Know which query and index changes are worth testing.
For teams that can share query text, plans and workload context, then test changes or obtain vendor approval.
Get tuning recommendations in priority order for the selected statements. It explains the proposed changes, their trade-offs and how your team can test them and plan a rollback.
Your team has a few important queries to improve. Get recommendations for those statements, with the reasons, trade-offs and tests needed before you change anything.
You collect. We assess.
Self-Service
Your team runs the BriteThought Diagnostic Collector in the agreed environment, reviews the package and sends the results. We provide expert analysis and the written assessment. BriteThought doesn't need remote access or your credentials.
Standard
$995
Priority Same scope, reserved scheduling
$1,395
USD · excluding applicable taxes · per engagement
You receive the expert-reviewed report and one written review of collection status. When useful, we also include one request to rerun failed or missing modules or provide a limited alternative.
We don't troubleshoot your permissions, security, network or execution environment in Self-Service. Remote sessions and a findings call aren't included; you receive the report in writing.
Standard target. Written delivery target: within five business days after complete intake and sufficient evidence are confirmed.
Priority target. Analysis slot within one business day after sufficient data is admitted; written report target two business days after data admission.
You arrange appropriate access through your approved process. We collect the evidence and carry out the same core assessment, with selected live validation.
Standard
$1,695
Priority Same scope, reserved scheduling
$2,445
USD · excluding applicable taxes · per engagement
You receive the same core written deliverables, up to 2 hours total of collection and the listed live validation or investigation, plus one 30-minute findings call. Recommendations aren't permission to change production.
Your team approves access and makes local security decisions. Production changes and remediation need a separate agreement. Access restrictions can still prevent collection.
Standard target. Collection is normally scheduled within five business days after scope, payment and working access are confirmed. The written report target is five business days after complete intake and sufficient evidence are confirmed, once collection is complete.
Priority target. Collection start within two business days after working access and scope are confirmed; written report target two business days after collection.
This is a one-time assessment with a fixed scope and the listed deliverables. Production changes and remediation are separate. With Self-Service, your team collects the evidence; customer-specific collection troubleshooting is excluded.
Priority requires confirmed availability before payment. We’ll confirm whether the service fits, the scope and the dates before payment. This website does not accept payments or diagnostic uploads. See when work starts and what can pause it.
One instance, one named user database and up to five selected SQL statements or individual statements within stored procedures. A procedure with many statements doesn't count as one unlimited query.
We review up to five associated execution plans and the relevant index and statistics definitions, and propose up to five change items in total.
One workload capture window of up to 60 minutes, plus up to seven days of existing history. You identify representative parameters and data distribution; you can replace confidential values with useful, safe examples.
Managed includes up to two hours total of collection and live validation, plus a 30-minute findings call. Live validation checks the evidence; it doesn't apply tuning changes.
We may use instance-wide metadata for context. Detailed review of unnamed databases, additional instances, replicas or environments is excluded.
Your report
Findings, reasons and next steps.
Tuning recommendations and how to test them
You receive a ranked tuning memo for the selected statements, query, index or statistics recommendations where the evidence supports them, and a test and rollback approach. Candidate scripts are included where practical and where an applicable, validated standard template supports a material recommendation, with testing and rollback guidance. They are reviewed recommendations and still need testing and approval for your environment. We label unsupported changes and assign them to the vendor or application owner.
Review the recommendations before making changes
Tune-Up means a tuning assessment and proposed changes. Recommendations are included. Candidate scripts are included only where practical and where an applicable, validated standard template supports a material recommendation. Implementing them is a separate purchase.
Both modes include an executive summary, the evidence sources and their limits, and recommendations that identify who should act, the risks and what needs testing. We supply candidate scripts only when an applicable, validated standard template supports a material recommendation. Custom scripts, code rewrites and execution are excluded unless expressly listed. Send up to five report questions together within ten business days of delivery for one included round of clarification. Managed includes one 30-minute findings call. Self-Service includes written delivery; a call can be purchased separately.
When is the assessment complete?
We complete the service when we've reviewed the selected statements and available evidence and delivered the tuning memo. Finding no justified change is a valid result. We won't claim a before-and-after improvement unless a separately authorised test measures comparable workloads.
A time limit alone does not complete the service. We must document the investigation or executed test, its result and its limits. If required collection fails before we accept the evidence for analysis, the collection, upgrade and refund rules apply. An empty report is not a successful diagnosis.
