Introduction
To migrate from QuestionPro to Sprig, export every response to CSV and SPSS before you cancel, because QuestionPro’s published policy deletes a cancelled account’s data within seven days. Then rebuild each live survey from a review-mode PDF with Sprig’s AI File Upload Builder, redesign anything built on custom scripting or respondent authentication, and run tracked metrics on both platforms until the step change is measured.
Last updated November 2026.
What a Migration Actually Involves
A survey platform migration is four projects that share a deadline: an archive, a rebuild, a measurement bridge and an administrative cutover. QuestionPro removes a cancelled account’s data within seven days, so the archive can’t share a deadline with the rebuild.
The Four Projects
The archive preserves what you’ve collected. The rebuild recreates the instruments still in use, which is rarely all of them. The bridge protects any metric someone reports as a trend.
The cutover moves people, permissions, integrations and distribution, including in-product surveys on a website or web app and native mobile apps where they apply.
Why One Plan Fails
The four projects run on different clocks. The archive is bound by QuestionPro’s deletion window, the bridge by your reporting cycle, and the cutover by your identity and DNS teams.
A single plan with one date commonly hides the project that runs late until it’s the only thing left.
The decision rule for this section: if a survey has no owner, no consumer and no fielding date in the last year, it’s an archive item, not a rebuild item.
What Does Not Port
Several QuestionPro capabilities have no documented Sprig equivalent, and a few have a partial one. The table lists each with what you do instead.
"Not documented" means Sprig’s published documentation doesn’t describe the capability as of November 11, 2026, so confirm with Sprig before treating it as absent.
| What you have in QuestionPro | Does it port | What you do instead | Source |
|:---:|:---:|:---:|:---:|
| Constant sum | Not documented | Rebuild as rank order plus a weighting question, or move the study to maximum difference scaling (MaxDiff) | QuestionPro standard question types, Sprig question types |
| Upload files and signature | Not documented | Collect files or signatures in a separate form and join by respondent ID | QuestionPro advanced question types, Sprig question types |
| Maps, date and time, calendar | Not documented | Use multiple choice with defined ranges, or validated text with a regular expression | QuestionPro standard question types, Sprig validated text |
| Total unduplicated reach and frequency (TURF) analysis and simulator | Not documented | Export responses and run TURF in a statistics package, checking any AI-drafted code against a known result | QuestionPro TURF analysis help, Sprig docs index |
| Gabor-Granger question type | Not documented | Rebuild as a sequence of purchase-intent questions at fixed price points | QuestionPro help navigation, Sprig question types |
| Van Westendorp question type | Partial | Rebuild the four price questions as separate numeric validated-text questions and analyze outside Sprig | QuestionPro Van Westendorp help |
| Apache Velocity custom scripting | Not documented | Redesign the instrument around skip logic, display logic and piping | QuestionPro custom scripting, Sprig study logic |
| Custom variables set mid-survey | Partial | Pass values in as user attributes before the survey loads | QuestionPro custom variables, Sprig attributes |
| Looping on a checkbox question | Not documented | Repeat the follow-up per option with display logic, capped at the options you actually need | QuestionPro looping, Sprig study logic |
| Respondent authentication | Partial | Use one-response invitation links or a user ID parameter on link surveys | QuestionPro authentication types, Sprig duplicate prevention |
| Offline collection in the mobile app | Not documented | Keep offline fieldwork on another tool or stop collecting it | QuestionPro mobile app help, Sprig docs index |
| Historical responses | Not documented | Archive outside both platforms, covered in the historical data section | Sprig export documentation, Sprig changelog |
Several QuestionPro items above sit on its Research Edition license, and several Sprig remedies, including display logic, response piping, quotas and MaxDiff, are documented as Enterprise-only.
Question Types With No Documented Equivalent
Constant sum, file upload, signature, maps, date and time, and calendar questions appear in QuestionPro’s documented question types and not in Sprig’s. Constant sum is the one that commonly breaks real programs, because forced-allocation questions often feed a budget or priority decision.
Rather than approximating it with a slider, rebuild it as a rank order plus a separate weighting question, or move the study to MaxDiff.
File upload and signature questions are typically compliance artifacts rather than research ones. The clean remedy is usually a separate form tool that captures the file or signature, joined back to the survey by a respondent ID.
Date and map questions generally rebuild as multiple choice with defined ranges, or as validated text using Sprig’s custom regular-expression format.
Survey Flow Constructs
QuestionPro’s custom scripting runs Apache Velocity code on branching and looping, and Sprig documents no scripting surface. An instrument that computes a value mid-survey and routes on it generally needs a redesign, not a rebuild.
Looping, which QuestionPro limits to multiple-select questions, has no documented Sprig equivalent, so repeat the follow-ups with display logic instead.
Custom variables port only in part. QuestionPro’s survey help documents up to 255 of them, passed through the survey URL or an email list upload.
Sprig’s nearest construct is user attributes, set before the survey through the SDK, a CSV upload, Segment, Google Tag Manager or the API. Values known before the survey move cleanly, and values assigned during it don’t.
