For IR Directors

Trusted, sourced answers across your data and the federal datasets you pull on

Directors of Institutional Research use Clema to bring IPEDS, peers, and their own institutional data into one plain-English conversation, with the source shown on every figure.

A day in the life of an IR office

Not a hypothetical. This is roughly what a week looks like for a 1 to 3 person IR shop, drawn from the patterns in our interviews with 50+ IR professionals across 19 states.

Monday, 8:40 a.m.

The board packet request with no cohort attached

The provost's chief of staff emails asking for retention rates by Pell status for the last five years, needed for Friday's Board Academic Affairs packet, with no cohort definition attached.

You build a first pass and send it back. An hour later: no, we meant first-time, full-time only, and can you also split by college. Two more rounds follow before the number is locked, days before the actual meeting.

With Clema: You ask Clema for the figure in plain English. It resolves against your locked first-time, full-time retention definition and asks the one clarifying question that matters (which college breakdown) instead of guessing. The number is locked the same afternoon, not Thursday night.

IPEDS week

A federal deadline collides with three unrelated inboxes

It's Fall Enrollment and Graduation Rates component week. You're cross-checking every submitted cell against last year's snapshot to catch drift, while a program review number, an assessment committee ask, and a reporter's public-records request land in the same inbox and can't wait until the federal deadline clears.

With Clema: The three unrelated requests get answered directly by the people who asked, sourced to the same data your IPEDS submission draws on. You keep your attention on the drift check instead of splitting it three ways.

Mid-cycle accreditation

A number from three years ago, with the analyst long gone

The visiting team asks for the exact cohort logic behind a persistence figure reported three years ago. The analyst who built that query left the office eighteen months ago and didn't document the exclusion rules, so the number has to be reverse-engineered from an old spreadsheet before the site visit.

With Clema: The definition and exclusion rules behind that figure are stored with the answer, not with the person who ran the query. You pull the methodology in minutes instead of reconstructing it from a spreadsheet nobody remembers building.

The night before

A budget hearing that needs peer numbers by morning

The VP of Enrollment Management asks for a quick comparison of admit-to-enroll yield by program against three named peer institutions, the night before a budget subcommittee hearing. It isn't quick: the peer numbers live in College Scorecard and IPEDS, the internal numbers live in the SIS, and reconciling naming conventions between them eats the evening.

With Clema: Peer figures from IPEDS and College Scorecard sit in the same conversation as your SIS data, so the reconciliation work Clema already understands the naming conventions on both sides doesn't become your evening.

How Clema helps

The work that eats your week, handled in minutes.

Sourced answers, every figure

Every response carries its data source, calculation method, and audit trail. You ship figures you can defend to deans, accreditors, and the cabinet without a second pass.

Federal data in the same conversation

IPEDS, College Scorecard, Pell, EADA, and PSEO sit alongside your institutional data. Ask a peer-comparison or completion-rate question and get the federal source and year shown.

Strategic analysis, accessible

Clema understands higher-ed terminology out of the box, so your team spends its time on interpretation and recommendation rather than wrangling definitions and joins.

The old way vs. the Clema way

None of these are exaggerated. They're the pattern our research kept surfacing across IR offices of every size.

The old way

A vague board request kicks off 2 to 5 rounds of email clarification before the number is locked, days before it's due.

With Clema

You ask in plain English. Clema resolves the question against your locked cohort definition and asks the one clarifying question that matters, inline.

The old way

The retention number lives in three places (the warehouse, an analyst's shadow spreadsheet, a two-year-old dashboard) and they don't quite agree. Our whitepaper found 82.4% of IR offices report data fragmentation across multiple systems.

With Clema

One definition-locked source of truth. Every answer is sourced back to the system and methodology it used, so the same question gets the same number twice.

The old way

The analyst who wrote the cohort exclusion logic left eighteen months ago and never documented it. The number has to be reverse-engineered.

