As Amazon’s stunning pullout from New York fades into the news archives, its potent lessons for business negotiators risk being lost. Highly promising deals in diffuse multiparty settings with many potential spoilers, like Amazon’s planned headquarters in Queens, often collapse as a result of negotiating too narrowly with those who have formal power and authority. Negotiation experts have a patriarchal name for a version of this classic—and avoidable—mistake: Decide-Announce-Defend or DAD.
Along with gaining the full-throated support of New York Mayor Bill de Blasio and Governor Andrew Cuomo, Amazon officials understandably figured that the prize it offered New Yorkers would sell itself: 25,000+ jobs paying in excess of $100,000 each with all the ancillary economic benefits. Decide (on Long Island City, Queens), Announce (the choice), and Defend (from attacks) ... and, if you’re still standing, you win.
Except, Amazon decided, announced, defended, lost, and abruptly pulled out, blindsiding virtually everyone involved. As New York’s chief negotiator for the deal mourned, this “was supposed to have been a coronation but instead was more like a coronary.”
"Requiring full consensus in a multiparty deal makes you hostage to the most extreme or reluctant party."
This surprisingly common result is why an “A” is often appended to DAD: “DADA” means Decide-Announce-Defend-Abandon. An apparently irresistible deal blessed by the top authorities runs aground on unanticipated opposition. The trail of such failed deals is long; for examples, see here, here, and here. For instance, consider the award of the 2024 Olympics to Boston over Los Angeles, San Francisco, and Washington DC. Boston’s successful bid was driven by the support of the state’s governor, Boston’s mayor, and many of its most influential citizens. Yet a small group of opponents catalyzed a local movement that, despite being outspent 1,500 to one by the bid’s boosters, ultimately caused the city to back out in 2015.
[DAD can be appropriate for purely private deals without potential blockers, though upfront take-it-or-leave-it demands run higher rejection risks than more back and forth negotiations.]
The frequent failures of DAD-style negotiation have led some project advocates to seek consensus among all stakeholders. In a city like Queens, riven with many factions and political agendas, Amazon would never have reached full consensus and didn’t try. Requiring full consensus in a multiparty deal makes you hostage to the most extreme or reluctant party. When you can anticipate unconditional opponents, or skeptics with diverse agendas who may opportunistically band together, don’t hand them blocking power.
So let’s assume that, with many contenders, Amazon had powerful reasons to choose New York. Comparative advantages presumably ranged from a large and highly educated employee pool to big incentives and to local entertainment options galore—not to mention that, once Amazon’s new headquarters were built, much of New York’s congressional delegation could be counted on for political support . . . in addition to that from Washington State and elected officials from its other new headquarters in Virginia. Apart from avoiding the DAD and full consensus traps, what could Amazon have done to retain these New York advantages? What are the broader lessons for those facing similarly challenging negotiations?
Building sufficient consensus
The goal should be to build “sufficient consensus” for a “winning coalition” in spite of potential blockers. This means earning enough support among enough of the right parties to gain agreement on your proposal and ensure successful implementation. Building such a sustainable winning coalition involves systematic steps that my colleague David Lax and I call a "negotiation campaign".
- In a complex, multiparty setting, don’t take victory for granted, ever. Today, social media can quickly amplify the views of even a few vocal opponents, giving voice to latent negative concerns of many otherwise passive groups. As Amazon learned, an apparent “movement” can seemingly spring up from nowhere. It can rapidly gain traction, surprising and thwarting the confident protagonists of an apparently popular project.
- Actively monitor local currents and cross-currents of opinion. Reportedly, Amazon did not hire a single New Yorker to continuously engage with community groups to build support; most of its representatives shuttled between Washington and Manhattan. On-the-ground presence would have provided invaluable local intelligence on fast-changing currents of opinion. Beyond broad polling that indicated 70 percent support for the new headquarters, it is not clear whether Amazon commissioned frequent local tracking polls—by different demographic groups—to pick up concerns as well as hopes. Polling and people are your eyes and ears; without them, you risk flying blind into a killer storm.
- Identify and nurture potential allies before you need them. To Amazon, the supporters seemed self-evident; after all, more than 200 cities desperately vied for the prize it bestowed on New York. Yet well-organized opponents overcame the unorganized supporters of the deal. Old-school reliance on the mayor and governor, powerful power brokers, proved unable to mobilize sufficient backing. Beyond cultivating elite support, a project sponsor should systematically work with community groups and local leaders so they feel intense personal and tangible stakes in the proposal. Detailed preliminary discussions with construction trades should make the huge amount of new work crystal clear. Early “job fairs” with sample applications could help persuade lower-skilled groups that thousands of new support jobs and training opportunities would be forthcoming along with the $100,000+ job bonanza for high-skilled workers. Community groups looking for improved parks, sidewalks, and local amenities could be nurtured at relatively low cost with “good neighbor” credible commitments. Failing to send CEO Jeff Bezos to New York to stroke the egos of local supportive politicians and learn firsthand of any qualms was a missed opportunity. Having identified and nurtured supporters, they can be activated in favor of your project if and as needed.
- Identify all likely and potential opponents at the outset of the process. Project advocates often find themselves in reactive, defensive mode, having focused too late on opponents who have seized the initiative.
- Beware of opponents with diverse concerns joining forces to form a “blocking coalition.” With her upset victory for a Congressional seat, Alexandria Ocasio-Cortez and other Queens organizers already had names, addresses, phone numbers, and social media accounts ready to launch against this new corporate Goliath in the name of ordinary people. Even though polls suggested that a strong majority of New Yorkers generally supported the Amazon deal, all manner of grievances, fears, and agendas quickly coalesced into a vocal anti-Amazon movement, much in evidence at two hostile city council meetings. When the newly Democratic state senate appointed an opponent of the deal to an obscure board that required unanimity for Amazon’s project to proceed, the game was effectively over. Amazon picked up its marbles and went home, presumably to choose a more welcoming location or spread the planned headquarters employees across several cities.