Authentication, Analysis and Operations
QuestionPro documents several respondent authentication types, from a global password to unique usernames, email passwords and participant IDs. Sprig’s closest controls are one-response invitation links, your own invitation IDs, or a user ID parameter, all on link surveys.
Sprig’s approach covers duplicate prevention, but it isn’t a password gate, so treat it as a partial port.
TURF and Gabor-Granger are documented QuestionPro analyses, often used in pricing work, with no Sprig counterpart, and Van Westendorp ports only as four separate questions analyzed elsewhere.
Offline collection in QuestionPro’s iOS and Android app has no documented Sprig equivalent at all. Historical responses don’t move either, which is why they get their own section below.
The decision rule: any row marked "Not documented" goes to the instrument owner as a design decision before rebuild work is scheduled.
Who Should Not Migrate
Some teams shouldn’t migrate, and they’re better served by hearing it before the audit starts. Five readers fall into that group.
Teams That Collect Offline
If your program depends on fieldworkers collecting responses on tablets without connectivity, stop here. QuestionPro documents offline collection in its iOS and Android app, syncing to the account when a connection returns.
Sprig documents no offline collection. Rather than splitting one study across two tools, keep the offline program where it works and migrate only the online instruments.
Teams With a Regional Hosting Requirement
Sprig hosts its software in Amazon Web Services facilities in the USA and documents no regional hosting option. QuestionPro publicly describes a European data center and a Dubai data center, and says it can move existing customers to a preferred one.
If a contract, regulator or works council requires data to stay in a specific region, that requirement is disqualifying rather than inconvenient.
Teams Whose Output Is Specialist Analysis
A pricing or portfolio team whose deliverable is a TURF reach curve, a conjoint market simulation or a Van Westendorp chart produced inside the platform is giving up real capability.
QuestionPro documents all three on its Research Edition license. Sprig documents conjoint and MaxDiff as Enterprise question types, with no TURF, no simulator and no Gabor-Granger.
Teams Without Enterprise Access to the Remedies
Many of the remedies in this guide are documented as Enterprise-only in Sprig. The survey-building ones are display logic, response and attribute piping, quotas, rank order, MaxDiff and conjoint. The operational ones are the AI File Upload Builder, single sign-on (SSO), System for Cross-domain Identity Management (SCIM) and email distribution.
A team on another plan that relies on QuestionPro’s branching, quotas and piping will find that most of its instruments don’t rebuild at all. Confirm the plan before the audit, not after it.
A team that can’t move to an Enterprise plan should generally keep any survey that depends on those features where it is, and migrate only the instruments that use skip logic, randomization and the question types available on every plan.
Teams Mid-Tracker With No Room for an Overlap
A team whose next benchmark wave falls before a bridge study could finish, with no budget to field both platforms, should usually wait. Switching platforms changes the measured number for reasons unrelated to customers, and without an overlap you can’t separate the two.
The usual better plan is to finish the benchmark on QuestionPro, migrate everything else first, and move the tracker at the start of the next cycle with a bridge study attached.
The decision rule: if any of these five describes your core program rather than an edge case, don’t migrate that program, or don’t migrate yet.
The Pre-Migration Audit
The audit decides what moves, what’s archived and what’s retired, and it should finish before anyone rebuilds a question. Start by sorting QuestionPro’s survey list by last response date, and send every survey with no responses in 12 months to archive by default.
The Audit Checklist
Work through these in order and record each answer in one shared sheet:
- List every survey with responses in the last 12 months, and its owner
- Mark which surveys are still fielding today and through which channel
- Identify every downstream consumer of each survey’s data, including scheduled exports that land in an inbox
- Flag every metric reported as a trend, with the person who set its target
- Record each survey’s license-gated features: custom scripting, looping, quotas, conjoint, MaxDiff, TURF, authentication
- Find the contract clause that sets your end date, and name who owns it
- Confirm which Sprig plan you’re on, since several remedies are Enterprise-only
- Assign every survey one disposition: rebuild, redesign, archive or retire
Two items commonly set the sequence rather than the scope: the downstream consumers and the contract owner.
The Two Items Nobody Checks
Consumers matter because a survey can look dead in the platform and still feed a dashboard every Monday. The contract owner matters because QuestionPro’s deletion clock starts at cancellation, and the person who sends the cancellation is frequently in procurement rather than research.
Sizing the Rebuild
Count the dispositions before you estimate the work. Redesign items, the surveys that touch the does-not-port table, need an owner, an analyst and a test cycle, so estimate them separately from straight rebuilds.
Rather than estimating by survey count, estimate by disposition, since a straight rebuild is mainly a transcription and review task.
The decision rule: no survey is rebuilt until it has an owner, a consumer list and a disposition in the audit sheet.
Exporting from QuestionPro
QuestionPro documents three export routes that matter for a migration: raw data to Excel or CSV, raw data to SPSS, and the survey itself to PDF or Word. Each one captures something the others miss, so take all three for every survey you keep, with question codes and answer codes switched on.