What we’ll need
Your environment, the question and access.
The standard service covers one reachable primary SQL Server Database Engine instance on Windows, on premises or in a virtual machine you control. Its release and edition must be compatible with the current collector. A named instance counts as a separate instance.
Before you pay, we'll confirm the metadata permissions, approved host diagnostics and collection window, along with the exact permissions for your SQL Server version, runtime requirements, disk space and tested compatibility. Sysadmin or local administrator access isn't a requirement for every assessment.
Complex availability groups, failover clusters, replication, distributed applications, Azure SQL Database, Azure SQL Managed Instance, Linux, inaccessible hosting, unsupported versions, extra instances, replicas or environments need a review before we quote separately. Enterprise edition may qualify if the technical scope stays small.
Agree the five statement units before you pay. We need query text, relevant plans, index and statistics information, workload frequency, expected behaviour and application or vendor restrictions. Query IDs alone may not be enough. If you can't share enough structure or query evidence, we'll check an alternative before accepting the order.
Identify whether each plan is estimated or actual, and whether parameter values or row counts are representative. Your team supplies a suitable test environment or takes responsibility for implementation testing after the report. The base assessment doesn't require us to run expensive statements against production.
Collecting evidence
Check that the evidence can answer your question.
Permissions, security policies, endpoint protection, network restrictions, hosting, software configuration and familiarity with the collector can prevent some or all collection. We can't guarantee the collector will work in every configuration. Its initial checks and summary must show which modules completed, partly completed, failed or were unavailable.
Before analysis, we need the selected statement text or an agreed equivalent, schema and index information, relevant plans or a documented suitable alternative, and meaningful workload measurements. A list of missing-index recommendations alone isn't enough. If the plan cache or Query Store isn't available, a new capture may be needed.
If Query Store or optional history is missing, we record the gap; it doesn't automatically mean collection has failed. We may agree to use a limited sample and available cached plans instead. We won't silently enable Query Store or persistent tracing. Restarts and how long cumulative counters have been running affect how we interpret the results.
Collection is designed to read diagnostic information and take samples within agreed limits. Reads still use resources, and some metadata operations can compete with your workload. We can't guarantee zero impact in every environment. Use an approved window and stop if the impact becomes unacceptable. Changes, restarts, index operations and new persistent capture settings need separate authorisation.
We review predicates, joins, cardinality estimates, parameter sensitivity, scans and seeks, memory grants, relevant waits and index trade-offs. We also consider write overhead, overlapping indexes and vendor support restrictions. One tool suggestion doesn't justify creating or rebuilding indexes across the database.
Each proposed change identifies the statement or object, the reason for the change, the expected direction of benefit, main risks, what must be ready before deployment, rollback and validation. Confidence depends on the evidence. Scripts that need adapting are clearly marked. Implementation needs a separate order and recorded change approval.
If your environment prevents collection
When you can put Self-Service payment toward Managed.
If permissions, security, network, hosting or execution limits stop your team collecting enough evidence, we'll review the collection status and check that Managed collection is feasible. The credit applies to the equivalent Managed assessment of the same instance, databases and question. You don't need to prove who was at fault or blame the person running the collector.
Standard upgrade to the same service
Managed assessment
$1,695
Self-Service credit
−$995
Remaining to pay
$700
USD · taxes recalculated separately
Request the upgrade within 60 calendar days of your Self-Service purchase. We credit the product amount you actually paid, including any paid Priority premium, up to the agreed Managed price. Taxes are recalculated separately. Credits have no cash value and can't be transferred to another customer, product, environment or add-on.
A new Priority slot must be confirmed; a credit does not preserve an expired reservation. Requests to switch modes before analysis begins may use the same credit by agreement even without collection failure.
If Managed collection also can't proceed and analysis hasn't started, we cancel and refund the unperformed Self-Service assessment, including any unused Priority premium. If our collector has a defect, we must fix it, provide an agreed alternative or refund the assessment if it can't proceed. Collector defects must not force a paid upgrade.
An inconclusive report based on enough accepted evidence is different from a failure to collect the evidence. A completed assessment doesn't automatically qualify for a credit or refund. New symptoms, a wider scope or implementation need a new quote.