With Clema

Definitions, exclusion rules, and prior answers live in Clema, not in one person's head or an orphaned spreadsheet.

The old way

A dean can't find the right filter on the dashboard IR built for exactly this question, doesn't fully trust what he sees there anyway, and emails IR for a manual pull instead.

With Clema

Same governed data, asked in plain English, with the source shown so the dean doesn't have to just take IR's word for it.

See Clema answer a real IR question

An illustrative example of the kind of drill-down a director actually needs before a board packet. Your figures will use your institution's own data and definitions.

Clema
What was our first-time, full-time retention rate for the fall 2023 cohort, and how does it compare to the prior three cohorts?
Using the IPEDS Fall Enrollment (Retention Rate) definition your office has on file, first-time, full-time retention for the fall 2023 cohort was 82%, based on fall-to-fall persistence of that cohort's first-time, full-time degree-seeking students. Compared with fall 2020 through fall 2022, the rate has moved from 79% to 82%, a gradual upward trend. Source: your institution's submitted IPEDS Fall Enrollment component, cross-checked against the SIS enrollment extract refreshed this morning. Want this broken out by college or by Pell status next?

Example figures for this walkthrough · First-time, full-time retention

CohortRetention rate
Fall 202079%
Fall 202180%
Fall 202281%
Fall 202382%
Yes, break it down by Pell recipient status, and flag anything I shouldn't report because the group is too small.
For the fall 2023 cohort: Pell recipients retained at 76%, non-Pell recipients at 85%. One subgroup, first-time Pell recipients in the College of Engineering, has fewer than 10 students, so I'm flagging that cell for suppression consistent with your small-n disclosure policy rather than reporting a specific rate. Source: SIS enrollment and financial aid extract, same refresh as above.

Example figures for this walkthrough · Fall 2023 cohort

GroupRetention rateN
Pell recipients76%412
Non-Pell recipients85%1,180
First-time Pell, College of EngineeringSuppressed (n<10)8
I need this in the Board Academic Affairs Committee format from last year. Can you show me the definition footnote we used then so the numbers line up?
Here's the footnote your office used in the fall 2022 board packet: first-time, full-time, degree-seeking undergraduates per the IPEDS Fall Enrollment component, excluding students who died or entered military, religious, or foreign-aid service, using the fall census date snapshot. I've applied the same definition and exclusions to this year's figures so they're directly comparable, and I can export the table in the same layout for the packet.

What ad-hoc requests actually cost your team

From our study of 50+ IR professionals across 19 states. Read the whitepaper.

40-60%

of team capacity goes to ad-hoc data requests instead of strategic analysis

73.5%

of requests arrive vague or unclear, taking 2 to 5 rounds of clarification before fulfillment

220 to 3,200 hrs/yr

reclaimable depending on team size: 220 to 330 hours for teams of 1 to 3, up to 2,133 to 3,200 for teams of 8 or more

What the research says

This isn't just our whitepaper. The pressure on this role shows up in independent research from four different angles.

~Half

of IR office leaders say their staff size isn't sufficient for their current workload; only 56% call their process for managing that workload adequate.

Association for Institutional Research (AIR)

2025 → 2041

the national count of high school graduates peaks in 2025, then declines roughly 10.5% below 2023 levels by 2041, unevenly across regions.

WICHE, Knocking at the College Door (11th ed.)

29%

of provosts say they consistently have the resources to implement institutional initiatives, even though 86% say they like the job.

Inside Higher Ed, 2025 Survey of Chief Academic Officers

34%

of research and institutional research staff cite job security as a top reason they're looking for other jobs, above the higher-ed-wide average.

CUPA-HR, 2025 Higher Education Employee Retention Survey

We ran our own primary research too: 85% of institutions depend on a single person for institutional knowledge. See the Institutional Intelligence Gap study.

No rip-and-replace

Where Clema fits with what you already have

Clema sits on top of the systems you already run and the public federal datasets you pull on. It doesn't replace any of them, and it doesn't ask you to migrate data anywhere.