Raw Data to Excel and CSV
The Excel and CSV export carries options that decide whether the archive is readable later. Turn on Display Codes, Question Codes instead of Text and Unselected Checkboxes with 0, and turn on Show Questions Not Displayed so skipped questions are distinguishable from unanswered ones.
QuestionPro documents that above 5,000 responses it splits the Excel output into multiple sheets delivered as a zip file.
Raw Data to SPSS
The SPSS export keeps value labels, which a CSV commonly loses. QuestionPro documents it on its Advanced, Team Edition and Research Edition licenses, with an Include Randomization Data option that records block order.
Turn that on for any randomized survey, because the order a respondent saw is otherwise unrecoverable once the account closes.
Two documented limits are worth knowing. Comment box text is capped at 10,000 characters in the SPSS file and other open-ended types at 500, so keep the CSV as the copy of record for long text.
Looped surveys produce multiple rows per response ID, which QuestionPro notes SPSS doesn’t handle well.
The Survey Itself
QuestionPro’s print tool downloads a survey as PDF or Word, "generated as respondent sees it." That rendered copy shows wording and options but not logic.
QuestionPro’s review mode produces a copy that QuestionPro says "will contain all survey flow settings and information like validation, blocks, logic, etc." Take both, because the review-mode PDF is typically the most useful input to the rebuild.
The API Route
QuestionPro’s API reference documents endpoints that retrieve a survey’s questions and its responses. For a large account, scripting the export through the API is generally faster than clicking through many surveys, and it gives you a repeatable job you can rerun on the last day.
Name and Store the Files Consistently
Use one naming pattern for every file, such as the QuestionPro survey ID, the survey title and the export date, and keep the three exports for a survey in one folder.
Record the export settings you used in the codebook, because a file exported without question codes can’t be matched to a rebuilt survey later.
The decision rule: a survey isn’t archived until its CSV, SPSS file and review-mode PDF all sit in the archive location and someone has opened each one.
The Historical Data Decision
Sprig documents no way to import historical survey responses, so your QuestionPro history stays outside Sprig. That single fact shapes the archive, the reporting plan and the deadline.
What the Documentation Shows
Sprig’s export documentation describes CSV export only. Its CSV upload imports user attributes and users for targeting, requires a User ID column, and overwrites existing attribute values on a match.
Sprig’s changelog through October 27, 2026 lists no response import. Verify this yourself before planning around it, because a negative finding about documentation can go stale without notice.
The Clock on QuestionPro’s Side
QuestionPro’s data retention policy says data in a cancelled account is removed from its primary database systems within 24 hours, and then from all systems including backups within seven days. QuestionPro’s separate GDPR page describes a 30-day grace period instead, so plan to the stricter seven-day figure.
The same policy says data from a cancelled account can’t be recovered. Treat the cancellation date as the export deadline, and confirm the terms in writing with your account contact.
Where the History Should Live
Rather than forcing old responses into the new platform, keep them in a governed store your analysts already use: a data warehouse, a statistics server or a controlled shared drive.
Store the raw files, a codebook mapping question codes to wording, and the review-mode PDF for every survey.
What the Archive Has to Contain
An archive is only useful if someone who never used QuestionPro can read it. For each survey kept, store these five things:
- The CSV export with question codes and answer codes
- The SPSS file with value labels and randomization data
- The review-mode PDF showing logic and validation
- A codebook mapping each code to its wording and scale
- A note on fielding dates, channels and any mid-study changes
The fifth item is generally the one an analyst needs first when a number looks strange.
The Archive Is a Privacy Object
If the old responses contain personal data, the archive inherits your retention schedule and deletion obligations, so give it an owner and an expiry date.
The decision rule: if a metric’s history matters to anyone, its archive gets an owner, a codebook and a retention date before cancellation is sent.
Mapping Question Types
Most QuestionPro question types map to a documented Sprig type, and the restrictions matter more than the names. A type that exists only on Enterprise plans or only in one survey format hasn’t ported cleanly for a team outside that plan or format.
| QuestionPro type | Sprig type | Port | Restriction |
|:---:|:---:|:---:|:---:|
| Multiple choice, select one or many | Multiple Choice Single-Select and Multi-Select | Yes | Images in options on Surveys only |
| Star, smiley and slider ratings | Rating Scale | Partial | Display styles not named in the docs |
| Net Promoter Score | NPS | Yes | None listed |
| Comment box | Open Text | Yes | None listed |
| Email address, contact information | Validated Text | Partial | Email, numeric, URL, phone and custom formats |
| Multi-point and checkbox matrix | Matrix and Accordion | Partial | Rows and columns can’t be deleted after launch |
| Rank order, drag and drop | Rank Order | Partial | Enterprise, In-Product and Surveys, 2 to 10 options |
| MaxDiff | MaxDiff | Partial | Enterprise, Surveys in Standard Format |
| Discrete choice conjoint | Conjoint | Partial | Enterprise, link surveys |
| Presentation text, section heading | Text and URL Prompt | Yes | None listed |
How to Read the Mapping
Multiple choice, NPS, open text and presentation text move cleanly on every plan. Rating questions map to Sprig’s Rating Scale, but Sprig documents changes to range, labels and icons rather than named star, smiley or slider styles, so check the rendered result against the original.
