Outcome
Harvesting
t is notoriously tricky to evaluate interven ons intended
to address such complex issues as poverty, educa onal
achievement and community safety. They typically in-
volve diverse actors and pursue very broad, o en conflic ng
goals. Such interven ons tend to have mul ple moving parts
and ac vi es, and almost always generate an unpredictable
The OH process is organized around six (typically iterave) steps:
1. Design: iden fy the primary users of the evalua-
on, develop the evalua on ques ons, and meth-
ods to answer the ques ons
spla er of outcomes. They are well beyond the league of tra-
2. Dra Outcome Statements: review project docu-
few, clearly-defined outcomes.
3. Engage with Informants: interview people who
di onal evalua on prac ces, which focus on tracking a select
Outcome Harves ng (OH) is designed to overcome this
inadequacy. Developed by the very experienced interna onal
evaluator, Ricardo Wilson Grau, OH builds on the ideas of
several other evalua on methods:
 outcome mapping, which targets “behaviorial change”
in a system’s key actors as the cri cal outcome when
tackling complex issues. Examples of behavioral change
are shi s in policies (e.g., a living wage policy); notable
ac ons (e.g., a Chief of Police acknowledging racism
in the police services); and new rela onships (e.g.,
non-profits and local businesses working together on a
menta on and dra ini al outcome descrip ons.
are involved in the ini a ve to review, revise and
expand the ini al outcome descrip ons.
4. Substan ate: verify the outcome descrip ons with
people independent of the interven on but knowl-
edgeable about the outcome. Based on their input,
adjust the outcome descrip on.
5. Sense-Making: classify all outcomes, usually in
consulta on with informants, according to ques on
and/or emerging theme.
6. Facilitate Use: facilitate discussions with the proj-
ect team about the implica ons of the harvest for
future interven ons.
joint training program for jobs in high demand).
The key to effec ve OH – and what dis nguishes it
process to capture all types of results emerging from an
people outside the interven on substan ate the pre-
 most-significant change, which employs a par cipatory
interven on – intended and unintended, posi ve and nega ve, short term and long term. It considers diverse stake-
holder perspec ves on what makes each result significant.
 contribu on analysis, which seeks to get a general sense
of the contribu on of a group’s ac vi es to observed
outcomes. It does not (vainly) try to establish a defini ve
cause and effect rela onship between them.
most from other evalua on methods – is Step 4: having
liminary outcome descrip ons provided by par cipants
of the interven on. This guards against the natural
tendency of those closest to the project to overstate its
posi ve results and understate its weaknesses. It can
also surface ripple effects previously invisible to project
implementers.
The output of OH is a report which includes (a) a descrip-
have a few “harvests” under their belts, it may be useful
assessment; (b) a visual map of the outcomes, including if
with extra exper se in qualita ve research and an out-
on of the purpose, key ques ons and methods of the
and how they are linked, over me; and (c) a more detailed
descrip on of the key outcomes presented in the la er.
Typically, the reports are concise. For example, a report on
a two-year World Bank-sponsored project to improve the
financial viability of local waste management opera ons in
Bosnia ran merely ten pages. It outlined 18 key outcomes,
including changes in poli cal commitment, (e.g., the local
police force agrees to enforce fee collec on); policies (e.g.,
a municipal increase in waste collec on fees by 10%); and
new working rela onships (e.g., joint databases between
several municipali es).
While the steps and outputs of OH are simple, the design and
implementa on are demanding. It requires me, resources
to get the support of an external outcome harvester,
side perspec ve.
There s ll are plenty of ways to develop and refine a
method as young as OH. Here are two, based on my re-
view of ten applica ons of OH to World Bank-sponsored
projects: (a) iden fy the techniques that groups might
use to es mate the contribu on of project ac vi es to
outcomes; and (b) create a set of principles and prac ces
to guide the substan a on of outcome statements. Happily, some par cipants of the Outcome Mapping Learn-
ing Community – the world-class community of prac ce
from which OH par ally emerged – have already turned
their a en on to these and other ma ers. The future of
OH is in good hands.
and exper se to se le on produc ve evalua on ques ons,
Many important interna onal development agencies
independent observers to verify the outcome statements.
ua on) now use OH. No wonder: the method weaves
develop ght outcome descrip ons, and iden fy and engage
Social innovators may feel that they can complete most of
the work in-house. Like as not, this is premature. Un l they
This is one is one in a series of
What We Know So Far documents that summarize some of
the latest thinking or developments in the field of social innova on and community change.
Sources:
(a clientele with exac ng standards in ma ers of evaltogether several solid evalua on methodologies in an
accessible, robust and useful way. Anyone serious about
social innova on, community change and evalua on will
make a point of experimen ng with OH in 2016.
The Methodology: http://betterevaluation.org/plan/approach/outcome_harvesting
Ten Case Studies: http://betterevaluation.org/resources/example/cases_in_
outcome_harvesting
Outcome Mapping Learning Community: http://outcomemapping.ca