FOSS4G 2011 Denver Attendance Projections
Attendance Projections
My best "guesstimate" based on looking at each of the top 20 countries resulted in a three tiers:
- Low: 426
- Mid: 839
- High: 1431 (capped at 1200)
"Accurate" attendance projections are extremely important to the early success of a conference the size of FOSS4G.
+ Past FOSS4G data | ||||||
---|---|---|---|---|---|---|
Year | Country | Local | CAN | USA | Other | Total |
2006 | Switzerland | 156 | 37 | 52 | 320 | 565 |
2007 | Canada | 253 | 253 | 265 | 203 | 721 |
2008 | South Africa | 321 | 17 | 31 | 178 | 547 |
2009 | Australia | 218 | 24 | 28 | 166 | 436 |
2010 | Spain | ??? | ??? | ??? | ??? | ??? |
- Overall, FOSS4G is a highly "local" event with at least 57% of attendees from the host region and 25% of attendees from the host country.
Brain Storming
Guesstimate
The formula (below) doesn't seem to be working out. The identified variables have no correlation to the number of attendees. The best solution would be, on a country-by-country basis, try to guess how many would come.
This table gives some "best guesstimates" on a per-country basis with justification. Note that the host country attendance (including the US for 2007) has been changed to the average of the other years.
Country | Population | 2006 | 2007 | 2008 | 2009 | Average | Pessimist | Optimist | Guess | Notes |
---|---|---|---|---|---|---|---|---|---|---|
USA | 307212123 | 52 | 37 | 31 | 28 | 37 | 265 | 800 | 500 | 265 attended '07 in Canada. |
CANADA | 33487208 | 37 | 26 | 17 | 24 | 26 | 26 | 250 | 125 | 253 attended '07 |
SOUTH AFRICA | 49052489 | 2 | 5 | 3 | 2 | 3 | 2 | 10 | 5 | Need to carry energy from '08 |
AUSTRALIA | 21262641 | 4 | 4 | 2 | 3 | 3.25 | 2 | 10 | 5 | Similar to SA, carry energy |
ITALY | 58126212 | 54 | 21 | 17 | 12 | 26 | 12 | 25 | 18 | France, Italy & Germany are strong supporters. How can we increase attendance? |
FRANCE | 64057792 | 43 | 18 | 16 | 9 | 21.5 | 10 | 20 | 15 | |
JAPAN | 127078679 | 8 | 11 | 7 | 15 | 10.25 | 8 | 25 | 18 | Japan is same distance as Europe with large tech population. |
GERMANY | 82329758 | 50 | 11 | 11 | 13 | 21.25 | 11 | 20 | 15 | |
SOUTH KOREA | 48508972 | 0 | 11 | 0 | 17 | 7 | 10 | 25 | 20 | SK is like Japan only hungrier |
UNITED KINGDOM | 61113205 | 20 | 15 | 9 | 6 | 12.5 | 10 | 25 | 15 | Why is UK so under-represented? |
SWITZERLAND | 7604467 | 10.3 | 11 | 17 | 3 | 10.3 | 10 | 20 | 12 | |
SPAIN | 40525002 | 16 | 12 | 12 | 4 | 11 | 12 | 30 | 12 | Should carry energy from 2010 |
NEW ZEALAND | 4213418 | 0 | 3 | 1 | 25 | 7.25 | 1 | 15 | 2 | Definite Kiwi interest. Need to bolster |
NETHERLANDS | 16715999 | 16 | 12 | 0 | 8 | 9 | 8 | 20 | 10 | Why doesn't Dutch like SA? |
NORWAY | 4660539 | 8 | 6 | 10 | 4 | 7 | 4 | 15 | 7 | |
BELGIUM | 10414336 | 10 | 5 | 10 | 6 | 7.75 | 5 | 10 | 6 | |
SWEDEN | 9059651 | 3 | 4 | 9 | 7 | 5.75 | 4 | 12 | 6 | |
BRAZIL | 198739269 | 5 | 4 | 5 | 4 | 4.5 | 4 | 30 | 12 | Brazil is another hungry hot spot |
AUSTRIA | 8210281 | 4 | 5 | 5 | 2 | 4 | 4 | 6 | 5 | |
NAMIBIA | 2108665 | 0 | 2 | 8 | 0 | 2.5 | 0 | 4 | 2 | |
INDONESIA | 240271522 | 1 | 3 | 3 | 3 | 2.5 | 1 | 3 | 2 | |
ROMANIA | 22215421 | 1 | 2 | 6 | 1 | 2.5 | 1 | 4 | 2 | |
CZECH REPUBLIC | 10211904 | 9 | 4 | 2 | 1 | 4 | 2 | 6 | 4 | |
PORTUGAL | 10707924 | 3 | 2 | 2 | 2 | 2.25 | 2 | 3 | 2 | |
FINLAND | 5250275 | 8 | 3 | 2 | 0 | 3.25 | 2 | 5 | 3 | |
POLAND | 38482919 | 1 | 2 | 3 | 0 | 1.5 | 1 | 3 | 2 | |
CHINA | 1338612968 | 0 | 3 | 0 | 2 | 1.25 | 0 | 25 | 10 | China is big on Geospatial |
NEW CALEDONIA | 227436 | 0 | 0 | 0 | 4 | 1 | 0 | 4 | 2 | |
TAIWAN | 22974347 | 2 | 1 | 1 | 1 | 1.25 | 1 | 6 | 2 | |
MEXICO | 111211789 | 2 | 3 | 0 | 0 | 1.25 | 2 | 30 | 10 | Mexico is very close and hungry |
Totals | 418 | 1431 | 839 |
Survey
An alternate method would be to use a short survey that is widely publicized. This could generate good numbers for the FOSS4G community but would miss the non-community members.
