Machuret / ASO, no agency

ASO services, explained · September 2026

ASO services: what app store optimization actually involves, and what an agency means when it sells it.

Every ASO agency sells the same eleven things under different names, and almost none of them explain what each one really is, how long it takes, or what it should cost. This page does. It is written by an independent consultant who competes with those agencies, and it links to the reviews of every named one and the country-by-country guide. If you are choosing an ASO agency, read this first.

Start with the servicesSkip to what it costs

What ASO services are, in one paragraph

Gabriel Machuret

App Store Optimization is the work of making an app appear for the searches its buyers actually make, and convert when it does. That is the whole discipline. The levers are small and the developer controls most of them: the title, the subtitle, the keyword field, the first screenshot, the icon, the rating, and on Google Play the text the index reads. The signals the developer earns rather than controls, conversion rate, retention, velocity, are what the stores use to decide whether to trust the controlled ones. Everything an agency sells as an ASO service is either a decision about one of those levers, a way of proving a decision was right, or an activity that fills the months between decisions.

The reason ASO services are confusing is that agencies name them for how they bill rather than for what they do. ‘Ongoing optimisation’ is a retainer. ‘Semantic core’ is a spreadsheet. ‘Integrated paid and organic’ is a media desk. ‘ASO 2.0’ is a features list. Below, each service is described by what it actually is, how long it genuinely takes, what it should cost, and the red flags that show you are buying the label instead of the thing.

The eleven ASO services, honestly

What they are called, what they are, what they cost
01

The ASO audit

Sold as: ‘Comprehensive ASO audit’, ‘store presence analysis’, ‘discovery phase’. Often a month, often billed at full rate.

What it actually is

  • A read of your listing against five real competitors in the storefronts your installs actually come from. Not a template with your app name swapped in.
  • A keyword reality check: which terms you rank for, which you could, which are brand, which are noise. With traffic, not just positions.
  • A conversion read: what your first screenshot claims, whether the title says anything, whether the icon survives at thumbnail size, what the reviews say people expected.
  • A verdict. Sometimes the verdict is that the listing is fine and the product, the price or the paid mix is the problem. A good audit says so.
How long it takes

One week. Two if the category is unusual.

What it should cost

Included in a fixed engagement. Agencies bill it as a month of ‘discovery’.

Red flags
  • An audit that reads like every other audit
  • No named competitors
  • Findings without a prioritised list of changes
  • A discovery phase longer than a week
02

Keyword research

Sold as: ‘semantic core’, ‘keyword strategy’, ‘400 keywords tracked’. The most over-sold and least decisive service in ASO.

What it actually is

  • A short list of terms that have traffic, that your buyers actually use, and that you can realistically rank for given your ratings, downloads and competition. Usually twenty to forty terms, not four hundred.
  • A decision about which three go in the title, which in the subtitle, and what fills the keyword field without repeating any of them. Apple indexes those fields; it barely reads the description.
  • On Google Play, a decision about which terms the short and long description should carry naturally, because Play reads text and punishes stuffing.
  • The judgment call: which competitive term is worth six months of patience, and which long-tail terms will bring installs next week.
How long it takes

Three to five days, as part of the teardown.

What it should cost

Tools cost under $100 a month. The judgment is the fee. Agencies bill ‘monthly keyword optimisation’ forever.

Red flags
  • A spreadsheet delivered as the strategy
  • Keyword counts as a deliverable
  • Weekly keyword churn that never lets the stores settle
  • The same terms in title, subtitle and keyword field
03

Metadata: title, subtitle, keyword field, description

Sold as: ‘metadata optimisation’, ‘listing copywriting’. Usually the second month of a retainer and the last time the listing changes.

What it actually is

  • A title that carries your strongest term and still reads like a product. Thirty characters on iOS. Every one is a decision.
  • A subtitle that carries the second claim and the second term, not a repeat of the first.
  • A keyword field with no wasted characters: no repeats, no plurals the store already handles, no spaces after commas, no terms already in the title.
  • A description written for the humans who read it and, on Play, for the index. On iOS, the first three lines before ‘more’ and the promotional text are what matters.
  • Written and handed over ready to paste, with the reasoning, so your team can make the next change without anyone.
How long it takes

One week after the teardown.

What it should cost

Fixed, once. Then nothing until the stores have responded and there is a reason to change it.

Red flags
  • A 3,000-character App Store description presented as the work
  • Keyword density as a metric
  • Metadata ‘refreshed’ monthly regardless of data
  • No explanation of why each term is where it is
04

Creative and conversion: icon, screenshots, preview video

Sold as: ‘creative optimisation’, ‘CRO’, ‘store creative production’. Often the expensive part of an agency retainer because a studio bench is behind it.

What it actually is

  • A decision about the claim in your first screenshot. It decides more installs than every paragraph of text combined. Feature? Outcome? Social proof? Price? One claim, chosen for how your category’s buyer decides.
  • A screenshot order that tells a story in the two or three frames a visitor actually sees before tapping, with captions that read at thumbnail size.
  • An icon that survives at 60 pixels next to your competitors, and that does not look like theirs.
  • A preview video only if your category rewards motion, and only after the message is tested. For most non-game apps, the first screenshot does the work.
  • Local conventions where they differ: Japanese screenshots are dense and reassuring; Western ones are sparse. That is a decision, not a translation.
How long it takes

One to two weeks for the first set. Production by your designer, or any studio, to a brief.

