What makes you install an app from a developer you've never heard of?

Shopify Apps
1

Hey there! I build Shopify apps, and I’m trying to understand this from your side rather than guessing.

When you’re browsing the App Store and you find something that does what you need, but it’s from a developer you don’t recognise and it has few or no reviews — what happens next? Do you install it anyway, or do you go back and pick the one with 2,000 reviews even if it’s more expensive or more than you need?

Three things I’d genuinely like to know:

  1. What would make you try an unknown app on your store: free plan, a demo store you can click through, screenshots, a video, someone answering fast when you ask a question?
  2. What makes you close the tab immediately?
  3. Has a small app ever worked out better for you than the big one in the category? What made the difference?

I want to know what actually earns trust before an install, because from the developer side it’s very easy to build the wrong thing and assume it’s a pricing problem.

And if you’ve got a story about an app that let you down, that’s useful too.

2

For me, a small app earns a first install by being clear about the exact problem it solves and making the risk easy to control. A free or low-cost trial helps, but a focused listing, current screenshots, transparent pricing, and a developer who answers specific questions quickly matter more than review count alone. I would still choose the larger app for a business-critical workflow when reliability and support history outweigh the extra cost, but for a narrow need, a smaller app can be the better choice when it is simpler and more responsive.

3

Answering from the other side, since I am the unknown developer. Fourteen months, around 500 installs, 17 reviews.

Review count is a lagging signal, not a quality one. I read 483 one and two-star reviews across 126 apps, the big ones included. Support was the top reason merchants left (152), then billing (84), then broke after a theme or platform update (72). Missing feature was 12. The 2,000-review app fails on exactly those, and the count hides it.

What actually lowers your risk on a small app is testable before you install:

  • Does the listing name the outcome, or only the mechanic
  • Does the pricing page say plainly what happens when you uninstall
  • Does the developer answer a specific pre-install question, and how fast

That last one costs you one message and tells you more about the support you will get than any review count.

What closes the tab for me: a listing that lists features without ever saying what changes for the store.

Disclosure, I build a post-purchase upsell app, so I have a stake in the answer.

4

This is the most useful answer I have gotten anywhere, so let me use it properly rather than just thanking you.

Your three tests, applied to my own listing, honestly:

Outcome or mechanic: my opening line says the app does points and cart recovery. That is mechanic. The paragraph underneath does name the problem, but the five bullets are pure feature list. So I fail my own reading of your test, and now I can see it.

What happens when you uninstall: not stated anywhere. I had not considered that a merchant would want to know that before installing, which in hindsight is obvious. A merchant in another thread here mentioned checking for leftover script tags after removing an app, so this seems to be a real and unaddressed worry.

Pre-install question and response time: we do answer fast, but as you say, that is not testable from the listing. Nothing there tells you it is true.

If you have ten minutes to look at ours and tell me what closes the tab, I would take that seriously. Livy Loyalty, we launched August 5th. No obligation, and I would rather hear the harsh version.

One thing I might be able to trade. You read 483 negative reviews to find those patterns. I am currently emailing the merchants who wrote that kind of review, one at a time, opening with their specific complaint and asking what would make them try an unknown app. Ten this week. Different angle on the same question: you have what they wrote, I am getting what they say when asked directly. Happy to share what comes back, patterns and numbers, once I have enough to be worth anything.

The 152 versus 12 split is the part I keep rereading. Support is the top reason people leave and missing features is almost noise, and yet most of us spend our time on features.

And your own numbers, 500 installs and 17 reviews in fourteen months, are the most useful benchmark I have seen. That is roughly one review per thirty installs, which is well below what people quote when they talk about this.

5

The “risk easy to control” framing is the part I had not put into words, thank you. Two follow ups if you have a minute.

When you say transparent pricing, is that mainly about knowing the real cost as you grow, or about knowing what happens if you stop paying? Someone else here pointed out that almost no listing says what happens when you uninstall, and I had never thought about it.

And on business critical workflows going to established apps: is loyalty business critical for you, or is it in the category where you would take a chance on someone small

6

Adding one hard data point, since we are on the zero-review side of this right now. I run a small portfolio of four apps, all flat-priced, newest listings with no reviews yet.

We ran Shopify search ads on one of them in August: position 2 to 3 on the exact search term, 14 clicks, 0 installs. Merchants found the listing, opened it, and every single one walked. With zero reviews the listing has to carry all the trust by itself, and ours could not. Which matches what JohnQQ said: nothing about the risk was controllable, the only plan was paid.

So last week we added a free plan to that app as a direct test of the risk theory. If the blocker was price risk, installs should start. If it stays at zero, the blocker is deeper: social proof, or the listing copy itself. Happy to report back either way, because this thread is asking exactly the question we are spending money to answer.

On your pricing follow-up: as a buyer I think both matter, but “what happens if I stop paying” is the sharper one for anything that touches the theme. An app that leaves code behind after uninstall breaks the easy-to-undo promise even when the subscription was cheap, and merchants have learned that the hard way often enough that it now prices into the install decision itself.

