Best Conjoint Analysis Tools for Mobile Phone Surveys
“Mobile conjoint analysis is where most survey platforms fall apart. Conjoint asks someone to compare several profiles side by side, and a phone screen is roughly a quarter the width you need to lay those profiles out the way desktop tools do. So the tools either shrink everything until it’s unreadable, or they make you scroll and swipe to see options you’re supposed to be comparing at a glance.”
I tested eight conjoint tools on a phone and scored each one out of five. OpinionX scored 5, Conjointly and Alchemer 3, Sawtooth and 1000minds 2, Qualtrics 1, QuestionPro 0, and SurveyMonkey a negative score for selling conjoint surveys that don't exist. Reasoning and a screen recording for each below.
What is conjoint analysis?
Conjoint analysis is a survey method for ranking the importance of things that carry multiple layers of information.
Imagine you're buying a new smartphone. Each phone has several attributes (brand, storage, colour, battery life) and within each attribute there are options (for brand, that might be iPhone, Samsung, Google Pixel). Conjoint analysis answers which attributes drive people's decisions most, and which specific options within each attribute they prefer.
The method goes back to Green and Rao's 1971 paper on conjoint measurement, and it now covers several distinct variants, from full-profile conjoint to the choice-based version most tools ship today.
Participants see sets of two to five profiles at a time. Each profile carries one option from each attribute, and the task is to pick the profile they prefer in each round. After enough rounds, the system can estimate which attributes matter most and which options are driving the choices. If you want the maths behind that estimate, we've written up how conjoint analysis is calculated separately.
What you do with the output is a separate question, covered in our guides to the conjoint market simulator and marginal willingness to pay.
The important part for this article is the word sets. The participant has to compare several profiles against each other in one glance. Everything below is about whether a phone screen lets them.
Why does mobile conjoint analysis UX matter?
Almost everyone writes their survey on a laptop and sends it to people holding phones.
SurveyMonkey's State of Surveys report found that 58.2% of survey responses globally were collected on a mobile device in 2024, against 48.4% in the US, with both trending up every year since 2021 (SurveyMonkey, 2025).
That shift has already changed how researchers write questions. The same report shows matrix questions appearing in 43% of surveys in 2015 and only 23% by 2024 (SurveyMonkey, 2025). Nielsen Norman Group's reading-comprehension work found that understanding the same content on a small screen is roughly twice as hard as on a desktop, and that's the reason behind the drop: a matrix makes you scroll left and right to read the row labels and reach the answer columns, and people stopped putting up with it.
The cost of ignoring this is measurable. In a randomised experiment run on GfK's KnowledgePanel, 4,555 people were assigned a device and a survey template. Break-off rates dropped by half when the survey used both a responsive template and mobile-friendly question design, and median completion time fell from 24 minutes to 18 (Barlas, Thomas and Graham, GfK, presented at FESAC, June 2015). Their traditional template didn't reformat to fit the screen at all, so participants had to scroll left and right just to read the question. Pew's tracking puts smartphone ownership among US adults at 91%, so this isn't a minority of your sample.
Conjoint surveys fail on exactly this, because almost all of them use a horizontal table layout.
How were these tools scored?
Five criteria, applied to the same task on every platform: vote on one set of conjoint profiles, on a phone.
Can you see every profile in the set without scrolling? The method depends on comparison, so this one is weighted heaviest.
Can you see every attribute and option without scrolling? A profile you have to scroll through isn't a profile you can compare.
Is there any horizontal scrolling or swiping? GfK's criterion, and the clearest single predictor of drop-off.
How much of the screen carries actual information? Padding is space a phone doesn't have.
Can you reach the next button without hunting for it?
I recorded everything on an iPhone 13, which has a smaller screen than current models, at the default 100% zoom. Where a platform wouldn't let me build and test a conjoint survey myself, I've said so and scored accordingly.
Which tool is best for mobile conjoint analysis?
OpinionX: 5/5
OpinionX was the only platform in this test where voting on a set of conjoint profiles needed no scrolling at all.
Profiles stack vertically, and each profile is laid out in two columns so more information fits into the same vertical space. On the recording below, the full set is visible and votable on a single screen, on an iPhone 13, at 100% zoom.
Full disclosure, obviously: I run OpinionX. The recordings are there so you don't have to take my word for the scoring. You can also try the conjoint builder yourself on the free tier and put it on your own phone, or watch a live demo first.
