Service · Performance & feedback

Downward reviews for US law firms

Partner evaluations of counsel, of counsel and associates, gathered from everyone who actually staffed the work rather than the one partner whose name sits on the file.

Who this is for

  • Managing partners and practice group leaders
  • Chief People Officers and Directors of Professional Development
  • Firms whose evaluation forms have not changed in a decade

One reviewer, recalled at year end, is not an evaluation

A downward review is attributed by design. The associate is told who wrote it, because a judgment that shapes compensation and progression has to have an author. That is the opposite of an upward review, and it changes what the instrument has to protect against.

What it has to protect against is a single partner writing from memory. The NALP Foundation surveyed 106 leading firms in November 2025 and found supervising attorney comments in every evaluation program. The harder question is how many partners contribute, and on what exposure. An associate who billed 400 hours across six partners and is assessed by one of them has been evaluated on a sixth of their year.

The written half is where it goes wrong quietly. Harvard Business Review research found 76 percent of women's reviews contained negative personality feedback against 2 percent of men's. That is a finding about reviews in general rather than law firms specifically, but the mechanism is the same: unstructured prose invites comment on the person instead of the work. NALP found 30 percent of firms never compare evaluation data across any cohort, so a firm with that pattern has no way of seeing it.

How a cycle runs

A downward cycle is a reviewer-assignment problem before it is a questionnaire problem. Most of the design work happens before anyone is asked to write anything.

  1. Weeks 1–2Competency framework mapped to practice area and seniority, then approved by the firm. Rating scales anchored in observable behavior rather than adjectives.
  2. Week 3Reviewer assignment from matter and hours data, so each reviewee is assessed by the partners who actually supervised the work. Self-assessments open at the same time.
  3. Weeks 3–5Reviewers complete, with automated reminders and live completion reporting to the firm. Written comments prompted against specific competencies rather than a blank box.
  4. Week 6Aggregation, outlier flagging and cohort analysis by class year, practice group, office, gender and race where the firm collects it.
  5. Weeks 7–8Calibration pack to practice group leaders, then individual reports and a findings readout to leadership.

What you receive

Individual evaluation reports

Ratings, written comment and self-assessment against the same competencies, with benchmark context.

Self versus reviewer gap analysis

Where an associate rates themselves well above or below their reviewers, which is usually the most useful conversation in the cycle.

Calibration pack

How each partner's ratings sit against their practice group, so a consistently generous or harsh reviewer can be shown it.

Cohort analysis

Results compared across class year, practice group, office and, where the firm collects it, gender and race.

Development plans

Findings carried into a plan for each reviewee, with owners and a revisit date.

Readout to leadership

A working session on what the distribution shows about the firm, not only about individuals.

What changes when the evaluation is built rather than inherited

The form most firms still use

An SRA downward review

One nominal supervising partner writes the review
Every partner above a matter-hours threshold is asked, and light exposure is excluded rather than averaged in
A blank comment box at the end of the form
Written comment prompted against each competency, so the evidence sits next to the rating it supports
Ratings mean whatever each partner takes them to mean
Behaviorally anchored scales, with a calibration pack showing each partner how their ratings sit against the group
Results read one associate at a time
Cohort analysis by class year, practice group, office and demographic, which is where a pattern becomes visible
The review is the end of the process
Findings feed a development plan with named actions and a date to revisit them

Confidentiality

How the data is held, in practice

The rules below are the reason participation holds and the reason findings are worth acting on. They are enforced by where the data sits, not by policy alone.

Read the confidentiality pledge →

Attributed by design

A downward review is not anonymous, and it should not be. The reviewee is told which partners assessed them, because a judgment used in compensation and progression decisions has to have an author who can be asked about it.

Independent administration

SRA holds and processes the responses as an independent party under a data processing agreement, with your firm as controller. That keeps the record intact and the process consistent across reviewers, rather than sitting in a partner's inbox until the deadline.

Self-assessments held until reviewers submit

An associate's self-assessment is not visible to their reviewers until those reviewers have filed, so the rating is not anchored to what the associate said about themselves.

Cohort analysis reported in aggregate

Individual reviews are attributed. Analysis across class year, office, gender and race is reported at the group level, against a threshold agreed before the cycle opens.

Certified handling

ISO 27001 certified. SOC 2 Type I available under NDA. GDPR Article 28 and CCPA aligned, with SRA as processor and your firm as controller.

Why firms use an independent party

Every figure below is recorded in an evidence register with its definition, source and period. Illustrative figures used elsewhere in the design are excluded here by design.

35 years serving US law firms
Years of continuous operation, 1991 to present.
100+ elite Am Law firms as clients
Count of distinct Am Law-ranked firms that have engaged SRA. Cumulative, not concurrent.
100,000+ confidential surveys conducted
Count of individual survey responses collected.
1,000,000+ performance reports delivered
Count of individual performance reports produced and delivered.
ISO 27001 certified
Certified information security management system. Certificate number and scope available on request.

Performance & feedback

This is one program inside a broader practice

This is a workflow your firm shapes, not a fixed product. Firms run it for the populations, cadence and reporting they agree, and give the program whatever name suits them. It sits inside Performance & feedback.

What you can tailor

  • Who is reviewed, and who reviews them
  • Competencies and questions
  • An annual cycle, or another cadence you agree
  • Who sees which report

What an agreed scope covers

An engagement covers the review types, the populations and the cycle you agree. Adding a review type or a population is a change of scope rather than something a practice includes by default.

These are not one instrument. A downward evaluation is attributed to the partner who wrote it, because the associate needs to know whose judgment it is. An upward review is confidential and reported only above an agreed threshold, because an associate assessing the partner who staffs their matters answers honestly only when identification is impossible. Both are explained to participants before they answer.

How to run it

Run with your team

Use Artemis to configure and run the agreed program, with onboarding and standard support. Your team owns administration and follow-through.

Explore Artemis →

Have SRA manage the program

Agree the design, administration, analysis and readouts you want SRA to provide. Your firm retains the decisions and actions that follow.

Discuss a managed program →

Running one program in Artemis does not commit the rest of the firm to the same arrangement. You can choose differently for different work.

Send us your current evaluation form