7

I opened your listing. Before the copy, three things in the pricing panel, because that is where I stopped.

Your third plan is named Gratuit, French for Free, and it is priced at 49 dollars a month. I browse from France so I get the French listing. Whatever the cause, a plan called Free that costs 49 is the fastest way to lose a stranger.

That same plan lists its features in English while the rest of the page is French, so the block reads as unfinished.

And all three plans show the same caps: up to 1,000 orders and 1,000 credits. From outside, 9, 15 and 49 dollars buy the same limits, and all three say additional usage fees apply. I cannot tell what more money gets me, and I cannot tell what the app costs me next month.

That is the tab closing, and it is not the copy. Zero reviews, a trial that needs a card, and an uncapped usage fee mean there is no version of this where I control the risk.

One thing you are undervaluing: you have a demo store link. That is the only element on the page that lets a stranger test the app without trusting you first, and it sits below the fold with no reason to click it. I would put what it lets you do in the first screen.

8

Those numbers are really interesting. I had been assuming our problem was visibility. You bought visibility outright and it converted at zero.

We are working on adding a free plan too, this same week, for the same reason. I would love for us to share our results.

Are you running those ads only inside the App Store, or on other platforms as well?

9

Hi @julifredes Love that you’re asking instead of guessing :slight_smile: From my experience the first installs come from a clear listing + answering support fast, reviews come on their own after that then it just becomes a rinse and repeat after that. You’ve done the hard part already, good luck with the app!

10

App Store search ads only, nothing external, and that was deliberate. Two reasons.

Intent: a merchant typing the exact problem into App Store search is the warmest traffic an app can get. If that traffic converts at zero, the listing is the bottleneck, and external channels would just pay more to teach the same lesson to colder visitors.

Math: our CPC came out at 6.57 dollars at position 2 to 3. Against a 6.99 flat monthly price, even a perfect funnel needs months of retention to pay back one click. Any platform with a higher CPC is ruled out before conversion even enters the calculation. Flat low pricing and paid acquisition basically do not mix; content and search presence are what is left, which is why I spend my time here.

And deal on sharing results. Our free plan went live August 31, so I will post numbers once it has had a fair window. I am genuinely curious whether yours moves differently, because you have reviews on the listing already, and that isolates the one variable my zero-review data cannot.

11

For me, reviews are probably the first filter, but they’re not an automatic no. If the app is free or has a trial, I’m much more willing to test an unknown developer, especially if the listing actually shows the app working instead of just having a bunch of feature screenshots.

What usually makes me close the tab is when I can’t figure out what the app actually does within a minute or two. Vague descriptions, screenshots that don’t show the actual workflow, pricing hidden until after installation, or an app that wants a bunch of permissions without explaining why would make me skip it.

I’ve had smaller apps work out better too. Usually it’s because they solve one specific problem without trying to turn it into some giant platform. I’ve had that experience with SEO tools as well. Something like SiteGuru can be more useful for a smaller store than a huge SEO suite when you mostly need to know what needs fixing rather than another massive dashboard.

For an unknown app, I think being able to safely try it matters more to me than having thousands of reviews. If I can understand what it does, see the actual interface and remove it easily if it doesn’t work, I’m much more likely to give the smaller developer a chance.

12

Thank you. Clear listing plus fast support is what almost everyone in this thread has converged on, and it is not what I would have guessed a week ago. I had assumed features and visibility.

The listing half turns out to need real work: someone here reviewed ours today and found our pricing panel is broken in some languages. So I am doing the unglamorous version of your advice first

13

This is the most useful merchant answer I have gotten, thank you.

The one to two minute test is brutal and fair. And the specific things that close the tab, vague descriptions, screenshots that do not show the real workflow, pricing hidden until after install, permissions with no explanation, are all checkable on my own listing rather than matters of taste. I already know we fail at least one of them.

Your point about seeing the app working rather than feature screenshots connects with something a developer said here today: we have a demo store you can click through, and we buried it at the bottom of the page. It is the one thing that lets you test us without trusting us first.

The permissions one I want to ask about, because loyalty apps ask for a lot. We need customers, orders and discounts to work at all. Would explaining why each one is needed actually help, or is the amount itself the problem regardless of the explanation?

And your point about smaller apps winning by solving one specific problem instead of becoming a platform is one I want to hold onto, because the pull in this category is towards adding everything.

14

One correction before anything else, because it affects your experiment: we have zero reviews. Same as you. If you were expecting our free plan test to isolate the review variable, it will not, but that makes the comparison cleaner rather than worse. Two zero-review listings adding a free plan within a week of each other, different categories.

Your CPC math is the part I had not thought through. 6.57 dollars a click against a 6.99 flat price means paid acquisition is ruled out before conversion even enters it. Our entry plan is 9 dollars, so we are in the same trap and I had not done that arithmetic.

