How Do You Do ASO for a New App With No Reviews or Rankings?

A new app has none of the inputs ASO guides assume. It has no ratings, no reviews, no install history, no keyword rank and no category position. Most ASO advice is about improving those signals, which makes it useless on day one.

What is left is small and worth doing properly: the fields you write yourself, the screenshots, and a plan for getting the first ratings quickly. Everything else follows from installs you buy or earn somewhere else, which is why a new app’s ASO plan and its acquisition plan are one document.

What can you actually do for ASO before you have any reviews?

Before you have reviews you can control the metadata and the creative, and that is the entire surface available. The title, the subtitle, the keyword field on iOS, the short and long description on Google Play, the screenshots and the preview video all ship on day one and none of them depend on history.

Write them around two or three specific phrases rather than your category. A meditation app competing on “meditation” will not appear, while the same app targeting a phrase describing one specific job has a chance, because almost nobody is competing for it.

The screenshots deserve more of your time than the keywords do. A new app has very little search traffic to convert, so nearly every install in the first months comes from someone following a link or an ad. Those people see the listing, and the first three screenshots decide whether they continue.

What you cannot do is anything that depends on scale. Store listing experiments on Google Play need traffic to reach significance, keyword rank tracking tells you nothing when you rank for nothing, and review response has no reviews to respond to yet.

So the first month of ASO is a writing and design job with a measurement plan attached, not an optimisation job. Treating it as optimisation produces weekly metadata changes that cannot be evaluated.

How many ratings does a new app need, and how fast can you get them?

A new app needs enough ratings to clear four stars and look credible, which is usually a few dozen rather than a few hundred, and the native prompt gets you there faster than founders expect. Star ratings accumulate much faster than written reviews.

Phiture’s analysis of the iOS rating prompt across top US iOS apps found that 13.5% of users shown the prompt submit a rating, while 0.07% leave a written review. That second figure means about seven written reviews per thousand prompts, so review text is a slow asset and star count is not.

Apple allows up to three prompts per user per 365 days, and does not allow gating features behind a rating, offering anything in exchange for one, or showing the prompt only to users you expect to be positive. Those rules are enforced.

The variable you control is timing. Fire the prompt immediately after the app has done the thing the user installed it for, not on first launch and not after an error. A prompt shown before the app has earned anything collects the rating of a stranger.

Asset	How fast it accumulates	What it does for you
Star rating	13.5% of prompts shown (Phiture)	Clears the four-star credibility threshold
Written review	0.07% of prompts shown (Phiture)	Gives browsers something specific to read
Keyword rank	Months, driven by install volume	Earns search impressions you did not pay for
Listing conversion rate	Days, once there is traffic	Multiplies every install you buy

Read that table as a sequence rather than a menu. Ratings and conversion rate are available in weeks. Rank is not, and planning the launch around it wastes the window.

Do app ratings really affect downloads that much?

Ratings affect downloads substantially, and the specific numbers everyone quotes are weaker than they look. The direction is well supported: according to AppFollow’s analysis of rating impact, apps rated 4.0 and above take around 80% of app store revenue, and apps sitting at three stars lose roughly half of their possible downloads.

The famous figures are a different matter. Almost every ASO guide repeats the same two survey figures: moving from two to three stars raises conversion by 340%, and from three to four by a further 89%. Both come from an Apptentive consumer survey published in October 2015, which asked 350 US smartphone owners which apps they would consider installing.

Stated willingness is not measured behaviour, 350 people is a small sample, and 2015 is three App Store redesigns ago. The numbers are quoted as conversion lift anyway, usually without the method attached, which is how a decade-old intent survey became the most-cited statistic in the category.

Use it as direction and not as a forecast. Being above four stars matters, and the gap between 4.3 and 4.6 matters much less than these figures imply.

So treat the rating as a threshold to clear, not a number to keep tuning. Get above four, hold it there by fixing what the one-star reviews actually say, and spend the rest of the effort on the listing creative, where the response is measurable in days.

Which keywords should a brand new app target?

Target long-tail phrases with real search volume and almost no competition, which in practice means two to four word phrases describing a specific job rather than a category. Both stores weight install volume and engagement, which a new app does not have, so competitive terms are unavailable regardless of how well the metadata is written.

Pick on the ratio of volume to difficulty rather than on volume alone. A term with moderate volume and low difficulty beats a high-volume term you cannot place on, because ranking second on a modest term delivers installs and ranking thirty-eighth on a big one delivers none.

Put the most valuable phrase in the title, the second in the subtitle, and use the iOS keyword field for variants, misspellings and adjacent nouns. Google Play reads the whole listing rather than a keyword field, so the long description carries that load there.