Cancellations and orders that aren’t ready
You can cancel for a full refund before we begin analysing accepted evidence or start Managed collection. Once work starts, we assess the work performed against the agreed deliverables; your rights under applicable law still apply. An order needs to start within 30 days. If it isn't ready by then, we'll send a reminder and agree cancellation or a written extension. Your agreement, supplied before payment, records the cancellation terms for the service.
Sharing technical evidence
Know what you’re sharing.
What the evidence may contain
Server and database names
Object and index definitions
File paths
Job history and error messages
Execution plans and query text
Query text and execution plans can include literals, parameter values or other sensitive information even when no table rows are exported. Don't assume diagnostic output is anonymous. Review the package before sending it and remove credentials or unrelated information.
Standard collection doesn't intentionally export application table rows, full database backups or passwords. Before sharing data, confirm the compatible collector and the evidence needed. If you can't share that evidence, talk to us before you pay so we can agree a suitable alternative or a different scope.
With Managed, we use named, temporary access through your approved channel, with only the permissions needed. That access is revoked when the work is complete. Permission to read isn't permission to make changes. Don't send credentials, query output, diagnostic packages, backups or sensitive records through general enquiries or ordinary email.
We use your diagnostic evidence to deliver the agreed service. Selling it, using it for public examples or model training, or sending it to external AI services needs separate informed authorisation and suitable agreements.
Use the enquiry form for a brief description of the problem. Keep credentials, query output, diagnostic packages, backups and sensitive records out of the form and ordinary email. Before you share evidence, we'll agree the scope, collection method and privacy, access and retention arrangements.
How long we keep the evidence
Raw packages and temporary working files have a 30-calendar-day retention window after report delivery or cancellation. Final reports are available for 90 days. We keep order and accounting records separately. Before collection, we'll confirm the arrangements and explain any delays deleting backup copies, legal holds or longer retention you've agreed to. This website has no diagnostic upload portal.
When work starts
Turnaround & Priority scheduling
These service targets apply after readiness and availability are confirmed. Enquire for a date; no Priority payment is accepted without a confirmed slot.
Standard
Written delivery target: within five business days after complete intake and sufficient evidence are confirmed.
Managed collection is normally scheduled within five business days after scope, payment and working access are confirmed. The report target starts after agreed collection is complete; collection and reporting are separate stages.
Priority Self-Service
Analysis slot within one business day after sufficient data is admitted. Written report target: two business days after data admission.
Priority Managed
Collection start within two business days after working access and scope are confirmed. Written report target: two business days after collection.
When timing starts and pauses
The clock starts only after scope acceptance, payment and required evidence or working access. BriteThought sends a readiness confirmation and dates. An upload receipt is not evidence acceptance.
The clock pauses for missing data, unavailable access, customer approvals, reproduction windows or an agreed third-party dependency. BriteThought notifies you of the reason, required action and revised dates. Pauses are not hidden and do not cover ordinary BriteThought workload.
Intake acknowledgement target: one business day. This acknowledges receipt; it is not completed technical review or an emergency response guarantee.
Hours and availability
Monday through Friday, 9 am to 5 pm America/New_York, excluding US federal holidays observed by the service.
Requests outside service hours enter the next business-day queue unless a separate appointment is confirmed. A symptoms window outside service hours needs a separately agreed appointment.
Priority requires confirmed availability before payment. Payment alone does not create an immediate response commitment. The booking confirmation states the actual reserved start and target delivery date. Priority changes scheduling, not scope, and does not provide remediation, after-hours coverage, continuous emergency support or a resolution deadline. An active outage, inaccessible server or suspected data corruption needs separate incident scoping.
If a confirmed Priority date is missed
If BriteThought cannot honour a confirmed Priority start or report date for reasons within its control, the Priority premium is refunded and the assessment retains its agreed scope. Customer delays and documented external dependencies do not trigger that remedy. Full cancellation and refund are offered when no agreed work has started and no replacement date is accepted. An unconfirmed slot will not be substituted after payment.
This is a one-time service with a fixed scope. You receive the agreed investigation or test and its written result. Production changes and remediation need a separate agreement, except for the operations expressly included in the isolated Recovery Proof test. We don't guarantee improved performance, uptime, identification of a root cause or recoverability. The service does not certify security or compliance.
Deploying or rebuilding indexes, updating statistics, forcing plans, changing compatibility, releasing code, extensive procedure rewrites and whole-database index cleanup aren't included.