And the intent argument is the one that lands hardest. If the warmest traffic that exists converts at zero, spending on colder traffic just pays more for the same lesson. That reframes every “we need more visibility” conversation we have been having internally.

One caveat on our side: another developer just reviewed our listing and found our pricing panel is genuinely broken in some languages, plan names shifted against prices. So our zero may be partly self-inflicted rather than structural. Fixing that first, then the free plan, so we know which one did what.

15

You are right and it is worse than one plan name. Gratuit at 49 is our Growth plan. The names and the prices have both been translated wrong, so a merchant browsing from France sees something that makes no sense at all. The caps are wrong there too: in English the plans show different limits, and the translation flattened them all to the same numbers. So the thing you could not make sense of was not the offer, it was a broken translation of it. I am going through every published language now, not just French. What a mess.

The real difference between Founder and Growth, which is invisible on the page even when it is translated correctly: Founder is a promotional plan for early adopters, limited to a number of spots. That is worth stating plainly rather than leaving a stranger to work out why one plan costs a fifth of another for similar features.

On controlling the risk, you are right that there is currently no version where a stranger can. Zero reviews, a trial that wants a card, and a usage fee. We are adding a free plan this week precisely so the app can be tried without a card at all.

On the demo store: it shows for me at the top left, under the install button, so I am curious where it sat for you. Is it that it appears lower on your screen, or that nothing on the page gives you a reason to click it? Because if it is the second, moving it will not fix anything and what it needs is a line saying what you can actually do in there.

Thank you for spending the time on this. It is by far the most valuable comment we have had, and I will report back on what changes and whether it moves anything, since that is a datapoint you can use too.

16

It was exactly where you say: top left, under Install. So it is your second hypothesis, and moving it would change nothing.

The label is the problem. “View demo store” names a destination, not what I get. I cannot tell whether it is a live storefront running your app or a folder of screenshots, and finding out costs a click I have no reason to spend.

Say what happens in there: Earn and redeem points on a live store, no install. Same sentence in the first gallery caption.

That one is worth doing today, ahead of the free plan, for a boring reason: it is a ten-minute copy change and the free plan is a week of work. They remove different objections, so ship the cheap one first.

On Founder: state the number of spots left. A promotional plan without a count reads as arbitrary pricing, which is the thing you are trying to get away from.

17

Zero reviews on both sides makes it a cleaner experiment, agreed, and your sequencing is right: fix the translated pricing first, then add the free plan, otherwise neither of us can tell which change moved the number.

One request for when you post results: note the date each change went live, and if Partners gives you installs by visitor locale, split by that. Your French page was showing a nonsense offer for weeks, so part of your zero may be regional rather than structural, and that would be a useful third data point next to my English-only listing. I will post ours the same way, with the free plan’s go-live date, so the before and after windows line up.

18

I am on the zero review side of this too, so this is from the developer seat rather than the merchant one.

The thing that changed how I write a listing was realising the merchant is not judging the app. They are judging how expensive it is to find out the app is wrong. So I try to make the first two minutes cheap. The free tier does one real rule end to end instead of a crippled demo, the install asks for the smallest scope set that can do the job, and the listing says plainly what the app does not cover rather than leaving them to find that out after installing.

On the pricing panel point above, I think that does more damage than the copy does. A plan name in one language sitting next to a number in another reads as unfinished, and unfinished is the one thing an app with no reviews cannot afford.

One thing I would add about reviews. The gap between install and review is long for small apps, and asking early gets you nothing. I only ask after the merchant has actually used the feature at least once, because a review request before the value has landed just tells them I care about the score.

19

Fourth data point for the ad math, because ours matches UprightApps almost exactly: exact category term, position 2 to 3, five clicks at 7 dollars each, zero installs. Same shape, different category. So whatever the fix is, it is not visibility.

The thing I keep coming back to in this thread is that every test people named - free plan, demo store, uninstall behaviour - is a way to let the merchant check us cheaply. Backup is a rough category for that: you cannot tell whether a backup app works without breaking something, and nobody breaks their store to evaluate software. Our answer was to break our own instead. We run a live store, damage it on purpose, restore it, and publish the results field by field, dates and all. It has not moved installs yet - the listing has no reviews, and I suspect nothing outruns that number in the first minute of looking - but support conversations start differently when the first question is already answered in public.

For context, mine is StoreVault, a backup and restore app, so the trust bar is probably as high as it gets: the product only matters on the worst day.

20

Four listings, four categories, the same shape: position 2 to 3, a click around 7 dollars, zero installs. At this point I would treat it as a property of zero-review listings rather than of any of our apps, which is oddly reassuring: it means the fix is on the page, not in the product.

Your break-it-and-restore-it log is the most honest version of the cheap check idea in this thread, because it answers the one question a merchant cannot safely ask themselves. I would put a single line from it in the first gallery caption, dates included, for the same reason Bristan gave julifredes about the demo link: proof only works if it lands in the first minute, and a listing gets about that long.