Blackbaud Database View Sunset: Raiser's Edge 2027

Blackbaud ends support for the Raiser's Edge NXT database view on September 15, 2027. Blackbaud Online Express, Blackbaud NetCommunity, and every database view plug-in, integration and customization end with it. Your donor data does not move. What changes is any export, import, report, plug-in or feed that only runs in the Citrix client.

This guide is for the advancement services or database manager at an independent school, university or nonprofit who has to turn that announcement into a plan. Where your contract or your own setup decides the answer, it says so.

When does the Raiser's Edge database view end?

September 15, 2027. Blackbaud's help center transition page says it plainly: "Database view-related solutions and functionality will end on September 15, 2027."

Tiffany Crumpton, Blackbaud's SVP of Fundraising, announced the date in early September 2026, and Trit Mulligan, Director of Product for Raiser's Edge NXT, wrote it up for customers in a Blackbaud Community post on September 3, 2026. If your plan was built on an earlier date, update it. Before that announcement, consultants were working from "the first half of 2027"; Heller Consulting's May 2025 summary showed June 30, 2027. You may see the same thing called the "RE 7 sunset," because Blackbaud's own help pages describe database view as Raiser's Edge 7.

One regional caveat. SmartTHING, a UK Blackbaud partner, reports that the September 15 date applies to the UK, Europe and North America, with Australia and New Zealand following later in 2027. Blackbaud's help page lists a single date, so organizations in Australia and New Zealand should confirm theirs with Blackbaud.

The same announcement listed what Blackbaud still has to ship in web view. These were its targets as of September 2026; check Blackbaud's What's Recent release notes to see which have landed.

Data managers should plan around the last row. If your office depends on Global Change or on an import type other than constituents and gifts, the web view version may not arrive until June 2027. That leaves about eleven weeks to rebuild and test before the cutoff.

What exactly is retiring, and what stays?

The database view is the Windows client many people still call RE 7. Blackbaud's help page puts it this way: "Raiser's Edge is the database view accessed through Citrix, and Raiser's Edge NXT is web view accessed from your web browser." Both read the same data. After September 15, 2027, only the browser side remains.

Retiring with it, per the announcement:

• The database view client and its modules as they exist today: Query, Export, Import, Global Change, Mail and Reports.
• Blackbaud Online Express and Blackbaud NetCommunity.
• Database view-based plug-ins, integrations and customizations. That covers anything that installs into or automates the Citrix client, whether Blackbaud wrote it, a partner sold it or your own staff built it.

Staying put:

• Your data. In Blackbaud's words, "there is no data migration: your constituent records, gift history, and organizational information stay exactly where they are."
• Backups. In a February 2026 post, Backup Support Remains Unchanged, Blackbaud said customers "can continue relying on Bulk Data Delivery and On-Demand backups" after database view is retired. The same post says Blackbaud expects customers to move over time toward "API-driven, cloud-aligned approaches." In a May 2026 follow-up, Mulligan said longer-term changes to backups would come "with at least 12 months notice."
• SKY API, which reads and writes the same data.

The published material is about Blackbaud-hosted database view. If you still run Raiser's Edge 7 on your own server, Heller noted in 2025 that Blackbaud had not published a separate end-of-support date for those installs, and we found none since. Ask your account team for your support status in writing.

What breaks on September 15, 2027?

Anything that runs only inside database view stops that day. Some things already live in both views. Blackbaud's web view Query FAQ says every database view query you can access also appears in web view, with edits synced both ways, except for query types web view doesn't support. Other things, like Export definitions and plug-ins, have no automatic path.

Watch the vocabulary when you rebuild. Attributes are custom fields in web view, annotations are alerts, proposals are opportunities, and solicit codes become communication preferences. Blackbaud's Database View vs Web View guide has the full list.

The same guide lists data that exists only in web view: solicit code dates and comments, prospect and opportunity status history, email history, attachments added through web view, RSVP status and JustGiving fields. Any process built on the database view, including a restored backup, has never seen those fields.

Plug-in vendors face the same deadline. In February 2026 a Blackbaud ISV partner whose product depends on volunteer data asked publicly for a Volunteer API timeline. Blackbaud has since moved: the SKY API changelog shows a Volunteer API in public preview from March 3, 2026, with breaking route changes in April and October 2026, and several Membership API endpoints promoted to long-term support in July 2026. Preview APIs still change, so ask each vendor whether their product already uses them.