Formula
A = ((P * I) / (D)) * 1/E
Where
- A = number of attendees from a country
- P = population of a country
- Seems to be too many contradictions, e.g., Switzerland is consistently over-represented and Japan under-represented.
- I = Industrialization
- GDP of the country
- HDI may be a better measure... but...
- Top 10 countries by average attendance (except South Africa) have high GDI
- Top 20 adds Namibia and Brazil with "lower" HDI
- 90% of attendees come from high-HDI nations, cannot account for variation
- D = Travel Cost Distance from country
- Proxy - air travel distance from largest city to destination (except hosts) using TravelMath
- Price of airfare in local currency?
- Effort in getting VISA?
- E = Some unknown general economic factor
- FOSS4G 2009 realized very low numbers due to the economic crisis
- DJIA?
Will this work for States within the USA? Assumes consistency across countries
- Is there some factor for USA/CAN?
Correlating Factors
- Economic Crisis
By Country
- Population
- Industrialization
- GDP as proxy
- Distance
- Local vs. Distant
- What is "local"?
- Number of Code Contributors
- Which projects?
- Should most countries be "ignored"?
- "Top 10" countries provide 75% of attendees
- "Top 20" countries provide 90% of attendees
- "Top 30" countries provide 95% of attendees
- Calculate for the Top 10 and extrapolate as 75%
By Region
- Distance
- Industrialization
- Oddly, Europe
Venue
One of the biggest decisions is what venue to hold the conference. Venues with larger capacities have a higher per-attendee cost. If we can come up with a reasonable range of number of attendees, we can make a better decision about the venue. The availability of a particular venue can also impact dates.
In Denver, there are several prospective venues:
- Colorado Convention Center
- Largest space: 5000 people
- Near light rail
- Denver Sheraton
- Largest space: 3000 people
- Near light rail
- Denver Tech Center Marriott
- Largest space: 1500 people
- Near light rail
- Denver Tech Center Hyatt Regency
- Largest space: 1350 people
- Near light rail
- Denver Rennaissance
- Largest space: 1200 people
- Not near light rail
- Sheraton Denver West
- Largest space: 750 people
- New light rail station won't be operational until 2013
- Inverness Hotel and Conference Center
- Largest space: 600 people
- Near light rail
Sponsorship
Sponsors (and exhibitors) like to know how many people they are reaching by investing in the conference.
Attendee Profiles
There is no one profile that fits all attendees. This is the nature of FOSS. Most significantly, the "heart and soul" of FOSS are the contributors who often are people who are "financially challenged". Here are some common traits:
1. Technically sophisticated
a. Event needs WiFi everywhere b. Power outlets are good - people will be using laptops in plenaries
2. Generally like caffeine and beer
Profiles
Low-income Contributor
This person may be a student, come from a developing nation, or just be under-employed. She will be paying out of pocket for the conference. She will not stay at the conference hotel and may even be interested in a hostel or "couch-surfing". She will use public transportation or borrow a bicycle (highlight the new Denver bike share program). Despite her low income, she will come armed with at least one laptop and a smart phone. She will use the conference WiFi and power outlets heavily.
a. Low conference fee with extras being optional b. Very low cost accommodation options c. Heavy use of wifi and power d. More likely to lead a workshop than attend one.
FOSS Professional
This person may be a consultant applying FOSS for customers or work for a large company integrating FOSS into the company's systems. Consultants will be paying out of pocket (as consultants always do) and others will have their employer either pay for or reimburse costs. She may likely stay at the conference hotel to increase networking opportunities. If she doesn't stay at the conference hotel, she will stay at a nearby hotel. She will come to workshops with a laptop but may only carry a smart phone to presentations. She will rely on public transportation or cabs. Most US government employees fall in this category. Many of these people like to self-identify with the "Low-income Contributor" as there is a certain romance to being heavily involved in contributing to FOSS.
a. Will pay for extras, especially if the conference fee is low. b. Likely stay at conference hotel if the rates are good c. Likely to attend many workshops
Business Leader
This person will be in upper management or an officer of a company. The company may be an established utility or a tech startup. Her time is more valuable than the cost of the conference. She will stay at the conference hotel but will likely only stay for a one or two days. She will rely on taxis rather than public transportation. May have a laptop but more likely will use a smart phone or a tablet with 3G rather than depending on WiFi.
a. May appreciate an "everything" registration option. b. Will stay at the conference hotel c. May attend a workshop or two but won't want to choose until she is here.
Colorado Front Ranger
This person is from the local area and may fit one of the above profiles. Her company may give her one or two days off to attend workshops or key presentations but may not pay for registration. She will not stay in a hotel and will either drive or use public transportation each day to reach the event. She will have a laptop and a smart phone and will use WiFi and power. May attend one social event but not all.
a. Would appreciate single-day registrations b. Will not stay at the conference hotel (or any hotel) c. Will likely attend many workshops - and may skip the rest of the conference.
Percent Makeup
Profile | Percent | Low | Mid | High |
---|---|---|---|---|
Low-income Contributor | 15% | 62 | 126 | 180 |
FOSS Professional | 30% | 124 | 252 | 360 |
Business Leader | 5% | 31 | 63 | 90 |
Colorado Front Ranger | 50% | 213 | 420 | 600 |
Totals | 100% | 426 | 839 | 1200 |