Industries / Media, arts and recreation / Zoos and aquariums
Visitors and education
Opening collections to the public and teaching visitors about animals and conservation, on site and at a distance.
7
Milestones achieved, from 1847 to 2024.
1
To watch for: announced, not yet confirmed outside.
2
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
1847
Gardens open to all
London's zoological gardens, open only to the society's members since 1828, opened to the public in 1847 to help fund the Zoological Society of London. The first reptile house (1849), the first public aquarium (1853) and the first insect house (1881) followed, by the London Museum's account.
Credit
Zoological Society of London
2000
A seafood guide in wallets
Seafood Watch began as a list of sustainable seafood choices drawn up by the Monterey Bay Aquarium in 1997–1999 for its "Fishing for Solutions" exhibit, and became a wallet-sized pocket guide that rates seafood "Best Choice", "Caution" or "Avoid". A case study counts more than 1,000,000 West Coast guides handed out since the program began in 2000.
LessonThe case study counts guides, downloads and restaurants signed up but gives no measure of changed buying, though shifting buying habits is the program's stated goal.
Credit
Monterey Bay Aquarium, funded by the David and Lucile Packard Foundation
2010
A global visit count
WAZA asked 12 national and regional zoo associations, with more than 1,000 member institutions, for their 2008 figures. About 600 million visits were reported; after using 1990 figures for regions that under-reported, Markus Gusset and Gerald Dick estimated "more than 700 million visits annually", which they called "most certainly an underestimate", and conservation spending of about US$350 million a year.
LessonThe same authors wrote that "it remains largely unclear how education initiatives affect visitor behavior".
Credit
Markus Gusset and Gerald Dick, WAZA (Zoo Biology, online December 2010)
2010
Claims of learning challenged
A 2007 study conducted by AZA (Falk and colleagues) was "widely heralded as the first direct evidence" that zoo and aquarium visits change people's attitudes over the long term. Marino, Lilienfeld, Malamud, Nobis and Broglio argued in Society & Animals that it had "at least six major threats to methodological validity" and that "there remains no compelling evidence" that zoos promote attitude change, education or interest in conservation.
LessonThe critics did not claim that zoos fail to teach, only that the evidence was weak, and asked for further study "using methodologically sophisticated designs".
Credit
Marino, Lilienfeld, Malamud, Nobis and Broglio
2016
A barrier fails
In May 2016 a three-year-old boy crossed the public barrier at the Cincinnati Zoo's Gorilla World and fell into the enclosure; zoo staff shot Harambe, a 17-year-old western lowland gorilla, saying he posed a "life-threatening danger" to the child. A USDA inspector wrote that the barrier, built in 1978, had no record of anyone crossing it before and was compliant until the breach. The zoo replaced the 32-inch (81 cm) barrier with one 41 to 42 inches (104 to 107 cm) tall, covered in nylon mesh.
LessonThirty-eight years without an incident did not show that the barrier was high enough.
Credit
USDA (inspector Cody Yager); Cincinnati Zoo
2020
The zoo comes home
After closing for COVID-19 on 14 March 2020, the Cincinnati Zoo streamed "Home Safari" videos on Facebook. A livestream of Moe the sloth drew 1.4 million views, close to the zoo's 1.5 million visitors a year, and its Facebook followers rose 50% between 14 and 24 March, against less than 1% at the St. Louis, National and Bronx zoos.
Credit
Cincinnati Zoo & Botanical Garden
2024
Learning pooled across studies
Xavier McNally and colleagues at the University of Sheffield, working with Chester Zoo, pooled 56 studies of zoo-led programs and found "a small to medium positive effect" on visitors' knowledge, attitudes and stated intentions (an effect size of 0.40 on a standard scale), with no sign that unpublished null results were missing. Effects were larger in studies at a single zoo than across several.
LessonOn behaviour, visitors "reported being more likely to act", so the pooled evidence is mostly about intentions.
Credit
Xavier McNally, T. L. Webb, C. Smith, A. Moss and J. Gibson-Miller (Conservation Biology, 2024)
Watch for
AI answers visitor questions
Hyperleap, an AI company, says its assistant at Mysuru Zoo, a government wildlife park in Mysuru, India, answers visitor questions in five languages, helps plan routes through the park and brings conservation stories into the conversation, and that the zoo handled six times as many visitor queries.
Achieved whenthe zoo or an outside evaluator publishes before-and-after query counts and measures whether visitors who used the assistant learned or did more than those who did not.
Claimed by
Hyperleap, July 2026 (updated November 2026), citing "the client's disclosed results"
Open
01
Measure what visitors do
A pre-registered field study (one whose design and analysis are published before data collection) that measures an action visitors actually take after a zoo visit, rather than what they say they will do. For example, visitors who opt in could get a sustainable-seafood or palm-oil shopping guide through a link, and the study would count, with consent, who opens and uses it a month later, comparing visitors who saw a conservation message with those who did not.
The briefCredit · unclaimed02
Tag every sign's technique
A public dataset of zoo and aquarium signs, each coded for the behaviour-change techniques it uses (such as giving a specific action, a reason or a prompt) from a published taxonomy of such techniques, so that later studies can test which kinds of sign go with measured learning.
The briefCredit · unclaimed
Open 01 · the brief
Measure what visitors do
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A pre-registered field study (one whose design and analysis are published before data collection) that measures an action visitors actually take after a zoo visit, rather than what they say they will do. For example, visitors who opt in could get a sustainable-seafood or palm-oil shopping guide through a link, and the study would count, with consent, who opens and uses it a month later, comparing visitors who saw a conservation message with those who did not.
Why it is still open
WAZA's own authors wrote in 2010 that "it remains largely unclear how education initiatives affect visitor behavior", and the 2024 meta-analysis of 56 studies found that, on behaviour, visitors "reported being more likely to act".
Built with what exists
The Open Science Framework (OSF) for pre-registration; a web form and link tracking with consent (any standard survey tool); a university ethics board's approval; an existing free guide such as Seafood Watch as the action to measure.
First step
Write the design and register it on OSF, and ask one zoo's education team for a two-week trial slot.
Done looks like
At least 1,500 consenting visitors at three or more zoos, randomly assigned to message or no message, with the share who used the guide at one and three months published whatever the result.
Open 02 · the brief
Tag every sign's technique
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A public dataset of zoo and aquarium signs, each coded for the behaviour-change techniques it uses (such as giving a specific action, a reason or a prompt) from a published taxonomy of such techniques, so that later studies can test which kinds of sign go with measured learning.
Why it is still open
McNally and colleagues identified only six specific behaviour-change techniques across 56 studies of zoo-led programs, and found that differences between programs did not explain why effects varied. No coded collection of the signs visitors actually read was found.
Built with what exists
Phone photos of signs; Tesseract or a vision AI model to read the text; Gemini Decision 1 to score each sign against a fixed list of yes-or-no questions, one per technique (see the Note "A thousand questions, one pass"); the Behaviour Change Technique Taxonomy v1 (Michie and colleagues, 2013); two human coders to check a sample.
First step
Photograph every sign at one zoo in a day and code the first 100 by hand.
Done looks like
At least 2,000 signs from at least 20 zoos coded and published, with agreement between the model and two human coders of at least 0.7 (Cohen's kappa, a standard agreement score) on a random 10% sample.