What should you inventory before September 15, 2027?

Start with anything that runs on a schedule or feeds another system, because those fail silently. Then work down to the tasks one person does by hand.

1. Name a transition owner. Blackbaud's announcement asks each customer to pick a web view transition leader, and the work needs one person who can say no to new database view builds.
2. List every saved query, with owner, last run date and what consumes its output. Flag the unsupported types listed above.
3. List every Export module definition and its destination: mail house, annual report, board packet, alumni office, finance, a vendor upload.
4. List import routines and their sources: event registrations, online gifts, payroll deduction files, matching gift files, alumni records from the registrar.
5. List every plug-in and partner add-in installed in database view, and get a written plan and date from each vendor.
6. Map every Online Express and NetCommunity function you use (forms, portals, directories, email) to a replacement, and export what you need to keep.
7. List reports built outside Raiser's Edge (Excel, Power BI, SQL) and trace each back to its source: export file, backup or API.
8. Test receipting in web view once the December 2026 features ship. Compare record counts, totals by fund, receipt number continuity and letter content against a database view run of the same batch.
9. Confirm who holds the Query Export and Backup Administrators permissions, and that more than one person can run each critical job.
10. Decide on backups. Price Bulk Data Delivery (Blackbaud does not publish pricing; ask your account team), or schedule monthly on-demand backups and download each within seven days.

Blackbaud's roadmap dates suggest an order of work:

If the sunset is pushing you to consider leaving Blackbaud altogether, the same inventory is the core of a replacement RFP. Every query, export and feed on the list is a requirement the next system has to meet.

What are the SKY API options for getting data out?

There are three useful routes. The Query API runs saved or ad hoc queries and returns a file. The entity APIs (constituents, gifts and so on) handle record-level and incremental reads. The Webhook API sends near-real-time change notices. Every call needs your subscription key in the Bb-Api-Subscription-Key header, per Blackbaud's SKY developer help, plus an OAuth 2.0 bearer token for a user who has authorized your app.

Query API

The Query API is the closest replacement for "run my query and give me a file," and Blackbaud's Query FAQ says its use is included with your Raiser's Edge NXT subscription. Blackbaud's Query connector definition shows the flow. You start a job with POST /query/queries/executebyid for a saved query or POST /query/queries/execute for an ad hoc definition. Then you poll GET /query/jobs/{job_id} until the status is Completed, and download the file from the returned sas_uri. Output can be CSV, JSON or JSONL, a choice Blackbaud added in April 2024. A sql_generation_mode of Export returns only your output columns, while Query adds the QRECID key. The definition warns that the results URL "will expire after 15 minutes," so download right away.

The user behind the token needs the Export permission under Query tasks in the Analysis role, which Blackbaud added in a May 2024 update. Since June 15, 2026, the Query API also supports single-row queries, the replacement for Export, in preview. Not all fields and filters work yet, and single-row queries live only in web view.

import time
import requests

BASE = "https://api.sky.blackbaud.com/query"
PARAMS = {"product": "RE", "module": "None"}
HEADERS = {
    "Bb-Api-Subscription-Key": SUBSCRIPTION_KEY,
    "Authorization": f"Bearer {access_token}",
}

# Start a job for saved query 1234 (user needs Query > Export permission)
start = requests.post(
    f"{BASE}/queries/executebyid",
    params=PARAMS,
    headers=HEADERS,
    json={
        "id": 1234,
        "ux_mode": "Asynchronous",
        "output_format": "Csv",
        "sql_generation_mode": "Export",
    },
    timeout=30,
)
start.raise_for_status()
job_id = start.json()["id"]

# Poll until the job finishes, with an overall deadline
deadline = time.monotonic() + 30 * 60
while True:
    resp = requests.get(
        f"{BASE}/jobs/{job_id}",
        params={**PARAMS, "include_read_url": "OnceCompleted"},
        headers=HEADERS,
        timeout=30,
    )
    resp.raise_for_status()
    job = resp.json()
    if job["status"] in ("Completed", "Failed", "Cancelled"):
        break
    if time.monotonic() > deadline:
        raise TimeoutError(f"Query job {job_id} still {job['status']}")
    time.sleep(15)

if job["status"] != "Completed":
    raise RuntimeError(f"Query job {job_id} ended as {job['status']}")

# The SAS URL is a short-lived credential (15 minutes). Download now.
download = requests.get(job["sas_uri"], timeout=300)
download.raise_for_status()
csv_bytes = download.content
expected_rows = job.get("row_count")

Check the field names against the endpoint reference for your subscription before you run this in production. The Query API started as a public preview, and its defaults have changed since; the changelog is where those changes show up.