Email and contact questions map to Validated Text, which documents email, numeric, URL, phone and custom regular-expression formats.
Matrix questions map to Matrix and Accordion, with one documented catch: after launch, rows and columns can’t be deleted or reordered, only added. Rank order allows two to 10 options and is Enterprise-only, on In-Product Surveys and Surveys.
MaxDiff and conjoint are the two specialist types that do map. Both are Enterprise-only in Sprig’s documentation, MaxDiff in Surveys using Standard Format and conjoint on link surveys. Teams running concept testing on conjoint should confirm their plan before moving those studies.
The decision rule: every "Partial" row gets a test response on the rebuilt survey before launch, checked against the QuestionPro original.
Rebuilding Logic and Survey Flow
Sprig documents logic at the question level through skip logic, display logic and piping, and QuestionPro programs that used logic for routing generally rebuild without trouble. Programs that used logic as computation commonly don’t.
Translating Branching
QuestionPro branching becomes skip logic on the question before the split, or display logic on the questions inside each arm.
Sprig documents that skip logic can’t route between two questions on the same page, so a branch inside a multi-question page has to be split across pages. Display logic is Enterprise-only and can use prior answers or user attributes.
Piping and Carry Forward
Response piping and attribute piping are both documented as Enterprise-only. Response piping needs at least one prior question and supports open text, rating, multiple choice, NPS, matrix and rank order answers, with fallback text required.
Attribute piping can fill the first question, but only when the respondent arrives with an identifier.
Custom Variables Become Attributes
QuestionPro custom variables passed in a survey URL or an email list upload map to Sprig user attributes.
Sprig documents that attributes can’t trigger a study and can only filter one. A response carries an attribute value only if it was set when the study was delivered.
Load attributes before launch, typically through a CSV upload keyed on User ID, rather than expecting them to arrive with the respondent.
Quotas
QuestionPro documents question-based, criteria-based and custom-variable quotas. Sprig documents response-based quotas on Enterprise plans, matched on screener questions using single-select multiple choice, rating scale or NPS.
A quota can’t be deleted from an active survey, so finalize quota design before launch rather than adjusting it in flight.
Randomization
Sprig documents randomization of response options, questions, blocks and random subsets of a block on all plans, and skip and display logic still apply after shuffling.
QuestionPro’s Include Randomization Data export is worth matching here: record the randomization design in your codebook so the old and new orders can be compared.
The decision rule: if an instrument’s logic computes a value, it’s a redesign with its own owner and review, not a line item in the rebuild.
Rebuilding the Instrument in Sprig
The fastest documented rebuild path is Sprig’s AI File Upload Builder, part of the Design Agent, which reads a PDF or Word document and drafts a survey from it. It drafts rather than converts, so every output is a draft until a person has checked it against the original.
Feeding It the Review-Mode PDF
Sprig documents that the builder accepts PDF and DOCX files up to 25 MB on all Enterprise plans and describes "a complete survey in under 2 minutes." Budget review time per survey, not generation time.
The review-mode PDF from QuestionPro is close to the format it reads. Mark the logic with the documented prefixes REQUIRED:, SKIP LOGIC:, DISPLAY LOGIC: and RANDOMIZE:, and wrap multi-question pages in page start and end markers.
What It Won’t Do
MaxDiff questions and prototype links aren’t supported through file upload and have to be added by hand afterwards. The builder also can’t recreate anything from the does-not-port table, because those constructs have no target to draft into.
Review every draft against the original wording and options, since a mistranscribed answer option breaks a trend line without any visible error.
The Scripted Route
Sprig documents a Model Context Protocol (MCP) server that lets AI clients such as Claude, ChatGPT and Cursor draft studies, edit them, configure logic, translate and set targeting.
Drafts stay in draft until someone launches them, and only administrators can turn the connection on. Sprig’s changelog also lists 37 API operations covering the study workflow, added October 27, 2026.
Translations
QuestionPro multilingual surveys need their translations rebuilt too. Sprig documents a CSV workflow, where you download a file with translatable fields in rows and languages in columns, fill it and upload it, and warns not to add or delete rows or columns.
The Design Agent’s AI Translations feature drafts translations of a finalized survey, and Sprig notes that regenerating them overwrites any edits.
Where Sprig fits. For a team rebuilding many surveys, the Design Agent and the MCP route generally turn transcription into review, though their drafts can mis-transcribe options or logic, so the review has to be line by line. The limitation is that both draft from what you give them, so a survey whose QuestionPro logic was never written down still needs a person to reconstruct it first.
