Innumerable breathless commentaries aside, there are simple lessons to learn from the DuPont-Trian proxy battle witnessed earlier this spring. Before we get to them, let’s touch on the cardinal rule of all corporate board elections.

As an astute elected official confided in us decades ago, all political elections for chief executive officers—whether president, governor, or mayor—are actually referenda on the incumbents more than tests of the challengers.

So, too, all activist battles are foremost referenda on the target companies. Everything else—from the importance of the proxy advisory firms to the competing board nominees—is subservient to that context.

So back to DuPont and its clash with Trian Partners. For anyone who might have missed it, Trian Partners, Nelson Peltz’s hedge fund, had tried to elect four directors to the board of the Delaware-based DuPont, arguing that it had failed to increase margins sufficiently and was burdened with a complicated and costly corporate structure. But while many activist battles this year succeeded in electing directors or otherwise negotiating a settlement that put activist nominees on the board, DuPont’s nominees won every seat. Trian lost.

Why? It’s simple, really. Trian picked the wrong company. Or to put it another way, DuPont had made itself the wrong company. To understand how, let’s take a quick post-election tour of four conflict areas that usually help activists win, but in this case, helped DuPont: performance, governance, proxy advisers, and the board nominees themselves.

Performance. Trian’s platform called for bifurcating the company, but its public pronouncements made clear that Trian really wanted the vote to be a referendum on DuPont’s board of directors’ performance and governance. Peltz argued over and over that DuPont was lagging its peers and had repeatedly missed earnings estimates, and that he could reduce expenses by some $4 billion a year. DuPont and its CEO, Ellen Kullman, argued that its returns had outpaced the S&P 500. It had improved margins by 65 percent over the previous five years.

What about the standard activist argument that the board was sitting on too much capital? DuPont had returned about $12 billion to shareowners from 2009 through 2014. What about breaking up the company? DuPont was also on top of that one, having previously announced a spin-off, though more limited than the one Trian advocated. Having anticipated attacks, DuPont went to great lengths to communicate its stance to investors.

In the end, Trian forgot the cardinal rule. An activist contest is not an even battle between competing visions. It is a referendum on incumbent management and directors. DuPont and its advisers deserve credit for a campaign that turned Trian’s misstep into a fatal disadvantage.

Result: DuPont demonstrated to most observers that it was anything but a fossilized relic and had seriously considered most of the strategic moves Trian suggested. That left Trian in the unenviable position of arguing for more or better—but not different.

Lesson learned: Opportunity cost is always a difficult case to make. When a target company has dedicated resources to cultivate a reputation for openness and a proven track record of performance, even if the company theoretically might have done better, it probably isn’t a good target.

Governance. Trian severely criticized the governance practices of the DuPont board, calling it “corporate governance from the ’70s and ’80s.” Peltz added that DuPont was “one of the worst examples of corporate governance that I’ve seen in years.” Those sort of charges usually get governance activists riled up and leaning toward change. But DuPont actually had steered resources, leadership time, and outside advice to bank goodwill among its shareholders. It had built a reputation in the governance community as one of the better-governed large companies. Even ISS, which wound up backing Trian’s candidates, ranked DuPont in the highest decile of companies for audit, shareholder rights, and compensation, and in the second decile for the board itself.

Result: A number of leading large investors with strong governance sensitivities, such as CalPERS, Blackrock, and the Canadian Pension Plan Investment Board, voted with DuPont.

Lesson to learn: Charges of governance failure get attention. But substance matters.

Proxy advisers. Activists and target companies avidly seek the endorsements from the leading proxy advisory firms, ISS and Glass Lewis. But their highest utility for institutional investors is in acting as a screening service. They flag which of the 10,000+ proxy decisions that a large investor must vote upon each year merit special attention. For relatively routine voting decisions, and even for low-profile contests, the proxy adviser recommendation may be the only, or most of, the information an investor has in making its decision.

DuPont was as high-profile as it gets. Information and analysis came at investors multiple times each day. Institutional investors hardly needed a proxy advisory firm to pay attention or to assess how their vote might affect the company’s value; they could do that homework themselves. Think of it this way: You probably note with interest your local newspaper’s endorsement for town council. But it’s highly unlikely to be the only informed opinion you hear about in a U.S. presidential race.

Results: Trian won the recommendations of the major advisory firms. That wasn’t enough to guarantee the victory.

Lesson to learn: The higher the profile of the proxy contest, the less the advisory firms’ recommendations matter.

Board nominees. DuPont actually followed the standard playbook and offered Trian a seat on the board. It was only when Trian said no to the offer, unless Peltz himself was named to the board, that DuPont made the final decision to take the contest all the way to an election. Trian is generally a well-regarded fund, but DuPont skillfully framed the activists’ “Peltz or no deal” position as more about ego than substance.

The argument struck particular resonance since Peltz already sits on three other major boards, which would make him “over-boarded” according to some governance guidelines. Meanwhile, DuPont’s board adopted a canny offense, replacing two incumbent directors with well-regarded new nominees. The strategy neutralized Trian’s charge that the company was saddled with an entrenched or stale board.

Results: A clean sweep for DuPont’s nominees. But the outcome was close, meaning that Trian remains in the game with significant shareholder support. DuPont’s board has no room to relax.

Lesson to learn: Board members have reputations, and picking on a well-respected board while holding out for Peltz alone as an acceptable director made Trian look obstinate. That’s not a good idea. Better to pick entrenched boards with obviously unsuited candidates running for re-election.

In the end, Trian forgot the cardinal rule. An activist contest is not an even battle between competing visions. It is a referendum on incumbent management and directors. DuPont and its advisers deserve credit for a campaign that turned Trian’s misstep into a fatal disadvantage.