Entity APIs, webhooks and quotas

To keep a copy current, pull only what changed. Blackbaud's synchronization guide says to call the list endpoints with last_modified, follow the next_link, and reuse the supplied next_link and sort_token so no records are missed. It adds two warnings. List endpoints do not report deleted records, and many lists won't show new changes in less than about 30 minutes because of internal sync times. The Gifts API definition shows GET /gift/v1/gifts with last_modified and pages of up to 5,000 records. Use the focused endpoints (addresses, consents, custom fields) for child data. The Webhook API sends a small message with record IDs when, for example, a gift is added, and your app fetches the details.

Budget for quotas. Blackbaud's tier page lists 25,000 calls per day on the Standard tier and 1,000 on the Free tier, with paid tiers at 100,000 and 250,000. Its rate limit page sets 10 calls per second, returns a 403 when the daily quota runs out, and resets at midnight UTC. Do the big historical load from a backup or query file, and use the entity APIs for daily changes.

Can you rely on Blackbaud's backups instead?

Partly. Backups stay supported after September 2027, and they are the only route that gives you a SQL Server copy of the database. Blackbaud's on-demand backup page also says they "capture database view information only and are not intended for full database restoration."

On-demand backups come with limits. A Blackbaud-hosted organization can request one per month, and each expires after seven days. Only an organization administrator or a user with the Backup Administrators role can create the request. Bulk Data Delivery is the separately contracted option for scheduled backups. A Raiser's Edge user who automated it describes downloading a nightly copy of most of the database and restoring it to their own SQL Server.

The restored copy uses the RE 7 schema, a large model with no Blackbaud documentation aimed at customer reporting. Hosted customers have asked for years for read-only SQL access to the live database and have fallen back on backups because that access isn't offered. Plan for a mapping phase: record and gift tables, code tables and table entries, attributes, and gift splits by fund, each checked against counts and totals from web view. AI-assisted analysis speeds up the profiling and join inference, but the checks against known totals are what make the mapping trustworthy. We describe that general method in our guide to legacy integration and reverse engineering with AI. Your Blackbaud agreement governs what you may do with their software and files. This is not legal advice; review your license agreement and MSA with counsel. See also whether reverse engineering software is legal.

Web view-only data is the larger gap. Email history, opportunity status history and the other fields listed earlier never reach the backup. For each one you need, confirm that an API endpoint or a web view query field exposes it, and raise any that don't with Blackbaud before 2027.

How do you build a donor data warehouse you own?

This step is optional. It is the one that keeps your reports and feeds working through the next interface change. Load history from a backup, add web view-only fields and daily changes through SKY API, and keep the result in a database your institution controls. Your feeds will still depend on Blackbaud's APIs, but your reports and downstream systems read from your copy.

A layout that works for a school or mid-sized nonprofit:

1. Restore a backup to your own SQL Server as a read-only raw history layer.
2. Build a staging layer that maps RE 7 tables to plain business entities: constituent, gift, gift split, fund, campaign, appeal, relationship, education, custom field. Use the system record ID as the join key everywhere. Web view Query can now output Import ID and System Record ID, which makes it easier to match file rows to API records.
3. Run a nightly job against the list endpoints with last_modified, following next_link and storing the sort_token for the next run. Mark records inactive rather than deleting them, since the API won't tell you about deletes.
4. Pull the web view-only fields you need through the API or a scheduled query job.
5. Check every run. Compare row counts with the Query API's row_count, total gifts by fiscal year against a standard report, and alert on drift.

We used the same pattern on a K-12 project that pulled five student information systems into one pipeline, where latency fell from four to six hours to under three minutes and the error rate fell from 3.8% to 0.02%, as described in our EdTech data ingestion case study. Schools moving student systems at the same time can compare notes with our posts on getting data out of PowerSchool and the Skyward Qmlativ migration, and our education industry page covers the wider set of systems schools are consolidating.

Where to start this month

Open web view Query and list the queries it can't run, then pull your Export definitions and sort both by how often they run. Pick the five that feed something outside Raiser's Edge and rebuild those first, while database view is still there to compare against. If you'd rather have someone build the warehouse and the automated feeds, our workflow automation team and education technology developers do this work, and you can reach us here.