- From the beginning, actively listen to the concerns of potential opponents and address them to the extent possible. This pullout from New York was not foreordained. Beyond cultivating supporters early, Amazon should have taken the initiative to listen and address concerns of skeptics. While Amazon would not likely have changed its no-union stance, for example, it could have engaged with New York unions to commit to construction projects, subcontractors, and service providers without respect to their union status. For those concerned with transportation and congestion, Amazon could have taken the initiative to recognize the potential problems and pledged to collaborate with local groups and authorities to solve them. For transportation and housing advocates, it could point to the $27 billion in taxes it projected to be paying over the next decade and commit to working with elected officials and interest groups to boost the stock of affordable housing and improve transit. It could even have earmarked a chunk of its $3 billion incentive package, say $1 billion, for transit and housing improvements near its new headquarters. This would have gone a long way to quiet opposition.
- Remember that negotiation does not end with a “yes,” but requires enough ongoing support for implementation and sustainability. The kind of negotiation campaign that I’ve sketched is designed to build a sufficient, sustainable “winning coalition” on behalf of an initiative like Amazon’s. But as this experience shows, an initial “yes” is only the entry point to a successful project, which requires sustained support for long-term success.
For Amazon’s project, opposition from organized labor was a given. Amazon is a strongly anti-union company, and New York has many powerful unions that despise Amazon’s broader labor practices. (It didn’t help that, in a public meeting, Amazon conspicuously refused to pledge neutrality around unionization issues.)
Next, consider people in and around Queens who were apprehensive about a giant new employer coming to town. Wouldn’t all these new employees cause traffic congestion and overwhelm the rickety subway lines that serve Long Island City? (It didn’t help that Amazon didn’t pledge to invest in better local transportation—and even carved out a provision for a helipad to permit Amazon execs to remain above it all.)
And for a rental-dominated community, wouldn’t this wealthy influx of Amazonians drive up rents, housing prices, and real estate costs, forcing long-time residents and local independent businesses to move?
Moreover, there was the $3 billion “incentive” package that New York offered to entice Amazon. This quickly became an albatross: Why did Amazon, whose value briefly exceeded a trillion dollars, headed by the world’s wealthiest man, need such incentives? (It didn’t help that Amazon paid no federal taxes on its $11.2 billion profit in 2018.) New Yorkers disapprovingly asked why this incentive money shouldn’t be used for more urgent local needs. Yet most of the $3 billion package would come from reductions in future taxes to be paid by Amazon. So if the company bailed, there would be no $3 billion pot for other purposes. Both the mayor and governor tried in vain to play catch-up on this “welfare for the rich” issue, arguing that new taxes from Amazon over this period would amount to some $27 billion, implying a nine-to-one return on the $3 billion of incentives.
Harder for Amazon to have anticipated was the local manifestation of an increasingly anti-corporate national mood, especially among progressive Democrats in bluer-than-blue New York, that came to a head in the 2018 elections.
Maybe de Blasio and Cuomo counseled Amazon to take such steps early in the process. However, after a few desultory efforts late in the game to meet local concerns, the Seattle giant had had enough and punted. After the fact, however, the New York mayor stated: “We gave Amazon the opportunity to be a good neighbor and do business in the greatest city in the world … Instead of working with the community, Amazon threw away that opportunity.”
[Ongoing sequel: after the pullout and conversations with Bezos, New York Governor Cuomo sought to coax Amazon back—in line with this section’s advice to build a supportive coalition and deal with opponents. Cuomo refused to approve the appointment of the project opponent to the board that could have vetoed the proposed Queens headquarters. Cuomo orchestrated the publication of a full-page letter to Bezos in the New York Times that included over 70 supportive signatories including the AFL-CIO and other unions, local business leaders, community groups, public housing tenant leaders, the local N.A.A.C.P. chapter, pastors, and elected officials. The letter indicated that Cuomo “will take personal responsibility for the project’s state approval,” and that Mayor Bill de Blasio “will work together with the governor to manage the community development process.” At this writing, Amazon has not commented nor is it clear that this effort will succeed, but such an after-the-fact campaign to build support and deal with opponents surely would have been more effective before the stunning pullout.]
Had Amazon approached the New York headquarters decision in line with the seven steps outlined above, experience suggests that its odds of success would have been far greater. Of course Amazon might have considered this approach but chose DAD instead. Playing hardball in this highly visible case could intimidate other cities seeking the jobs and economic boost that a new Amazon facility would bring: “Don’t push us or else we’ll expand elsewhere.” This uncompromising stance has worked when Amazon was threatened with special taxes or other burdensome measures; examples include Texas, South Carolina, and Seattle. Yet by pulling out of New York, Amazon lost the very real advantages that led it to choose Queens in the first place. And if the seven-step approach detailed above hadn’t paid off in a reasonable time, Amazon still retained the hardball pullout option.
Even with a worthy project, there will inevitably be opponents and naysayers; that is the nature of complex, multiparty contexts. But having an ongoing finger on the pulse of the communities and groups involved, having nurtured allies, having anticipated opponents and sought to address their concerns, and having built enough support and neutralized enough opposition to overwhelm those who remain unconditionally hostile, the odds of a fatal blocking coalition forming can be drastically reduced—and your project is far more likely to succeed.
James K. Sebenius is the Gordon Donaldson Professor of Business Administration at Harvard Business School, Director of the Harvard Negotiation Project at Harvard Law School, and most recently, co-author of Kissinger the Negotiator: Lessons from Dealmaking at the Highest Level.