. 
“[C]ircumstantial allegations about


      1
       The district court also dismissed two other claims, both suggesting that the
plan and TIAA had entered into prohibited transactions with each other. See 29
U.S.C. § 1106(a)(1)(B), (D). The dismissal of these claims is not before us,
however, because the plaintiffs have opted not to pursue them on appeal.
                                          -4-
[the fiduciary’s] methods” based on the “investment choices a plan fiduciary made”
can be enough. Meiners v. Wells Fargo & Co., 
898 F.3d 820
, 822 (8th Cir. 2018).

                                          A.

      The first claim clears this pleading hurdle. It alleges that fees were too high
and that WashU should have negotiated a better deal.

       Perhaps the clearest example is Vanguard’s “mixed” lineup of funds. By
mixed, we mean that there were different share classes in its lineup, including
“institutional” and “retail” shares. For some investments, Vanguard offered lower-
cost institutional shares, which have higher initial investment requirements. Instead
of offering these shares across its entire lineup, however, it offered retail shares for
other funds, even though minimum investment requirements are “routinely waived”
for individual investors in large retirement-savings plans. The failure to replace
these shares with their lower-cost counterparts breached WashU’s fiduciary duty, in
the plaintiffs’ view, because higher fees led to lower overall returns. They believe,
in other words, that WashU’s lack of diligence cost them money.

       We allowed similar allegations to proceed in Braden. Braden involved a
retirement-savings plan filled with retail shares despite “a very large pool of assets”
and a “highly competitive” 
marketplace. 588 F.3d at 595
. Taken together, we
concluded that these facts were enough to create an inference of a “flawed” process,
even if the complaint itself did not “directly address[]” how the plan made its
selections.
 Id. at 596
.

      We reach the same conclusion here. The complaint alleges that the
marketplace for retirement plans is competitive, and with $3.8 billion invested,
WashU’s “pool of assets” is large. From these facts, two inferences of
mismanagement are plausible from WashU’s failure to offer more institutional
shares. The first is that it failed to gain access to them because, as the complaint

                                          -5-
alleges, it did not negotiate aggressively enough with Vanguard. The second is that
it was asleep at the wheel: it failed to pay close enough attention to available lower-
cost alternatives. Either way, a “failure of effort [or] competence” is enough to state
a claim for breach of the duty of prudence.
 Id.

       WashU has 
two responses. One is that a plan without retail shares would not
have covered the plan’s costs because, according to WashU, institutional shares do
not make large enough revenue-sharing payments. The other is that, with a more
competitive marketplace and growth in the plan’s total assets, it has demonstrated
diligence by shifting into more institutional shares over time. Taken together, it
says, the only plausible inference is that it has prudently managed the plan by
negotiating for the best possible deal under the circumstances.

       WashU has identified one plausible inference, but it is not the only one. On a
motion to dismiss, we must draw every reasonable inference in favor of the plaintiffs.
See Sabri v. Whittier All., 
833 F.3d 995
, 998 (8th Cir. 2016). So the fact that
mismanagement is another plausible inference means that this claim cannot end here.
See Whitney v. Guys, Inc., 
700 F.3d 1118
, 1130 (8th Cir. 2012) (“[A] well-pleaded
complaint may proceed even if it strikes a savvy judge that actual proof of those
facts is improbable, and that a recovery is very remote and unlikely.” (alteration in
original) (quoting Bell Atl. Corp. v. Twombly, 
550 U.S. 544
, 556 (2007))). 2

                                          B.

      We reach the opposite conclusion on the other claim, which puts the spotlight
on three specific investment options. According to the complaint, WashU should
have jettisoned them because they were poor performers and cost too much.


      2
       The complaint also includes a number of other allegations, but it is not
necessary to address them in light of our conclusion that the plaintiffs have otherwise
alleged enough for this claim to survive.
                                         -6-
      A fiduciary can breach its duty of prudence if it fails to monitor and remove
imprudent investment options. See Tibble v. Edison Int’l, 
135 S. Ct. 1823
, 1829
(2015). It is no defense to simply offer a “reasonable array” of options that includes
some good ones, see Tussey v. ABB, Inc., 
746 F.3d 327
, 335–36 (8th Cir. 2014), and
then “shift[]” the responsibility to plan participants to find them. Hecker v. Deere
& Co., 
569 F.3d 708
, 711 (7th Cir. 2009).

        For an investment-by-investment challenge like this one, a complaint cannot
simply make a bare allegation that costs are too high, or returns are too low. 
Meiners,
898 F.3d at 822
. Rather, it “must provide a sound basis for comparison—a
meaningful benchmark.”
 Id. In Braden, 
a combination of a “market index and other
shares of the same fund” met this requirement, but there is no one-size-fits-all
approach.
 Id. (emphasis added) 
(discussing 
Braden, 588 F.3d at 595
–96).
Plausibility depends on the “totality of the specific allegations in [each] case.”
 Id.
(citation omitted).
                                          1.

       The first investment is TIAA Real Estate Account. According to the
prospectus, its “principal strategy is to purchase direct ownership interests in
income-producing real estate, primarily office, industrial, retail and multi-family
properties.” 3 (Emphasis added). In other words, its holdings consist of numerous
individual properties like malls, office buildings, and apartment complexes.