Clema

Conversational layer. Ask in plain English, get a sourced answer.

sits on top of, doesn't replace

Your systems

SISData warehouseTableauPower BIExcel

64.7%

of IR offices report their self-service dashboards go underused

82.4%

report data fragmented across multiple systems that don't talk to each other

"The issue is not the absence of tools; it's the absence of integration."

Clema is the layer that makes the infrastructure you already have actually get used: one plain-English question, answered against your existing systems, with the source shown.

What Clema gives your team

Sourced

Every figure carries its origin and method

One window

Federal and institutional data together

Higher-ed native

IPEDS, accreditation, and IR terms built in

IR Director FAQs

What does Clema give an IR Director?

A single conversational surface over your institutional data and the federal datasets you pull on, with sources shown on every figure. Ask for completion rate, FTE, or graduation rate by cohort in plain English and get the calculation with the NCES source attached.

Does Clema replace our BI tools and dashboards, or the SIS?

Neither. Clema sits on top of your SIS, data warehouse, and BI tools like Tableau and Power BI rather than replacing them.

Our whitepaper research found 64.7% of IR offices report their self-service dashboards go underused, often because people don't know where to look or don't trust what they find there, not because the dashboards are badly built.

Clema gives people a plain-English way into the same governed data instead of asking every dean and VP to become a dashboard power user.

If two people in different offices ask the same question, will they get two different numbers?

No, that's the specific failure mode Clema is built to close.

Clema resolves questions against your locked definitions (your official first-time, full-time retention cohort, or your IPEDS graduation-rate methodology) rather than whichever spreadsheet or dashboard filter a given user happens to land on.

Every answer is sourced back to the definition and dataset it used, so a dean and a provost asking the same question in the same week get the same number, with the same citation.

How does Clema handle IPEDS, accreditation, and state reporting cycles, not just ad hoc requests?

Clema connects directly to public federal datasets, including IPEDS, College Scorecard, Pell, EADA, and PSEO, alongside your internal systems. With that in place, you can:

  • Cross-check a submitted IPEDS figure against prior-year values
  • Pull the historical methodology behind a number an accreditation site visit team is asking about
  • Ask for completion rate, FTE, or graduation rate by cohort and get the calculation with the NCES source shown, without rebuilding the query by hand
How do we verify an answer before it goes into a board packet or accreditation report?

Every Clema answer comes with a visible source trail: which dataset or system it pulled from, which definition or IPEDS methodology it applied, and when the underlying data was last refreshed. That means your staff review the sourcing, not re-derive the number from scratch, before it goes to a provost, board committee, or accreditation site visit team.

What happens to institutional knowledge if our IR director or sole analyst leaves?

This is a known risk in the field, not a hypothetical one.

Our companion research on 20 IR/IE professionals found 85% of institutions depend on a single person for institutional knowledge, and 55% sit in what we call the Large Gap tier, where core definitions are effectively one resignation away from being lost.

Clema keeps definitions, cohort logic, and prior answers in a system of record rather than in one person's head or an undocumented query, so a departure doesn't mean rebuilding institutional memory from scratch.

How does Clema handle data governance and security for IR-owned definitions?

Clema is SOC 2 Type II certified and FERPA-ready by design, built on:

  • Read-only credentials
  • SAML 2.0 SSO
  • TLS 1.3 encryption
  • Full audit logging on every query your team runs

Clema doesn't create a new copy of student-level data outside your environment's security boundary; it works within the access controls you already have.

Your team keeps ownership of the underlying metric definitions; Clema makes them consistently askable instead of drifting across shadow spreadsheets.

How long does it take to start getting value from Clema?

Most IR teams connect their SIS, warehouse, and IPEDS extracts and are live in two to four weeks, including analyst training. Start free on a sample data source first, so your team sees a sourced answer before committing to the full connection.

Get started

Try Clema on your data

Start free on the Free plan, or book a demo with IR scenarios built for your institution.