Conjointly: 3/5
Conjointly uses a table view on desktop and switches to a mobile version that turns each column into a separate card, which correctly avoids horizontal scrolling.
The problem is density. You still scroll vertically to see all the options in a set, so you're comparing profiles from memory rather than side by side. The same survey fits into less than half the screen space on OpinionX.
There's also a structural limit worth knowing before you pick it. Conjointly treats survey questions (multiple choice, rating scale, text input) differently from research methods (conjoint analysis, MaxDiff).
You can add extra survey questions to a study, but each questionnaire can carry only one research method section. So you can't put MaxDiff and conjoint in the same survey, and you can't run two separate conjoint sections in one survey. If you need to mix methods, that's a hard stop. We go through the rest of the differences on the OpinionX vs Conjointly page.
On price, Conjointly's Professional plan is $2,895 per team per year (Conjointly pricing, checked 25 August 2026), which is a little over three times OpinionX's $900 Analyze plan.
^ Screen recording taken in July 2025. Interfaces change, so check Conjointly's site for their current version.
Sawtooth Software: 2/5
Sawtooth Discover scores 2 out of 5 because it needs both horizontal and vertical scrolling.
The horizontal swiping cards mean only one profile is ever on screen when you vote. Then you scroll vertically to find the "Next" button. The mobile experience was also buggy: I hit error messages and got stuck more than once.
Worth updating since I first wrote this: Sawtooth no longer publishes prices for Discover Professional or Premier, both of which now say "Request Pricing". They have added a free tier capped at 50 participants per survey, which does at least mean you can test the mobile experience yourself before committing to a quote. Do that. There's a fuller breakdown on the OpinionX vs Sawtooth page, and a three-way comparison if Conjointly is also on your shortlist.
QuestionPro: 0/5
QuestionPro offers no examples and no free trial of their conjoint surveys, on the grounds that we're all too stupid to test it ourselves. Their words: "Due to complex nature of the analysis and the level of expertise required to support such an offering, unfortunately we cannot provide free trials."
I've said it before and I'll say it again. People need to stop acting like conjoint analysis is a mind-meltingly complicated research method. It isn't. The reason companies say it is, is so they can quote you five figures for what is otherwise just another survey method. Advanced methods on QuestionPro sit on a plan you have to contact sales for.
A big 0/5, and please, let's put an end to the condescending complexity warnings. The feature-by-feature version is on the QuestionPro comparison page.
Alchemer: 3/5
Alchemer earns a 3/5. No horizontal scrolling, and the "next" button sits in a sticky footer, which is hard to spot at first but cuts down scrolling once you've made your choice.
What costs it the other two points is empty space. The levels and options take up about 22% of each voting card in the example below, so you scroll through padding that a tighter layout wouldn't need.
SurveyMonkey
Whatever their blog posts and website suggest, SurveyMonkey does not have a conjoint analysis question type. What they have is a service: you commission their in-house researchers to build a standalone conjoint project for you, quoted per project.
So I couldn't test the mobile conjoint UX, because there's no conjoint survey there to test. Yeah, that'll be a -1/5. There's more detail on that gap on our SurveyMonkey conjoint comparison page.
Qualtrics: 1/5
Those sneaky people at Qualtrics, it's like they hacked my scoring criteria. Their conjoint UI on mobile needs no horizontal or vertical scrolling and has high information density.
It achieves this by not really being a conjoint survey. Qualtrics' demo mode doesn't allow editing, so I couldn't confirm whether a better format exists that just wasn't available to me. The recording below is genuinely labelled "conjoint analysis" inside Qualtrics. Make of that what you will.
1000minds: 2/5
1000minds isn't a standard conjoint tool. It's built for policy decision-making and comes with a set of opinionated constraints I found frustrating.
Their patented PAPRIKA scoring algorithm is what they sell on, and it also appears to restrict you to two profiles per voting set. You can't edit options after publishing either, so I ended up in a loop: publish, test, delete all participants, unpublish, edit, republish, test again.
On mobile it uses horizontal swiping cards, the same as Sawtooth. As the recording shows, that left me swiping back and forth between profiles trying to hold both in my head before voting.