     The complaint points to another plan offering—the Vanguard REIT Index
Fund—as a possible benchmark, but it is different. In contrast to TIAA Real Estate


      3
       Like the district court, we consider the relevant fund prospectuses and plan
disclosure documents because they are “embraced by the pleadings.” 
Meiners, 898
F.3d at 823
.
                                         -7-
Account’s direct-purchasing strategy, Vanguard REIT Index Fund “employs an
indexing investment approach designed to track the performance” of an index
“composed of stocks of publicly traded equity real estate investment trusts.”
(Emphasis added). It invests in intermediaries, various real-estate companies, which
in turn manage a portfolio of actual real-estate assets. TIAA Real Estate Fund, by
contrast, just cuts out the middleman and holds the real-estate assets itself. This is
the first difference between the two investments.

       The second is that their management approaches differ. TIAA Real Estate
Account is actively managed—meaning that professional investment managers try
to beat the market through picking individual investments—and Vanguard REIT
Index Fund is not. See Downes & 
Goodman, supra, at 9
–10. In fact, the latter is
indexed, which is a form of passive management that tries to mimic a market index,
in this case one full of real estate investment trusts. See
 id. at 342. 
Some investors
like having direct exposure to real estate, and analysts continue to debate whether
active or passive management is a better approach. See generally Dan M. McGill et
al., Fundamentals of Private Pensions, 788–89 (8th ed. 2005) (discussing the
advantages and disadvantages of index funds).

       It is true that some analysts think it is better for investors to put their money
in real estate investment trusts rather than real estate itself. And it might even be
better still to do so through an index fund rather than an actively managed portfolio.
But it is not imprudent for a fiduciary to provide both investment options. See
Meiners, 898 F.3d at 823
–24. They have different aims, different risks, and different
potential rewards that cater to different investors. Comparing apples and oranges is
not a way to show that one is better or worse than the other. See
 id. at 823.
                                          2.

        The second investment is CREF Stock Account, which “invests at least 80%
of its assets in [a] broadly diversified portfolio of common stocks.” It is a variable

                                          -8-
annuity, which is technically an insurance contract, but it behaves similarly to a
mutual fund. As the value of the underlying holdings goes up, so does the price of
the contract. See Downes & 
Goodman, supra, at 808
.

        Once more, the complaint points to index funds as a potential benchmark. Yet
again, the complaint is comparing apples and oranges. Index funds are passively
managed, whereas CREF Stock Account has an actively managed component. In
fact, it uses a “combination” of management strategies, including a “proprietary,
quantitative modeling” approach with “mathematical models and computer
programs” in an effort to beat the market. The biggest difference, however, is that
variable annuities can provide a steady stream of income at retirement—what TIAA
calls “income for your life”—and mutual funds do not. Even if stock index funds
will typically outperform a variable annuity like CREF Stock Account, as the
complaint alleges, it is not imprudent to provide options with differing features from
which to choose, regardless of whether some perform better than others. 4 See Renfro
v. Unisys Corp., 
671 F.3d 314
, 327 (3d Cir. 2011).

      In one last effort to find a truly comparable benchmark, the complaint also
points to three actively managed funds: Vanguard PRIMECAP, Vanguard Capital
Opportunity, and Vanguard Diversified Equity. They are closer, but not close
enough. CREF Stock Account has holdings from around the globe, with around


      4
         The complaint claims that the relevant market is captured by the Russell 3000
Index, which tracks the 3,000 largest stocks in the United States by market
capitalization. The plan disclosures treat this index as a close “broad-based
securities market ind[ex].” See 29 C.F.R. § 2550.404a-5(d)(1)(iii). Nevertheless,
we are not persuaded that the Russell 3000 Index, standing alone, is a “meaningful
benchmark” in this breach-of-fiduciary-duty case. 
Meiners, 898 F.3d at 822
. After
all, it is not a fund, much less an actively managed one, and it does not offer the
security of an income stream for life. Cf. 
Braden, 588 F.3d at 595
–96 (considering
a comparison to market indices as one factor among many in allowing a breach-of-
fiduciary-duty claim to proceed).
                                         -9-
30% of its portfolio invested in international securities. The other funds, by contrast,
have a lower percentage of international stocks. With such a comparatively large
portion of its portfolio in international stocks, CREF Stock Account is essentially a
one-stop shop for those investors who want broad exposure to both domestic and
international equities. This does not mean it is necessarily better or worse. It is just
different, even if it has “some similarit[ies].” 
Meiners, 898 F.3d at 823 
(explaining
that plaintiffs may not “dodge the requirement for a meaningful benchmark by
merely finding a less expensive alternative fund or two with some similarity”).

       Moreover, even if one of these other actively managed funds were an
appropriate benchmark, the complaint fails to connect the dots in a way that creates
an inference of imprudence. There is no question, as the complaint alleges, that
CREF Stock Account performed more poorly than they did over certain periods of
time. But fiduciaries are not required to pick “the best performing fund.”