The decision rule: rebuild tracked instruments first, high-volume operational surveys second, and the long tail last, retiring anything the audit left without an owner.
The Measurement Discontinuity
Changing survey platforms changes the number you track, and the change isn’t a change in your customers. A team that migrates a tracker and reports the step to executives has reported an artifact as a finding.
Two Effects, Not One
A migration changes mode effects and selection effects at the same time. Vannieuwenhuyze, Loosveldt and Molenberghs, in "A Method for Evaluating Mode Effects in Mixed-Mode Surveys" (Public Opinion Quarterly, 2010), separate the two: selection effects come from different people answering, and measurement effects come from the same people answering differently. Separating them needs a comparable dataset collected alongside, not a judgment call afterwards.
How Large the Mode Effect Can Be
Pew Research Center’s 2015 report "From Telephone to the Web" fielded 60 questions to 3,003 respondents randomly assigned to phone or web.
Pew found "a mean difference of 5.5 percentage points and a median difference of five points across the 60 questions." On satisfaction with family life, 62% of phone respondents said "very satisfied" against 44% on the web.
Satisfaction items moved most, and satisfaction items are what Net Promoter Score (NPS) and customer satisfaction programs most often track.
Pew also states that "there is no way to determine whether the telephone or the Web responses are more accurate." Neither number is the true one, so report the bridge estimate as the size of the break rather than as a correction.
House Effects Without a Mode Change
Diana Schumann, Hawal Shamon and Jürgen-Friedrich Hake, in "The Importance of House Effects for Repeated Public Opinion Surveys" (International Journal of Public Opinion Research, DOI 10.1093/ijpor/edz039), found that "house effects can occur even if the survey agencies use identical question wordings or target populations." A platform switch is the same situation: same questions, same customers, different infrastructure.
Two Precedents
The US Bureau of Labor Statistics ran a Parallel Survey from July 1992 to December 1993 on a sample about one-fifth the size of the Current Population Survey before its 1994 redesign.
The Bureau’s best estimate of the redesign’s effect on the annual average unemployment rate was "an increase of .56 percentage point."
The University of Michigan Surveys of Consumers ran web collection in parallel with phone for over seven years, then phased the switch from April to July 2024, reporting a 0.97 time series correlation between the two methods.
A June 2026 Chicago Fed Letter by Brave, Henken and Jolley still estimated an initial measurement effect of about 10 index points on Michigan’s index. The same letter said a similar move to online collection by the Conference Board had a "fairly small" effect.
What the Precedents Teach
A high correlation and a level shift aren’t in conflict. Correlation preserves the shape of a series and says nothing about its level, so an overlap lets you detect and size a break rather than prevent one.
A team that reads only direction month to month generally comes through a switch fine. A team with a numeric target or a contractual threshold doesn’t.
The Commercial Tension, Stated Plainly
The methodologically correct answer is to run both platforms at once, and you’re migrating partly to stop running both.
The costs are covered in the parallel run section below. Pretending the overlap is free gets the recommendation ignored, so plan it as a cost with a defined end.
No Published Rule Sets the Overlap Duration
No published rule sets an overlap duration. The US Office of Management and Budget’s Standards and Guidelines for Statistical Surveys (2006) advises agencies to "Develop adjustment methods, such as crosswalks and bridge studies," and defines a bridge study as one that "continues an existing methodology concurrent with a new methodology."
The OMB guidance gives no duration, and the precedents above ranged from 18 months to seven years.
The decision rule: for every tracked metric, decide before migrating whether anyone depends on its level or only on its direction, because a level dependency needs a bridge study and a direction dependency usually needs only a note.
Designing the Bridge Study
A bridge study runs the old and new instruments at the same time on comparable samples, so the step change can be measured rather than guessed. It’s the procedure the OMB guidance names, scaled to a survey program.
What to Hold Constant and What to Compare
Hold the wording, answer options, order and scale direction identical, since small presentation changes frequently move rating items. Keep the sample frame as close as you can, which is often the hard part, since an emailed QuestionPro list and an in-product Sprig intercept reach different people.
Compare the tracked metric’s mean and distribution, the response rate and the completion rate.
The Bridge Study Specification
Use this as the plan for each tracked metric:
- Name the metric, its owner and every report that uses its level
- Field the QuestionPro and Sprig versions in the same weeks to randomly split halves of one frame where possible
- Size each arm with a power calculation for the smallest change the owner would act on
- Run through at least one full reporting cycle, and through a seasonal peak if the metric has one
- Estimate the difference with a confidence interval, not a single point
- Agree in advance what result counts as continuous and what counts as a break
Where Sprig fits. Sprig’s panel and in-product channels reach a different frame from an emailed QuestionPro list, so match the frame only where both arms can draw from the same list. The limitation is that Sprig documents no weighting or significance testing, so the comparison in steps 3 and 5 runs in your own statistics tool.
When the Frame Can’t Be Matched
If the old program emailed a list and the new one intercepts users in the product, no bridge design fully separates mode from frame.
