Services / Programming and hosting

Programming & Hosting

We build and host your instrument on Qualtrics, then wire it directly to the myCLEARopinion panel. You get an enterprise survey platform and a verified B2B audience on one project, without managing a software licence, a sample vendor and a programmer separately.

Send a questionnaire in whatever state it is in — a Word document, a deck, a half-finished draft. We program the logic real B2B studies need, test it properly before anyone sees it, and run a soft launch so that incidence, length and drop-off are known quantities rather than end-of-field surprises.

At a glance

PlatformQualtrics, hosted by us
LogicNested quotas, piping, masking, randomization
AdvancedMaxDiff, conjoint, van Westendorp, TURF
DevicesDesktop, tablet and phone, tested on each
SampleOur panel, your list, or both
You provideQuestionnaire, quotas, timeline
We provideBuild, QA, hosting, monitoring, data
48 hrs
Typical first test link
10%
Soft launch before full field
30+
Languages supported
2
Revision rounds included

How it works

Five stages from document to clean dataset.

The expensive mistakes in a survey are all made before launch. Most of this process exists to find them while they are still cheap.

01

Review the instrument

We read the questionnaire before quoting it and come back with an estimated length, the questions that will cost you completes, and anything that cannot be programmed as written.

02

Program

Built in Qualtrics with the logic the design actually calls for: nested quotas, piping, masking, carry-forward, randomization and custom validation, styled to your brand if you want it.

03

Test properly

Every path walked by a person, not just the happy one. Machine-generated test data fills the dataset so you can check your tables and banners before a single real respondent arrives.

04

Soft launch

About ten percent of sample goes first. Real incidence, real length, real drop-off points, all confirmed against the estimate before the rest of the budget is committed.

05

Field and deliver

Live quota monitoring through field, then cleaned data in the format your analyst wants: SPSS, Excel, CSV, or banner tables built to your spec.

Capability

If the design calls for it, we can program it.

Nothing here is an upgrade or an add-on module. It is what a B2B instrument routinely needs.

Nested and interlocked quotasPiping and dynamic textCarry-forward and maskingMaxDiffChoice-based conjointVan Westendorp and Gabor-GrangerTURFConstant sum and allocationDrag-and-drop rankingImage and concept testingHeat map and markupVideo open-endsFile uploadAttention and trap questionsMulti-languageCustom branding

What the data says

Length is the decision that costs you the most.

Every extra question feels free when you are writing the questionnaire. It is not free once it is in field, and the bill arrives as abandoned interviews.

Completion rate by instrument length

Share of respondents who start and finish, B2B audiences

100%75%50%25%0%94%44%51015202530Length of interview (minutes)

Past twenty minutes the curve turns sharply, and it turns hardest on the senior respondents you most want to keep. This is the conversation we would rather have with you at review than at end of field.

Device mix

How B2B respondents actually take surveys

Desktop51%
Phone41%
Tablet8%

Nearly half your respondents are on a phone, often on a job site. A wide grid that looks fine on your monitor is where they abandon.

What QA catches before launch

Issues found per 100 programmed questions, by type

051015Logic and skip errorsPiping and text errorsQuota misconfigurationMobile layout breaksTranslation issues149653Issues found per 100 programmed questions

None of these are exotic. They are what happens to any instrument of real complexity, which is why testing is a stage rather than a favour.

What slows a build

Ranked by how much each tends to add to the timeline

Late questionnaire changes. A new question in the middle re-tests every path after it, not just itself.

Interlocked quotas. Cells that depend on each other take longer to build and far longer to verify.

Conjoint and MaxDiff. Design generation and validation are their own step before programming starts.

Translations. Every language multiplies the testing, and layout breaks differently in each one.

Approval chains. The build is usually waiting on sign-off, not on programming.

Figures on this page are placeholders for layout review. Replace with observed platform data before publishing.

Compliance and data

Collected properly, stored properly, handed over cleanly.

Hosting someone else is hosting their data. That comes with obligations, and we would rather you knew exactly how we handle them.

Consent at entry. Respondents are told who is collecting, what for, and how long it is kept, before the first question.

Enterprise platform security. Hosted on Qualtrics rather than a homegrown stack, with the certifications an enterprise procurement team will ask you about.

Identifiers separated. Personal identifiers stay apart from response data, so what reaches your analyst is anonymised by default.

Deletion and access requests. A respondent asking to be removed is handled, and you are not the one fielding the request.

Your data is yours. Full export at the end of the project, in a format you can actually open, with no hostage-taking on the way out.

What to send us

A questionnaire in any state, and four details.

It does not need to be clean. Half of what we do at review is turn a document into something programmable.

The questionnaire. Word, Excel, a deck or a link to an existing survey. Mark the skip logic however you already mark it.

Quota plan. Cell definitions and targets, and whether they interlock.

Where the sample comes from. Our panel, your own list, a third-party provider, or a mix.

Languages and branding. Which markets, and whether it should look like your brand or stay neutral.

Delivery format. SPSS, Excel, CSV or banner tables, and who is doing the analysis.

Need respondents too? Survey sample plugs straight into the instrument we build, and full service covers design through final report on one statement of work.

Common questions

Before you send the questionnaire.

Do we need our own Qualtrics licence?

No. The study runs on our instance, so there is no licence to buy and no seat to administer. If you already have Qualtrics and would rather it lived in your account, we can build there instead.

Can you program a study we are fielding with someone else?

Yes. Programming and hosting is sold on its own. We set up entry and exit redirects for whichever sample provider you are using and reconcile against their statuses.

How long does a build take?

A straightforward instrument is usually a test link within two business days. Conjoint, heavy interlocking or multiple languages add to that, and we will tell you how much at review rather than after you have committed to a date.

What if we want changes after it is built?

Two rounds of revisions are included before launch. Changes after soft launch are possible but not free of consequence: anything that alters a question mid-field splits your dataset, and we will say so plainly before doing it.

Can it look like our brand?

Yes, and sometimes it should not. Branding lifts response among your own customers and biases answers in a competitive study. We will give you an opinion either way.

What do we get at the end?

A cleaned dataset in your format of choice, a full variable map, and banner tables if you want them. Everything exports; nothing is trapped in a viewer you have to keep paying for.

Send us the questionnaire. We will tell you what it will really cost you.

A review comes back with an estimated length, the questions that will cost you completes, a build timeline and a price. No obligation, usually within one business day.

Contact usDownload Panel BookExplore B2B verticals