Summary: mobile conjoint analysis scores compared
| Platform | Score | Why |
|---|---|---|
| OpinionX | 🟢 5/5 | The only platform tested where a full set of profiles is visible and votable without scrolling. |
| Alchemer | 🟡 3/5 | Vertically stacked profiles are the right call, but the wasted space makes voting inefficient. |
| Conjointly | 🟡 3/5 | No horizontal scrolling, but too little on screen at once. One research method per survey. |
| Sawtooth | 🔴 2/5 | Horizontal and vertical scrolling, plus errors mid-survey. Free tier now lets you test it. |
| 1000minds | 🔴 2/5 | Horizontal swiping blocks comparison, and PAPRIKA caps you at two profiles per set. |
| Qualtrics | 🔴 1/5 | Scores well on scrolling by barely qualifying as conjoint analysis. |
| QuestionPro | 🔴 0/5 | Won't let you test conjoint yourself, on the grounds that it's too complicated for you. |
| SurveyMonkey | 🔴 -1/5 | Markets conjoint surveys. Doesn't have a conjoint question type. It's a commissioned project. |
Seven of the eight scored 3 or below, and they fail in one of two ways: a horizontal layout a phone can't fit, or a vertical one that shows too little at a time to compare anything.
For a full price-by-price breakdown across ten platforms, including the ones that don't publish figures, see our separate guide to conjoint analysis software compared on price.
What should you check before picking a mobile conjoint analysis tool?
Build the smallest conjoint survey each shortlisted platform will let you make, then open it on your own phone. It takes about five minutes a tool.
Vote on a real set. Not the marketing screenshot, not the desktop preview. Send yourself the live link.
Count the scrolls. Every scroll or swipe between profiles in one voting set is a comparison the participant has to make from memory.
Use your longest attribute. Layouts that hold up with three short options break when one option is a full sentence, so build the test with your actual copy.
Try it one-handed, since most people answer surveys standing up.
Run the number of profiles you actually need. A layout that works for two profiles per set often collapses at four.
Ask whether participant-level data exports. Otherwise you're stuck with whatever charts the tool decides to draw.
Watch someone else do it. You built the survey, so you know what everything means. They don't.
If you can't build and send a test survey without talking to a salesperson, that tells you something about the platform before you've seen the interface.
When does mobile conjoint analysis UX matter less?
A controlled panel on desktop. If you're fielding through a panel provider that routes to desktop, or running sessions in a lab, mobile layout stops being the constraint and methodological depth matters more.
Academic work needing adaptive designs. If you need adaptive conjoint or a deep base of academic literature behind your method choice, Sawtooth's breadth is the better fit and the mobile experience is a cost you accept.
Very small attribute counts. Two attributes with two levels each will fit a phone on almost any layout. The differences here show up once your design gets realistic.
If conjoint turns out to be the wrong fit entirely, we've listed the alternatives worth considering, including how conjoint compares to AHP.
Conjoint isn't always the right method either. We've written on when not to use conjoint analysis, and if your question is about the relative importance of a list rather than trade-offs between bundles, MaxDiff is usually the better instrument. The MaxDiff tools have their own mobile problems, and Uber's combined MaxDiff and conjoint study shows what running both looks like in practice.
Frequently asked questions
Which conjoint analysis tool is best for mobile surveys? OpinionX scored 5/5 in this test, the only platform where a full set of conjoint profiles is visible and votable on a phone screen without scrolling. Alchemer and Conjointly both scored 3/5.
Why do conjoint surveys work badly on phones? Most conjoint interfaces are built as horizontal tables, with one column per profile. A phone screen isn't wide enough, so the tool either shrinks the text or makes participants scroll and swipe between profiles they're meant to be comparing at a glance.
Does mobile survey design actually affect results? Yes. GfK's randomised experiment with 4,555 participants found break-off rates halved and median completion time dropped from 24 minutes to 18 when surveys used a responsive template with mobile-friendly question design.
How many survey responses come from mobile? 58.2% of survey responses globally were collected on a mobile device in 2024, according to SurveyMonkey's platform data, with the US at 48.4% and both figures rising every year since 2021.
Can you run conjoint analysis on SurveyMonkey? Not as a survey you build yourself. SurveyMonkey has no conjoint question type; conjoint is sold as a project their in-house research team builds for you, quoted separately.
Only three of these eight made it easy to finish a conjoint survey on a phone, and none of that shows up on a feature list. If you need people to test yours with, we've written up where to find free survey participants.
Want more like this? Over 42,000 researchers and product people get The Full-Stack Researcher in their inbox. Subscribe for the next teardown.
Every method on OpinionX, conjoint included, works on the free tier: $0, unlimited surveys, unlimited researcher seats, capped at 25 participants per survey so you can check the design works before you pay $900 for the year (full pricing).