Your own brand name is the one term you will rank for immediately and it needs checking. An app that does not appear first for its own name usually has a name colliding with an established app or a common word, which is a naming problem and gets more expensive to fix every month.

Review the list after eight weeks against what actually happened. The terms a new app ends up ranking for are rarely the ones it targeted, and the correct response is to take the accidental wins seriously rather than to keep pushing the original list. The broader mechanics of both stores are in how to optimize an app for the App Store and Google Play.

What goes in the listing when you have no social proof?

With no social proof the listing has to carry the argument on its own, which means the first screenshot states the outcome rather than showing the interface. A browser with no reviews to read and no rating to trust is deciding entirely on what you claim and how specific the claim is.

Specificity substitutes for proof better than polish does. A caption naming a concrete result reads as a claim someone would have to stand behind. A caption calling the app simple and beautiful reads as nothing, and it is what most new listings lead with.

Match the listing to where the traffic is coming from. If installs are arriving from one creator or one ad angle, the first screenshot should repeat that promise in those words. Custom product pages on iOS let you run a separate listing per source, which is the single most useful ASO feature available to an app with no organic traffic to protect.

Say what the app does not do. A new app with no reviews has nobody to vouch for its limits, and naming the limits yourself is the cheapest credibility available. It also reduces refunds and one-star reviews from people who installed expecting something else.

Leave the preview video until there is traffic to test it against. Production is expensive, the conversion effect is inconsistent across categories, and the same budget spent on three screenshot variants tells you more. A fuller treatment of the launch problem is in how to market an app with no users and no reviews.

Can paid ads buy you a ranking on day one?

Paid ads buy you presence immediately and ranking only indirectly. An Apple Ads placement appears above organic results for a term you cannot rank for, which is the only way a brand new app shows up in store search at all in its first weeks.

The inventory is also changing. AppTweak’s tracking records a further ad placement in App Store search results introduced by Apple in March 2026, which means more slots and more competitors bidding on exactly the terms a new app is trying to establish organically.

Ranking follows installs, and installs per day count for more than installs in total in both stores. Buying volume does raise that figure, so paid and organic are connected rather than separate. What a burst cannot do is make the ranking stick, because engagement and retention are weighted too and low-intent installs supply neither.

Use paid for two jobs at this stage instead. First, generating enough listing traffic to make conversion rate testing possible, which is the one ASO lever that responds in days. Second, finding out which promise converts, so the metadata can be rewritten around the winner rather than around a guess.

Running the listing, the creative and the paid plan as one piece of work is what makes this stage move, and it is the service described on our app launch page.

How do you tell whether any of it is working in month one?

Judge month one on conversion rate and rating, not on rank. Store impressions to installs responds within days of a screenshot or headline change once there is traffic, which makes it the only ASO number that gives usable feedback inside a launch window.

Track four things weekly. Listing conversion rate, split by traffic source. Average rating and the count behind it. The number of distinct terms you appear for at all, which is a better early signal than position on any single one. And installs per day, since that is the input both ranking algorithms weight.

Change one thing per cycle. The temptation with no data is to rewrite the title, replace the screenshots and adjust the keyword field in the same week, after which nothing that happens can be attributed to anything.

Expect rank to look flat for weeks and treat that as normal rather than as evidence of failure. The first keyword movements usually show up on the long-tail terms nobody else wanted, and those are the ones worth building on.

All of this stops applying once you have install volume. At that point rank and velocity are things you can actually move, the standard ASO playbook starts working, and the metadata you can rewrite freely in week one is carrying rankings you would rather not disturb.

Write down what you changed and when, in a file rather than a chat thread. Three months in, the question you will want answered is which listing change preceded the week conversion rate moved, and nobody remembers.

The short version

ASO for a new app is metadata, screenshots and a plan for the first ratings. Everything the standard advice optimises, meaning rank, velocity and category position, is downstream of installs you do not have yet, so it cannot be worked on directly in month one.

Get above four stars quickly and stop tuning the decimal.

According to AppFollow, around 80% of app store revenue goes to apps rated 4.0 and above, and Apple’s native prompt is the mechanism: Phiture found that 13.5% of users shown it submit a rating against 0.07% who write a review, so fire it after the app has done something useful and accept that written reviews will lag.

Treat the famous “+340% from two to three stars” figure with caution, since it traces to a 2015 Apptentive survey of 350 people about what they would consider installing.

Target two or three long-tail phrases you can plausibly win, put the best one in the title, and check that you rank first for your own name.

Use Apple Ads to appear in search before you can rank, noting the further search placement Apple added in March 2026, and use the traffic it buys to test screenshots. Judge month one on listing conversion rate and rating, not on keyword position, and change one thing at a time.