This service does not include ongoing monitoring, unlimited revisions, an enterprise audit, compliance certification, licensing advice, an application rewrite, migration, hardware procurement, a production restore, corruption repair, representation in a vendor dispute or routine DBA work. It does not certify security or recoverability.
Why BriteThought
Careful analysis. A report you can use.
Evidence before assumptions
Reports distinguish what was observed, what may be contributing and what remains unknown. Each recommendation explains its evidence, its limits and the next useful test.
A specialist reviews the evidence
A qualified specialist analyses the evidence and is responsible for your report, using documented collection methods and review checklists. Consistent report formats and a process for further technical review support the assessment. Participation by a particular founder or engineer isn't guaranteed.
A report your team can use
Use the report with your team, vendor or chosen implementer. It sets out priorities, responsibilities and validation steps so you can decide what happens next. Production changes remain a separate decision.
Common questions
Before we start.
Do you rewrite all our stored procedures?
No. We review the selected statement units and may suggest changes within the agreed limits. A larger rewrite is a separate project.
Can you guarantee faster queries?
Performance depends on data, parameters, workload and environment. We provide recommendations supported by the evidence and a way to test them.
Why do we need vendor approval?
Changing a vendor database or application can affect support. Your team needs to confirm which changes are allowed.
Does Managed include making the recommended changes?
With Managed, we collect the evidence and carry out the investigation or validation listed for the service. You receive recommendations; making those changes or providing ongoing support needs a separate scope.
Does Priority include support around the clock?
Priority requires confirmed availability before payment. Payment alone does not create an immediate response commitment. The booking confirmation states the actual reserved start and target delivery date. Priority changes scheduling, not scope, and does not provide remediation, after-hours coverage, continuous emergency support or a resolution deadline. An active outage, inaccessible server or suspected data corruption needs separate incident scoping.
Can we share the report and ask follow-up questions?
You can share the report with your organisation, advisers, vendor and chosen implementer. Send up to five questions about it together within ten business days of delivery; one round of clarification is included. This does not include new evidence, symptoms or implementation. Managed includes one 30-minute findings call. Self-Service includes written delivery; a call can be purchased separately.
Is Self-Service just an automated report?
No. Your team collects the evidence, and a qualified specialist analyses it and writes the report. Self-Service costs less because your team handles collection.
What if our team can’t collect enough evidence?
We include one written review of your collection status and, when useful, one request for a targeted rerun or limited substitute evidence. Troubleshooting permissions, security, network or execution problems specific to your environment, and remote sessions, are excluded. If environmental limitations prevent collection, your payment may qualify for credit toward the equivalent Managed assessment, provided we can perform it. The upgrade rules on this page include a 60-calendar-day request limit. A standard upgrade leaves $700 to pay, excluding tax adjustments.
What if we can’t provide access for Managed either?
We confirm whether Managed collection can proceed before you upgrade. If we still can’t collect enough evidence and analysis has not begun, we refund the unperformed Self-Service purchase, including any unused Priority premium, under the refund rules on this page. If a defect in the BriteThought collector prevents the assessment, we must correct it, provide an agreed alternative or refund you if the assessment cannot proceed. Collector defects must not force a paid upgrade.
Can diagnostic collection affect production?
Collection is designed to read diagnostic information and take samples within agreed limits. Reads still use resources, and some metadata operations can compete with your workload. We can't guarantee zero impact in every environment. Use an approved window and stop if the impact becomes unacceptable. Changes, restarts, index operations and new persistent capture settings need separate authorisation.
After the report
Decide what happens next.
A supported recommendation can be quoted through a Change Implementation Block. If the slow operation is still unclear, consider Slow Application Triage. Wider configuration concerns may suit a Health Check.
You can take the report to your team, vendor or another provider. Any further work needs its own clear question and scope. You don’t need to buy another service to have the agreed report explained.
We agree a separate quote for optional follow-up work before it starts. It stays within that scope and does not add ongoing support or automatic charges. Your own team or vendor can act on the report without buying more services.
Tell us what you need to find out.
Tell us what’s happening, how it affects your business and what you want to find out. We’ll check whether the service fits, explain what collection or testing involves and agree when work can start. Please leave credentials and diagnostic evidence out of your enquiry.
This is a one-time assessment with a fixed scope and the listed deliverables. Production changes and remediation are separate. With Self-Service, your team collects the evidence; customer-specific collection troubleshooting is excluded. Priority requires confirmed availability. See scope, data handling and upgrade terms.