What it should cost

The brief and the hypothesis are the fee. Production is a per-piece purchase from whoever you like.

Red flags
  • Screenshots that share a layout with the agency’s other clients
  • A video commissioned before the first screenshot has a measured conversion rate
  • Beautiful, on-brand frames that say nothing to a stranger
  • Creative ‘refreshed’ before the current set has been tested
05

A/B testing

Sold as: ‘data-driven creative testing’, ‘experimentation programme’. Sometimes a software subscription in disguise.

What it actually is

  • Apple’s Product Page Optimization and Google Play’s store listing experiments, which are free and built into the consoles you already own. Third-party pre-launch testing tools exist and are useful for some situations; they are not required.
  • One variable at a time. One screenshot claim against another, with everything else held still, so you know which change earned the installs.
  • Enough traffic and enough patience to read a result. Small apps need weeks, not days. An agency on a monthly clock will call it early.
  • The discipline to test the hypothesis, not the tool’s suggestion. A test tells you which of two ideas did better. It cannot tell you both were wrong.
How long it takes

Two to six weeks per test, depending on traffic.

What it should cost

Free in the consoles. The hypothesis and the discipline are the fee.

Red flags
  • Testing proposed before anyone has stated what the first screenshot should claim
  • A software seat required to run tests
  • Multiple variables changed in one test
  • Tests called after three days
06

Localisation

Sold as: ‘global ASO’, ‘localisation into 30 storefronts’. A cost multiplier agencies love, because every storefront is another line.

What it actually is

  • First, a decision about which storefronts actually matter: where your installs come from, where they could, and where the competition is beatable. Usually three to six, not thirty.
  • Then, for each, a category read, not a translation: different competitors, different search behaviour, sometimes a different first-screenshot claim. Japanese, Korean and Arabic storefronts in particular are decided, not translated.
  • Then native copy, written after the decisions are made, by someone who searches in that language.
  • Apple’s cross-localisation quirks used deliberately: some storefronts index more than one language, and that is free keyword space.
How long it takes

One to two weeks per priority storefront.

What it should cost

Native copy is inexpensive. Deciding which storefronts and what to say is the fee. Thirty storefronts before the primary one works is waste.

Red flags
  • Localisation proposed before the primary listing has been tested
  • A translated keyword list
  • Thirty storefronts in the proposal
  • Western screenshots with foreign text
07

Ratings and reviews

Sold as: ‘reputation management’, ‘review management’, and in the grey zone ‘rating improvement’. Read the last one twice.

What it actually is

  • A prompt strategy: asking satisfied users at the right moment, inside the app, using the native rating dialogs. Timing is the whole skill.
  • Reply discipline: responding to reviews, especially negative ones, because replies are read by the next visitor and, on Play, affect ranking signals.
  • A loop back to product: the reviews tell you what people expected and did not get. That is a listing problem or a product problem, and the audit should say which.
  • Never purchased, incentivised or exchanged reviews. Both stores prohibit it, detect it, and penalise the developer account, not the vendor.
How long it takes

Ongoing, by your team, after a one-week setup.

What it should cost

Nearly nothing. The prompt logic is a product change; the reply habit is a calendar entry.

Red flags
  • A target star rating with a delivery date
  • ‘Review management’ that shades into review writing
  • Rating campaigns sold by the unit
  • No mention of the in-app prompt
08

In-app events, custom product pages, promotional content

Sold as: ‘advanced ASO’, ‘ASO 2.0’, ‘full-funnel store presence’. Real features, over-labelled.

What it actually is

  • Apple’s in-app events and Google Play’s promotional content: time-boxed features that get their own store placement and can be found in search. Useful for apps with seasons, launches or live-ops. Pointless for a static utility.
  • Custom product pages: alternative iOS listings you can send paid traffic to, so the page matches the ad. This is where paid and organic genuinely meet, and it is the one place ‘paid informs organic’ is more than a pitch.
  • A decision about whether your app has anything to say more than once a quarter. If not, the feature is not for you, whatever the proposal calls it.
How long it takes

A few days per event or page, once the base listing works.

What it should cost

Small, per piece. Not a reason for a retainer.

Red flags
  • ‘ASO 2.0’ as the headline
  • Events proposed for an app with no seasons
  • Custom product pages without paid traffic to send to them
  • Base listing untested while advanced features are sold
09

Apple Search Ads and Google App Campaigns

Sold as: ‘integrated paid and organic’. Adjacent to ASO, not ASO. Where most agency margin actually lives.

What it actually is

  • Paid search on the App Store and Google Play. A media discipline that scales with spend, produces daily numbers, and is billed by agencies as a percentage of budget.
  • Genuinely related to ASO in two ways: the listing decides paid conversion, so a better page lowers every paid install’s cost; and paid keyword data tells you which terms convert before you commit the title to them.
  • Set up properly once: campaign structure, brand-term decision, negative keywords, a bidding discipline your team can run. Then run by your team, not by a retainer tied to spend.
How long it takes

One week to set up and hand over.