Say so in the read-out, and report the bridge estimate as the combined effect of both rather than as a pure mode effect.
The decision rule: if the confidence interval for the difference excludes zero, the series broke, so report a new baseline with the bridge estimate attached rather than splicing the two.
Users, Roles, SSO and Provisioning
Sprig documents Security Assertion Markup Language (SAML) 2.0 single sign-on on Enterprise plans, with setup guides for Auth0, Google Workspace, KeyCloak, Microsoft Entra ID, Okta and OneLogin. QuestionPro documents SAML sign-on on its Research Edition, including a "Restrict to SSO only" setting.
Enforcement and Roles
Sprig documents two enforcement settings, one for all non-admin users and one for everyone, and warns to confirm an Organization Admin can sign in through the identity provider before enforcing it for all.
Access is organized by Workspace, with roles such as Viewer, Editor and Workspace Admin.
The Provisioning Gap
The provisioning gap is specific. Sprig documents SCIM on Enterprise plans for deprovisioning only: it soft-deletes inactive users but "does not create users through SCIM," and only Okta and Microsoft Entra ID are verified. Plan user creation as a manual or domain-based step with a named owner, since a migration often adds many users at once.
Compliance, as Documented
Sprig’s documentation lists SOC 2 Type 2, HIPAA and the Data Privacy Framework, with encryption in transit and at rest.
Sprig’s question-type documentation notes that Video and Voice questions aren’t supported in HIPAA-compliant environments, which matters for a health program moving open-ended video prompts. Check each of your QuestionPro compliance requirements against Sprig’s current documentation rather than against a badge on a marketing page.
The decision rule: enforce SSO only after an admin has signed in through it, and keep a dated user list until SCIM deprovisioning has been tested.
Integrations and Downstream Data
Inventory the Consumers
Inventory the people and systems that receive survey data rather than the integrations that are switched on. A scheduled CSV that lands in someone’s inbox is a consumer, and it won’t appear on any integration list.
For each consumer, record what it receives, how often, in what format, and what happens if it stops. That last answer sets the order of the cutover.
Getting Data Out of Sprig
Sprig exports CSV through a link emailed to the requester, capped at 950,000 rows, and its MCP connection reads up to 1,000 records at a time.
Sprig’s email distribution is documented as Enterprise-only and on link surveys only, with a verified custom sending domain, CSV contact lists that need Email, First Name, Last Name and User ID columns, and a 24-hour cooldown per recipient.
Getting Data Into Sprig
Targeting data flows the other way. Sprig documents attributes arriving through its SDK, a CSV upload, Segment, Google Tag Manager or its public API, so whichever system held your QuestionPro email lists now feeds Sprig attributes instead.
The domain verification step involves whoever manages DNS, who is typically not in the migration meeting and often works on a separate queue. Start custom sending-domain verification in the first week, because Sprig can’t create an email run until at least one domain is verified.
The decision rule: every consumer on the audit sheet gets a named replacement feed and a test date before the QuestionPro feed is switched off.
The Parallel Run
The parallel run is the period when both platforms field live surveys, and it exists for the bridge studies and for catching rebuild errors. It costs the most when it has no end condition.
Making It Affordable
Run in parallel only the instruments that feed a tracked metric, which is typically a small share of an account. Operational surveys with no trend reporting can generally switch on a date.
Rather than parallel-fielding everything, keep the overlap to the handful of trackers in your experience measurement program, and set an end condition for each based on its bridge result rather than on the calendar.
What the Overlap Costs
An overlap costs license time on QuestionPro, analyst hours to read two datasets, and respondent attention, since the same customers may be asked twice.
Splitting one frame into two random halves, one per platform, avoids double-surveying anyone while still producing a comparison.
Sequencing Against the Contract
Work backwards from the QuestionPro end date. The final full export, the bridge read-out and the consumer cutover all have to land before cancellation, with the deletion clock in mind.
If the bridge can’t finish in time, negotiate a short extension rather than ending the series blind.
The decision rule: a tracker leaves parallel only when its bridge study has reported and its owner has accepted the result.
Cutover and Decommissioning
The cutover is a sequence with dependencies, and each phase has an owner rather than a date someone guessed. Run it in this order:
- Research operations completes the audit and assigns dispositions
- Each survey owner exports the CSV, SPSS file and review-mode PDF, verified by opening them
- Survey owners rebuild and test instruments, trackers first
- Survey owners and analysts launch bridge studies for every tracked metric
- Data owners repoint each downstream consumer and confirm one delivery
- The platform admin moves users, configures SSO and tests SCIM deprovisioning
- The contract owner runs a final export, then sends cancellation
Each phase depends on the one before it, and phase seven depends on all of them. The final export matters because responses keep arriving during the parallel run, so the archive from phase two is already out of date by phase seven.
What to Keep After Cancellation
Keep the archive, the codebook, the audit sheet and every bridge study read-out. The audit sheet records why each survey was retired, which answers the question someone asks a year later about a missing report.
