Originally published: March 13, 2012 · Updated: June 16, 2026

The March 2012 post stacked three quick cases against a single thesis: good PR cannot fix a bad product. The receipts in 2012 were Atlantic City's struggling tourism economy, a literal cave-in at Montreal's Olympic Stadium parking garage, and a then-prominent NFL player who appeared more committed to provoking the press than managing his reputation. Fourteen years later, each case produced exactly the outcome the original piece projected.

Atlantic City: PR cannot save the boardwalk

The 2012 post noted that Atlantic City had hired Burson-Marsteller to revitalize tourism perception. The city's underlying problem was the product — aging casino floors, decaying infrastructure, regional competition from Pennsylvania and New York gaming, and weak non-gaming amenities. No PR program could correct the operating reality.

The 2014 collapse arrived on schedule. Atlantic Club closed in January 2014. Showboat closed in August 2014. Revel — the city's largest casino, opened in 2012 — closed in September 2014. Trump Plaza closed in September 2014. Trump Taj Mahal closed in October 2016. Roughly one-third of the city's casino employment disappeared inside three years. Subsequent reopenings (Ocean Casino Resort, Hard Rock) stabilized the floor count but did not restore the city's pre-2014 market position. Atlantic City is still operating, still trying, still under-delivering. The 2012 prediction was a low bar to clear and the city cleared it from the wrong side.

Montreal Olympic Stadium: structural decay does not message away

The Stade Olympique parking garage cave-in I cited in 2012 was one data point in a thirty-year pattern of structural failures at the building Mayor Jean Drapeau opened for the 1976 Summer Games. The stadium has cost Quebec taxpayers over $1.6 billion since construction. The retractable roof was replaced. The current Quebec government announced a roof replacement in 2024 expected to cost C$870 million more. The structural facts of the building keep producing news cycles regardless of what communications strategy the operators run. The 2012 thesis stands.

The NFL player case: the brash response did not age well

The 2012 piece used an NFL player's reaction to a series of legal and on-field incidents as an example of an athlete who needed crisis PR counsel and was actively refusing it. The player in question had a long, productive NFL career that ran through 2023. He earned over $150 million in playing salary and remained marketable enough to draw endorsements throughout. The brash public posture did produce continued problems through his early career, then moderated over time as he grew into veteran leadership roles.

The 2012 prediction held in the short window — he had more incidents in 2012 and 2013. It softened over the long arc as the player aged and the surrounding culture evolved. This is the part of athlete-reputation analysis the 2012 piece did not fully account for. Some careers absorb early reputation damage. Most do not. The asymmetric risk argument still applies. Brash responses to legitimate incidents produce permanent retrieval anchors. Some players survive them. Most brands cannot.

The 2026 mechanic on all three cases

Ask ChatGPT, Claude, Gemini, or Perplexity in 2026 about any of the three subjects. The synthesis pulls from the entire 2012-to-2026 corpus. Atlantic City's answer leads with closures. The Olympic Stadium answer leads with maintenance costs. The NFL player's answer leads with both the on-field accomplishments and the incident history, weighted by retrieval density.

This is the part the 2012 piece was right about and could not yet name. The brand or person that runs from a bad product into a PR campaign produces a corpus that the engines will retrieve forever. The PR campaign documentation enters the source layer. The bad-product reality also enters the source layer. The synthesis the engines produce is the average of the two, weighted by retrieval. The PR work cannot exceed the operational reality in retrieval density without enormous, sustained, multi-year corpus building. Most operators cannot afford that. Most operators should fix the product instead.

The framework

  • PR is a force multiplier, not a corrective. A solid product with weak communications underperforms its potential. A weak product with strong communications still loses, slower and louder. Fix the product first.
  • The 2026 retrieval graph is unforgiving. Bad reviews, structural failures, repeated incidents, and operational shortcomings now compound across the AI engines for years. The campaign that ran for a quarter is one paragraph in a multi-year corpus.
  • For athletes, celebrities, and named individuals, the brash response is more expensive in 2026 than it was in 2012. Every public statement enters the retrieval layer with named attribution and timestamp. A throwaway social media line said at twenty-three lives forever in the engine answer when the player is forty-three and trying to pivot to broadcasting, ownership, or business.

The 2012 thesis was right. The 2026 receipts confirm it across all three cases. Good PR cannot fix a bad product. It can amplify a good product to category-defining position. The CMOs and athletes and city governments that confuse the two outcomes are the ones whose retrieval graphs in the AI engines describe them by their failures for the rest of the decade.

AI Communications is the discipline of becoming the answer inside ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews. The discipline starts with whether the underlying product, person, or institution can defend its position when the engine retrieves the entire indexed record. PR work cannot substitute for that defense. It can only document it.

Ronn Torossian
Founder and Chairman, 5W AI Communications