What it should cost

Setup is part of an engagement. Ongoing management is a media business; buy it as media if you want it, and never confuse it with organic work.

Red flags
  • ASO priced as an add-on to a percentage-of-spend fee
  • Brand-term bidding recommended before organic brand ranking is checked
  • Organic results reported only inside the paid deck
  • ‘Paid data informs organic’ used to justify a spend increase
10

Reporting and monitoring

Sold as: ‘transparent monthly reporting’, ‘real-time dashboards’, ‘quarterly business reviews’. The service that exists so the retainer can recur.

What it actually is

  • A changelog: what changed on the listing, when, and what search installs and conversion did in the weeks after. One page. Readable by a founder in two minutes.
  • Console data you already own: App Store Connect and Play Console show impressions, product page views, conversion and search installs. No dashboard adds information; it adds formatting.
  • A monthly look, not a daily one. The stores respond over weeks. A daily dashboard on an organic listing is a screensaver.
How long it takes

Ten minutes a month, by your team, after handover.

What it should cost

Nothing. Reporting is a description of work. It is not work.

Red flags
  • Reporting listed as a deliverable
  • A dashboard login as the main handover
  • Ranking movement without install movement
  • A report longer than the list of changes
11

‘Boosts’, guaranteed rankings, keyword installs

Sold as: ‘ASO’, ‘app promotion’, ‘guaranteed top-10’. This is the one service on the list that is not a service.

What it actually is

  • Paid downloads from operated devices and accounts, triggered on a search term, to fake the signal that real people chose your app. Purchased ratings and reviews work the same way.
  • Prohibited by Apple’s guidelines and Google’s policies in plain language. Detected by systems built around install velocity that retention does not support.
  • The outcome, in order: the ranking arrives; retention collapses; the store discounts the signal; sometimes the developer account gets a notice the seller will never see.
  • Cheap by the unit. The most expensive thing on this page once you count the account.
How long it takes

Days to arrive. Weeks to fade. Months to recover from.

What it should cost

Never.

Red flags
  • Installs, ratings or reviews sold by quantity
  • Guaranteed positions within days
  • ‘Safe’ or ‘organic-looking’ described as a feature
  • Payment by card on a self-serve platform, no strategy call
See what it looks like: the ASO World review →

The honest order

Six weeks, no decks
Week 1

Teardown

Your listing, five competitors, your keyword reality. The verdict, including whether the listing is even the problem.

Week 2

Rewrite

Title, subtitle, keyword field, description, creative order. Handed over ready to paste, with the reasoning.

Weeks 3–6

Test

One first-screenshot hypothesis against another, in the free console tools. Everything else held still. Read when the data says so, not the calendar.

Week 6+

Handover

The keyword map, the test log, the prompt strategy, the Search Ads setup, the training. Your team runs the next round.

This is the whole engagement, and it is why one consultant can do it for a fixed fee while an ASO agency needs a retainer: the work has a beginning and an end. The agency vs consultant guide makes the comparison in full, and the proposal guide shows how to spot the eleven services being padded into twelve months.

What ASO services cost

Typical market rates · illustrative, not quotes

The eleven services, bought monthly from an agency

$36k–$180k$3,000–$15,000/mo × 12, depending on tier and countryAbout a quarter reaches the listing; the rest is the agencyMonths four to twelve are mostly reporting

The services that matter, bought once

$9,500Teardown, rewrite, test, handover. Fixed scope, fixed fee.Gabriel, on every deliverableYour team runs the next round without anyoneApply to work with me

Country-specific ranges are on each country page, and the full breakdown of where a retainer goes is in the pricing guide.

Questions about ASO services

“Which ASO services actually move installs?” +

Three: the decision about your first screenshot, the title and keyword field, and a disciplined test that proves the first two. Everything else supports those or exists to fill a retainer.

“Do I need all eleven?” +

No. Most apps need the first five once, the sixth for two or three storefronts, the seventh as a product habit, and the ninth set up and handed over. The tenth is a calendar entry and the eleventh is never.

“What should ASO services cost in total?” +

Done once by a consultant on a fixed scope, in the region of $9,500. Done monthly by an agency, $36,000 to $180,000 a year, of which roughly a quarter reaches the listing. The pricing guide has the ranges by country and tier.

“Is Apple Search Ads part of ASO?” +

Adjacent. The listing decides paid conversion and paid data informs keywords, so they touch. But Search Ads is a media business billed on spend, and an agency that sells both has a stronger incentive on the paid side.

“Can my team do this in-house?” +

After a proper handover, yes. The tools are free or cheap, the consoles are yours, and the discipline is one considered change and then patience. The part that is hard to do in-house the first time is the judgment, which is what the engagement transfers.

Buy the services
not the
subscription.

A limited number of engagements at a time, in one language: what we change, when it ships, what it costs. If your listing is fine and the problem is somewhere else, I will tell you on the first call and point you there.

Replies come from Gabriel, usually within a working day.