Keep the bridge read-outs next to the metrics they adjust, which is generally where a confused reader starts looking.
The decision rule: cancellation is the last action in the migration, never the first.
Validating the Migration
Validation confirms that the rebuild matches the original and that nothing downstream broke. Check each item and record the result:
- Compare every rebuilt survey’s wording and options to the review-mode PDF, line by line
- Submit test responses through every logic path and confirm the routing
- Confirm each "Partial" mapping renders as intended on the live channel
- Reconcile the archive’s response counts against QuestionPro’s reported totals
- Open each SPSS file and confirm value labels survived
- Read each bridge study result and record the decision
- Confirm every downstream consumer received one delivery from the new feed
- Confirm departed users can’t sign in
Two items deserve the most care: the archive reconciliation, which catches a truncated export, and the departed-user check, which catches a provisioning gap that otherwise surfaces in an audit months later.
Approving the Migration
One person should approve the migration as complete, typically the research operations lead, with each checklist line initialed by the person who checked it.
Rather than declaring completion when the last survey launches, declare it when the archive is reconciled and every bridge study has reported. Those two are the items that can’t be fixed once QuestionPro is gone.
The decision rule: the migration is done when every item above has a recorded pass and a named checker.
Where QuestionPro Wins
QuestionPro leads Sprig on several checkable capabilities, and a team that used it for years will know them. These are the ones that matter for a migration.
Specialist Pricing and Portfolio Methods
QuestionPro documents a conjoint market simulator, MaxDiff preference shares, TURF with a reach simulator on checkbox and matrix questions, and a Van Westendorp question type, all on its Research Edition license, and lists a Gabor-Granger question type in its help center. Sprig documents conjoint and MaxDiff question types but neither simulator, no TURF and no Gabor-Granger.
Statistical Methodology You Can Name
QuestionPro’s MaxDiff documentation names its estimation model, a multinomial logit, and lays out the steps from raw utilities to preference shares. A team that has to defend a number to a methodologist can point to that page.
Sprig documents no significance testing or weighting, and its cross-tab recipes run in a connected AI client such as Claude rather than in Sprig, with a minimum cell size instead of a statistical test.
Programmable Survey Flow
QuestionPro’s custom scripting runs Apache Velocity code on branching and looping, alongside up to 255 custom variables and reference-data lookup tables. That makes it a programming surface for complex instruments. Sprig’s logic is question-level and has no documented scripting layer.
Respondent Authentication and Offline Fieldwork
QuestionPro documents several respondent authentication types and offline collection in its iOS and Android app. Sprig documents duplicate prevention on link surveys and no offline collection, which concedes both for any program built around controlled access or fieldwork without connectivity.
Question Breadth and Regional Hosting
QuestionPro’s documented question types include constant sum, file upload, signature, maps and date fields that Sprig doesn’t document. QuestionPro also publicly describes European and Dubai data centers, against Sprig’s documented US-only hosting. Those are marketing pages rather than contract terms, so a team with a residency requirement should get QuestionPro’s hosting location in writing too.
A Larger Review Base
On G2, QuestionPro holds 4.5 out of 5 from 1,163 reviews and Sprig 4.3 out of 5 from 199, both retrieved November 11, 2026.
QuestionPro’s Capterra rating is 4.8 from 544 reviews on the same date. Sprig’s G2 review base is about one-sixth the size of QuestionPro’s, so read the gap between the two scores with that difference in mind.
How to Use This List
Treat each item as a test rather than a footnote. If a program’s core output depends on one, keep that program on QuestionPro. If it’s an edge case, plan its workaround from the does-not-port table. Several items appear twice in this guide, once as a migration problem and once here, because each is a capability you give up and not only a task you inherit.
The decision rule: if your program’s core output depends on any capability in this section, keep that program where it is.
Common Mistakes
These six mistakes are sequencing mistakes, not rebuild mistakes, and each one appears earlier in this guide as a step:
- Cancelling before the final export, which starts QuestionPro’s deletion clock
- Exporting CSV only, which loses SPSS value labels and randomization order
- Rebuilding every survey, including ones nobody has opened in a year
- Reporting a tracker’s first Sprig reading as a trend, with no bridge study
- Assuming custom variables port, when only values known before the survey do
- Skipping SCIM testing, which leaves departed users with access
Two Mistakes Specific to QuestionPro Exports
The first is trusting the SPSS file for open text. QuestionPro documents a 500-character limit on most open-ended types in the SPSS export, so long answers survive only in the CSV.
The second is missing the zip split: above 5,000 responses, the Excel export arrives as several sheets in one zip file, and an archive built from the first sheet is incomplete.
Put Each Mistake on the Audit Sheet
Rather than relying on memory, put each of these six on the audit sheet as a line with an owner and a check date. Cancellation is commonly owned by procurement, so the export line needs a named person in procurement as well as in research.
The decision rule: any task that can’t be undone, such as cancellation, deletion or a first trend report, waits until the tasks that protect it have passed.
