siggraph, dismissing it a show for geeks, movie makers, and game developers—all of which is correct. 

However, the stuff shown at Siggraph, Laval (15K), FMX (8K), GDC (27K), and other like-minded conferences is fascinating, futuristic, and fantastic. The most common expression heard at those shows is, “Oh wow!” Siggraph is the most futuristic of the lot, and the papers and demos presented at Siggraph show up in movies, games, simulators, and other, often surprising, places two or three years later. 

However, with the exception of Laval’s public days, and the grifters who slip into GDC, only serious people attend the pixel-pushing shows, and for that reason, I declared that attendance of 10K to 15K is an appropriate size for such conferences. Therefore, you can imagine my delight when I learned that my forecast for the attendance at Siggraph this year was off—significantly off. Instead of 10K to 11K visitors, Siggraph 2016 in Anaheim drew 14,000 people, just 45 short of how many came to L.A. last year, and L.A. is considered the mecca of venues for Siggraph. The ten-year average for Siggraph is 18K, and the five-year average is 16K, so Anaheim was in a respectable range. 

Most important, not only was attendance up, the number of exhibitors was up to 150, exceeding the 143 in L.A. last year—and that’s where the bulk of the show’s revenue comes from. 

Siggraph (established in 1974) is one of several special interest groups (SIG) within the ACM (Association for Computing Machinery, established 1947), and is the largest SIG within ACM by quite a bit. So if Siggraph sneezes, ACM catches a cold. ACM needn’t worry; Siggraph is quite healthy and robust. I’m happy to report.