How Sprig Supports the Migration
Sprig’s support for a migration runs through three documented tools: the Design Agent for the rebuild, the Field Agent for distribution, and the Synthesize Agent for reporting on the new surveys once they’re live.
The Design Agent’s AI File Upload Builder drafts surveys from the review-mode PDFs, and its AI Translations feature drafts translations of a finalized survey, which Sprig notes are overwritten if regenerated.
The Synthesize Agent groups responses into named themes with summaries, and lets you open the responses behind every theme.
The Field Agent handles delivery across in-product, email, link and panel surveys, and asks contextual follow-ups based on each response, which researchers can turn off per study.
Researchers remain responsible for the final wording, the logic review and the interpretation, since every agent output is a draft until a person approves it.
Where Sprig fits. Sprig’s documented channels are in-product, email, link and panel surveys, which cover most of what a QuestionPro online program runs. The limitation is that it can’t host your historical responses, so the archive and the bridge studies stay your team’s work.
Frequently Asked Questions
Can I Import My QuestionPro Responses into Sprig?
Sprig documents no way to import historical survey responses as of November 11, 2026. Its CSV upload imports user attributes and users for targeting, not responses. Keep your QuestionPro history in a governed archive outside both platforms, with the CSV, SPSS file, review-mode PDF and a codebook for each survey, and report old and new data side by side.
What Is the Fastest Way to Rebuild a QuestionPro Survey in Sprig?
The fastest documented route is to download each survey’s review-mode PDF from QuestionPro and upload it to the Design Agent’s AI File Upload Builder, available on Sprig Enterprise plans. Mark logic with the documented prefixes first. MaxDiff questions and prototype links have to be added by hand, and every draft needs a line-by-line check against the original.
What Doesn’t Transfer from QuestionPro to Sprig?
From QuestionPro, historical responses, Apache Velocity custom scripting, looping and offline collection have no documented Sprig equivalent. Neither do constant sum, file upload, signature, maps and date questions, total unduplicated reach and frequency (TURF) analysis or Gabor-Granger. Custom variables, respondent authentication, Van Westendorp and several question types port only in part. Each needs a redesign decision from its owner before the rebuild starts.
How Long Does a QuestionPro Migration Take?
No published figure exists for how long a QuestionPro to Sprig migration takes, and the honest answer depends on three counts from your audit: surveys to rebuild, surveys to redesign, and tracked metrics that need a bridge study. The bridge studies usually set the length, since each runs through at least one reporting cycle. Work backwards from your QuestionPro contract end date.
When Does My QuestionPro Data Have to Be Exported?
Export before you cancel. QuestionPro’s data retention policy says a cancelled account’s data leaves its primary databases within 24 hours and all systems, including backups, within seven days. The policy also says that data can’t be recovered, and a separate GDPR page describes 30 days, so plan to seven. Run a final export after both platforms stop fielding together, because responses collected during that overlap aren’t in your earlier archive.
Will My NPS or Satisfaction Scores Change When I Switch?
Expect NPS and satisfaction scores to move when you switch platforms. Pew Research Center’s 2015 phone-versus-web experiment found a mean difference of 5.5 percentage points across 60 questions, with satisfaction items moving most. Run a bridge study, fielding the old and new surveys at the same time, on every tracked metric so you can measure the step and report a new baseline rather than a false trend.
How Long Should I Run QuestionPro and Sprig in Parallel?
No published rule sets a duration. The US Bureau of Labor Statistics ran 18 months of parallel collection before its 1994 survey redesign, and the University of Michigan ran web in parallel with phone for over seven years. Size the overlap with a power calculation and your reporting cycle, and end it when the bridge study reports.
Can Sprig Run the Conjoint and MaxDiff Studies I Built in QuestionPro?
Sprig documents conjoint and MaxDiff question types on Enterprise plans, conjoint on link surveys and MaxDiff in Surveys using Standard Format. Sprig documents no TURF analysis, market simulator or Gabor-Granger. A study whose deliverable is a simulation or reach curve needs its analysis run outside Sprig, or should stay on QuestionPro.
Does Sprig Support SSO and SCIM?
Sprig documents SAML 2.0 single sign-on on Enterprise plans with setup guides for six identity providers, and SCIM for deprovisioning on Enterprise plans. Sprig’s SCIM doesn’t create users, and only Okta and Microsoft Entra ID are verified. Plan user creation separately and test deprovisioning before you rely on it.
Can I Collect Survey Responses Offline in Sprig?
Sprig documents no offline collection as of November 11, 2026. QuestionPro documents offline surveys in its iOS and Android app that sync when connectivity returns. If offline fieldwork is central to your program, keep that program on a tool that supports it and migrate only your online surveys.
Should We Migrate at All?
Teams shouldn’t migrate a program that depends on offline collection, regional data hosting, or in-platform TURF, Gabor-Granger or a conjoint market simulator, none of which Sprig documents. For in-product, email and panel surveys with defined owners, a migration is usually manageable if you export first, rebuild trackers first and bridge every tracked metric.